'Slow Cinema':
Temporality and Style in Contemporary Art and Experimental Film
Submitted by Matthew Flanagan to the University of Exeter as a thesis for
the degree of Doctor of Philosophy in English in October 2012.
This thesis is available for Library use on the understanding that it is
copyright material and that no quotation from the thesis may be published without
proper acknowledgement.
I certify that all material in this thesis which is not my own work has
been identified and that no material has previously been submitted and approved
for the award of a degree by this or any other University.
(Signature) ………………………………………………………………………………
1
ABSTRACT
This thesis explores a stylistic current within contemporary art and experimental film that will be referred to
here as 'slow cinema'. This type of cinema privileges a number of distinct and recognisable tropes: the
application of the long take, an undramatic narrative or non-narrative structure, a tendency toward realist or
hyperrealist representation, and a pronounced stillness of composition and visual content. This thesis details
the evolution of slow cinema over the course of the last three decades (in line with both dominant and
marginal cultural transitions under neoliberalism), and places its aesthetics in the context of relevant
innovations in post-war modern and experimental film. The first chapter outlines the contemporary shape of
'slow cinema' by describing its most typical characteristics (in relation to narrativity and modes of realism);
the second chapter focuses on the device of the sequence shot and the continuing relevance of the criticism
of André Bazin in relation to contemporary durational film; the third chapter situates slow cinema in the
context of an opposition to the compression of temporality in neoliberal culture; and the fourth chapter
encompasses a study of two recent tributes to the cinema of Yasujiro Ozu, followed by an extended reflection
on the digital regime of contemporary film production. As a whole, this thesis aims to map a set of unique
aesthetic strategies across a number of post-war and contemporary durational films, and to place the field of
'slow cinema' within a suitably broad framework of related film-historical, cultural and socioeconomic
trends.
2
CONTENTS
Title Page 1
Abstract 2
Table of Contents 3
Introduction 4-23
Chapter One: Key Characteristics of Contemporary Slow Cinema 24-62
Chapter Two: The Moderns, Bazin, and the Durational Sequence Shot 63-98
Chapter Three: Undramaticness, Time, and the Neoliberal Economy of Contemporary Cinema 99-135
Chapter Four: The Still Life, Two Tributes to Ozu, and the Digital Regime 136-212
Conclusion 213-216
Works Cited 217-228
3
INTRODUCTION
This thesis examines the emergence of a current within post-war cinema that has, over the course of the last
decade, become increasingly commonly known as 'slow cinema'. The label 'slow cinema' refers to a model of
art or experimental film that possesses a set of distinct characteristics: an emphasis upon extended duration
(in both formal and thematic aspects); an audio-visual depiction of stillness and everydayness; the
employment of the long take as a structural device; a slow or undramatic form of narration (if narrative is
present at all); and a predominantly realist (or hyperrealist) mode or intent. This conception of slowness has
been present in modern cinema since its emergence after the Second World War, but has become increasingly
prevalent as an institutionalised mode of film practice during the last three decades. The focus of this thesis
is primarily the expansion of an increasingly austere form of slow cinema in the last three decades (with an
emphasis on the more populous last two), as well as its earlier evolution from post-war modern and
experimental cinema.
The broader context of modern cinema from which slowness emerges will be assumed here to result
from the radical break in image and sound production effected by the trauma of the Second World War, a
history conforming to Gilles Deleuze's concept of the schism between the movement-image and time-image,
and mapped in sections 1A and 2B of Jean Luc-Godard's Histoire(s) du cinéma (1998). Contemporary slow
cinema is an eventual descendant of the international modern cinema that emerged in the late 1940s, one that
attempted to restore belief in “the tatters of this world” by creating an aesthetic regime that reflected the
post-war struggle to (re)connect with a new reality (Deleuze 2005b: 166). This impulse began with post-war
Italian neorealism and continued most forcefully in European modernist film of the 1950s and 1960s, and the
high modernist, structural and materialist cinema of the 1960s and 1970s. The notion of 'slowness', in this
context, comprises one of the most potent signifiers of the Modern in post-war cinema, and the forms that it
has assumed have been diverse and changeable.
The purpose of this thesis is to provide an extended study of the evolution and unique qualities of
this mode of art and experimental film. It is, equally, intended to contribute to a broader understanding of the
role of duration in formal aspects of recent art and experimental cinema—an approach that necessitates, as I
will propose, a more expansive model of slow cinema than is most commonly circulated at present. The
4
contribution of this thesis to the field of study of recent durational art and experimental film will comprise
four interrelated aspects: a theoretical modelling of the emergence and parameters of post-war modern
cinema (in relation to the writing of André Bazin and Gilles Deleuze); an attention to the film-historical
precedents of contemporary slow cinema (in terms of its break from both classical and modernist cinema); a
survey of slow cinema's position in our hypermediated culture of the neoliberal era; and a focus on key
exponents of its cinematographic style—that is, a discussion of relevant and exemplary filmmakers, films,
individual shots and technologies.
I should note, at the outset, that the label of 'slow cinema' is not strictly the most ideal term to
comprehensively account for the set of stylistic tendencies that will be addressed throughout this thesis. As a
collective term, 'slow' too readily suggests a binary opposition based on speed and motion, and signifies a
range of contemporary films, filmmakers and styles in a manner that might be considered to be excessively
panoptic. I will directly address the distinction between 'fast' and 'slow' in terms of speed (or, rather, in terms
of the compression of time in contemporary culture) in my third chapter, but otherwise I will consider slow
cinema's salient opposition to 'fastness' to be a secondary aspect of the term here.
However, I shall retain the concept of 'slow cinema' throughout this thesis primarily because it has,
over the course of the last decade, become commonly accepted as a broad signifier of a certain mode of
durational art and experimental film (as I will outline in the following sections). No alternative label could be
expected to account for the complex network of stylistic convergences referred to here with absolute
precision, and I still consider 'slow' (in its subtle evocation of temporality and subjective positioning in
relation to the world) to be the most fitting container. Throughout this thesis I will also use the terms
'durational' and 'observational' to refer to the unique qualities of the contemporary cinema under discussion,
as well as other terms that have broader implications and connotations in relation to post-war film, such as
'undramatic' and 'modern'. In my usage of the label throughout this thesis, 'slow' is simply intended to refer to
a field of cinema that shares common traits and aesthetics: an emphasis on the passage of time in the shot, an
undramatic narrative or non-narrative mode, and a rigorous compositional form that is designed for
contemplative spectatorial practice.
In the remaining part of this introduction I will broadly summarise the established basis of slow
cinema that I am drawing from, and aiming to expand, here. The following sections will detail existing
5
considerations of the notion of slowness in post-war modern cinema; the role of the long take in unifying this
trend in recent cinema; its classification in existing criticism (a section that will include a brief literature
review of the field); and a summary of the content and approaches of the chapters to come.
Slowness in post-war film
The notion of a type of art or experimental film that might be considered 'slower' than commercial or
institutional norms is certainly familiar in an informal sense, and tends to conform to a certain logic. That is,
a familiar stock of filmmakers are, in casual discourse, frequently associated with slowness as either a
convenient point of reference or, offhandedly, as a pejorative. The makeshift canon of 'slow' that has emerged
in this unscientific fashion signals work in both narrative (art cinema) and non-narrative (avant-garde,
experimental or underground film and video) traditions that shares, at least, a certain undramatic sensibility
or nascent durational style. For instance, in his article “Is Ozu Slow?” (2000), Jonathan Rosenbaum lists
“Robert Bresson, Carl Theodor Dreyer, Hou Hsiao-hsien, Abbas Kiarostami, F.W. Murnau, Yasujiro Ozu,
Jacques Rivette, Jean-Marie Straub and Danièle Huillet, Andrei Tarkovsky and Jacques Tati” in this context,
whilst Raul Ruiz, in the first volume of his Poetics of Cinema, adds Michael Snow and Andy Warhol to an
otherwise identical list of what he refers to as 'provocateurs' of privileged moments of boredom (Ruiz 2005:
13).
David Campany, after noting the contrast between the uniquely 'fast' cinema of the late silent Soviet
montagists (who, alongside the European surrealists, comprise the vanguard of the first wave of Modernist
cinema), summarises the broadly consensual context from which this familiar canon of 'slow' later emerged:
After the Second World War, European and North American culture began to be dominated by the
ideologies of mainstream cinema, television, lifestyle culture, saturation advertising and mass
distraction. In this new situation speed lost much of its critical edge and most of its artistic
credentials. To be radical in this new situation was to be slow (Campany 2007: 10).
Campany suggests that art, faced with the economic and cultural onslaughts of large-scale post-war capitalist
6
expansion and an intensification of the spectacle of modernisation, opted for slowness as “the only creative
option.” This new sense of temporality soon characterised a set of landmarks in art and film, namely the
cinema of “Vittorio de Sica, Roberto Rossellini, Ingmar Bergman, Robert Bresson, Michaelangelo
Antonioni, Pier Paolo Pasolini, Chantal Akerman, Andrei Tarkovsky, Wim Wenders, Krzysztof Kieslowski,
Aleksandr Sokurov, Béla Tarr, and many others” (Campany 2007: 10-11).
Such representative canons of slowness encompass a broad range of stylistic tendencies, and
predominantly refer, at root, to the formal and institutional categories of art cinema initially advanced by
David Bordwell (1979) and Steve Neale (1981). In Neale's formulation, the institution of art cinema,
predominantly supported by the state, emerged within national cinemas as a reaction to the dominance of
American exports in international exhibition markets since the silent era. As a result, currents within
international art film have varied historically and geographically since they derive in part from another
simultaneous function: “that of differentiating the text or texts in question from the texts produced by
Hollywood. Hence they change in accordance with which features of Hollywood films are perceived or
conceived as dominant at any one point in time” (Neale 1981: 14). One of those clear differentiations has
been a broad tendency toward slowness, a temperament that quite plainly forces a departure from the
strictures of the classical Hollywood film in order to impose a set of new ones in their place.
In David Bordwell's distinction, art cinema is seen to define itself “explicitly against the classical
narrative mode, and especially against the cause-effect linkage of events” (Bordwell 1979: 57). It is an
ambiguous cinema that opts to cast narratives and characters adrift within an aesthetic framework that
incorporates both the objectivity of documentary realism and an intense subjectivity of psychological
breakdown: as a rule, “the art cinema is less concerned with action than reaction; it is a cinema of
psychological effects in search of their causes” (Bordwell 1979: 58). Perhaps most significantly, the art
cinema for Bordwell is realist, emerging in the context of Italian neorealism and remaining strongly
committed thereafter to spatial and sonic representations of material reality, and analogue realisations of the
passage of time. This context sets up Mark Betz's quip that, in recent decades, “the sum produced by adding
the variables 'time' and 'art film'” has, frequently, been the generic label “'slow'” (Betz 2009: 5).
Recent durational film has its roots in this tendency toward realism and slowness within art and
experimental film between the late 1940s (Roberto Rossellini's Germania, anno zero and Luchino Visconti's
7
La Terra Trema [both 1948] are probably the most influential works from this period) and the end of the
1970s (centring, during the course of that decade, on work by Chantal Akerman, Marguerite Duras, JeanMarie
Straub and Danièle Huillet and Andrei Tarkovsky, amongst others). This tendency toward slowness
deepened during the 1990s, as watershed films like Béla Tarr's Sátántangó (1994), Tsai Ming-liang's Vive
l'amour (1994) and Fred Kelemen's Frost (1997) appeared, and during the last decade it has become more
prominent still (as I will document later). It might be argued, however, that a great many—if not all—of slow
cinema's foundational stylistic tendencies (its formal signifiers of slowness) were introduced long before the
contemporary cinema—that is, between Italian neorealism and the emergent structural and materialist
concerns of the 1960s and 1970s.
Since the turn of the millennium, marginal filmmakers (alongside the rather anomalous Gus Van
Sant, who has, before and after his 'Death Trilogy' of Gerry [2002], Elephant [2003] and Last Days [2005],
been accustomed to working in a more commercial capacity) have begun to adopt the aesthetic mannerisms
of slowness at a steadily increasing rate. Alongside an endurance of experimental practice, this development
has given rise to an institutional form often referred to as the 'festival film'—a type of art film that has
become somewhat disconnected from national arthouse distribution (albeit less so in certain European
countries such as, for instance, France or Austria) because it is commonly considered too austere or rarefied
for wider distribution. Adrian Martin has referred to the designation of the 'festival film' as “a truly hideous
term,” not least because it signifies that the destiny of the field of slow film is to be publicly projected only
in 'ghettoes' of international festivals rather than the nominally more accessible local arthouse circuit or more
commercial venues. Martin has noted that, as a result, the work of key contemporary filmmakers such as
Philippe Garrel, Pedro Costa and Apichatpong Weerasethakul “are branded, virtually from the outset of their
public life, as unfit for general (or even moderately specialised) distribution and exhibition” (Martin 2009:
105). During the course of the last decade, the concepts of 'slow cinema' and the 'festival film' have often
become intertwined in popular discourse, circulating as a shorthand for both a certain type of unassuming,
minimalist film often entirely restricted to the festival circuit (or very limited national distribution), and a
more challenging, austere mode of durational art or experimental film marginalised even in that context too.
The most prominent filmmakers whose work—in part or in sum—has in some way shaped and
influenced the field (and practice) of what I am referring to as slow cinema include Chantal Akerman,
8
Lisandro Alonso, Theo Angelopoulos, Apichatpong Weerasethakul, Sharunas Bartas, James Benning, Pedro
Costa, Lav Diaz, Philippe Garrel, Hou Hsiao-hsien, Peter Hutton, Jia Zhang-ke, Fred Kelemen, Abbas
Kiarostami, Liu Jiayin, Sharon Lockhart, Raya Martin, Albert Serra, Aleksandr Sokurov, Jean-Marie Straub
and Danièle Huillet, Béla Tarr, Tsai Ming-liang and Gus Van Sant. Although some of these filmmakers have
been active since the 1960s, refining or rethinking their particular aesthetics up to the present day (or, for
Danièle Huillet and Theo Angelopoulos, 2008—the year when their final works appeared), some of the
younger filmmakers listed (such as Lisandro Alonso, Liu Jiayin, Raya Martin and Albert Serra) have mostly
started filming and exhibiting work in the last decade. Taken together, this diverse group of filmmakers are
connected only by a tendency toward undramaticness or realism by way of variants on a durational long-take
style, but their particular approaches collectively represent a formidable aesthetic achievement. On the basis
of this particular grouping, slow cinema represents a current within modern and contemporary cinema that
extends beyond the informal canon of slowness suggested by Campany and Rosenbaum, and has entailed a
more rigorous and attenuated approach to duration than the vast majority of post-war art cinema.
The long take
A key formal property that binds together the field of slow cinema, alongside a preference for the subversion
or dissolution of classical narrative structure (already common in art and experimental film), is what Tiago
Magalhães de Luca has identified as “the hyperbolic application of the long take” (de Luca 2011: 21). This
central feature, in de Luca's description, gives rise to a mode of realism in which “spatial and temporal
integrity is preserved to hyperbolic extremes. As a result, in spectatorial terms, narrative interaction is
dissipated in favour of contemplation and sensory experience,” and the spectator is “invited to adopt the
point of view of the camera and protractedly study images as they appear on the screen in their unexplained
literalness” (de Luca 2011: 23-24). The long take, when employed for this purpose, enables a principally
realist display of actions or incidents in their entirety (often unspectacular, or pertaining to the everyday);
extended fixed studies of objects, places or bodies; and swathes of temps morts or temporal voids that occur
before, during and after an event around which a particular shot is organised.
Undramatic long takes might be used to structure an unconventional form of narration (which is
9
always distended rather than compacted); be punctuated by fragmented narrative activity; or simply stacked
end to end, divorced from narrative relations. A combination of these strategies has often led to extreme
departures from dominant practice in classical or art film, and a small number of feature-length durational
films consist of abnormally few shots that focus on subjects (in particular, landscapes and meteorological
phenomena) that are not commonly afforded close attention in other modes of film practice. Abbas
Kiarostami's Five: 5 Long Takes Dedicated to Yasujiro Ozu (2003), Sharon Lockhart's Double Tide (2009)
and James Benning's Ten Skies (2004), 13 Lakes (2004) and Ruhr (2009) are key examples of this trend,
whilst, even more radically, Benning's Nightfall (2011)—a stationary real-time study of light changing from
day to night in a forest high up in California’s Sierra Nevada mountain range—consists of a single ninety
minute shot.1
It will therefore be assumed here that all recent slow films feature average shot lengths (ASLs) that
vastly exceed the norms of both international mainstream and art cinema. Although it might not be an
entirely appropriate benchmark, David Bordwell notes that representative ASLs within commercial
American cinema have steadily decreased since the immediate post-war period in which William Wyler,
Alfred Hitchcock, Otto Preminger and Vincente Minnelli adopted long takes as viable stylistic options within
a mature classical system. In the late 1940s and 1950s, ASLs in Hollywood studio films averaged nine to
eleven seconds, only to steadily decline during the following two decades (Bordwell 2005: 150). By the
1980s, such rates had effectively vanished from mainstream cinema, and ASLs measuring double rather than
single digits, such as Steven Soderbergh's remake of Andrei Tarkovsky's Solaris (2002), are now extremely
rare (Bordwell 2006: 122-123).
By contrast, undramatic art cinema has tended towards lengthier ASLs since significant films such as
Michelangelo Antonioni's Cronica di un amore (1950) and Carl Theodor Dreyer's Ordet (1955) (of 32.8 and
64.1 seconds, respectively) as well as during more experimental currents such as the political modernism of
Jean-Luc Godard, Nagisa Oshima and Miklós Jancsó that emerged in the 1960s and 1970s. Representative
ASLs for exemplary slow films vary, but a selective sampling that I suggested as a guideline in an article
published in 2008 is still applicable: “35.1 seconds in [Carlos] Reygadas' Silent Light (2007), 35.7 in Albert
Serra's Honor de cavalleria (2006), 65.1 in Gus Van Sant's Gerry (2002), 66.7 in Hou Hsiao-hsien's Café
1
This is not as uncommon a strategy within experimental and underground film as might be assumed, see MacDonald
2008: 240-254.
10
Lumière (2003), 136.6 in Paz Encina's Hamaca paraguaya (2006), 151.4 in Béla Tarr's Sátántangó (1994)
and 884.8 in Abbas Kiarostami's Five: 5 Long Takes Dedicated to Yasujiro Ozu (2003)” (Flanagan 2008). Of
course, de-contextualised ASLs tell us little of the specific formal design of individual films, and are
frequently open to distortion—for instance, both Serra's Honor de cavalleria (2006) and Tsai Ming-Liang's
Goodbye, Dragon Inn (2003) contain very brief sequences of jump cuts and rapid montage that significantly
reduce their overall metrics, skewing a glancing impression of their overall rhythmic design—but they do
provide a useful indication that unbroken duration (within the shot) is a dominant concern within certain
works.2
This preference for extended duration distinguishes slow cinema from other key works of post-war
modern cinema. In relation to the canons advanced by Rosenbaum and Campany above, it is the undramatic
long-take cinema of Carl Theodor Dreyer, Michelangelo Antonioni and Andrei Tarkovsky (alongside the
influence of experimental filmmakers such as Andy Warhol and Michael Snow) that provides the genesis of
the recent slow cinema, rather than the reflective, patient styles of, for instance, Robert Bresson or Krzysztof
Kieslowski. The latter do not tend to pursue extended duration as a formal strategy, and, in that sense, the
editing of observational elements in their work might be considered to be less forbidding. The stylistic
tendencies of the 'Bressonian' strand of art cinema (a lineage traced by Yvette Bíró in her essay on minimalist
aesthetics in contemporary film, “The Fullness of Minimalism” [2006]) is perhaps best outlined by Susan
Sontag in her analysis of “reflective art” in “Spiritual Style in the Films of Robert Bresson” (originally
published in 1964). Sontag suggests that Bresson's 'detached' method is designed to “induce a certain
tranquility in the spectator, a state of spiritual balance that is itself the subject of the film,” and notes that his
work resists the creation of empathy by mediating emotional power to provoke “disinterestedness” and
reflection (Sontag 2009: 180, 177). Although Bresson's narrational style is often strongly linear and “antidramatic,”
it is important to note that its application is particularly “emphatic,” and reliant on “poetic”
montage techniques that actively compress rather than distend shots and sequences (Sontag 2009: 179, 183).
2
All other ASLs throughout this thesis have been taken by myself, and adjusted for factors such as PAL speed-up
(converting 24 to 25 fps) if the viewing source has been a DVD or AVI (Blu-ray, fortunately, should retain 24 fps as
a rule). It is perhaps worth noting that I have been surprised at the discrepancy between some of my own
measurements of certain films and those by serial shot-counters such as David Bordwell, and I have also been wary
of trusting (without reservation) other sources such as the crowd-sourced Cinemetrics database:
http://www.cinemetrics.lv/
11
Fig. 1: Honor de cavalleria (2006)
The undramatic long takes of the contemporary 'festival film' have become somewhat notorious,
gaining a reputation for their rarefied excesses of duration in relation to dramatic incident. As a result,
numerous festival screenings have seen audiences leave partway through films due to impatience or
befuddlement at unfamiliar stylistic methods, and the films' insistence on duration over action has at times
been considered to be overtly confrontational.3
Anecdotal tales of the reaction afforded to Pedro Costa’s
Colossal Youth at its first screening at Cannes in 2006 recall those of the whistling and catcalls that echoed
after the première of Antonioni's L'Avventura at the same festival in 1960—which, as Mark Betz notes, also
involved “cries of 'Cut!'” during the film's long takes (Betz 2009: 5-6)—and many other slow films have
become equally renowned for their spectatorial challenge.
A few shots in Albert Serra's second and third feature-length films, Honor de cavalleria and El cant
dels ocells (2008), have provided oft-cited examples of this. Mark Peranson has described the first screening
of Honor de cavalleria at the Quinzaine des Réalisateurs in 2006 as “infamous and potentially legendary”
because of the audience's reaction to the film's opening sequence, which consists of eleven shots (in as many
3
This seems to be the implication of Nick James' much-quoted Sight and Sound editorial, “Passive Aggressive,” in
which he averred that, in films that dwell “too much” on landscape and their milieu, “there are times, as you watch
someone trudge up yet another woodland path, when you feel an implicit threat: admit you’re bored and you’re a
philistine” (James 2010: 5).
12
Fig. 2: El cant dels ocells (2008)
minutes) that chart the onset of night with minimal incident. The opening sequence of Honor (an
experimental adaptation of Cervantes' Don Quixote) is emblematic of the tone of the remainder of the film:
instead of the introduction of a linear narrative framework, we watch Don Quixote rest, looking toward the
sky as night falls, whilst Sancho Panza ambles in and out of the frame, aimlessly following an instruction to
find a laurel wreath [Fig.1]. The camera, when not static, is as aleatory in its slow movements as Sancho's
body, and the increasingly poor light slowly obscures the content of the image. Peranson recalls that, at an
early morning screening at Cannes in 2006, “as the light gradually dims over the lengthy real-time shot,
which approaches absurdity but skirts it ever so slightly, the audience’s reaction became more and more
hostile. Uncomfortable murmuring spawned the familiar sounds of seats slamming, rustling of bags, and,
soon enough, mass exodus” (Peranson et al 2007: 20). Echoing this not uncommon collective reaction, Doug
Cummings recalls a screening of Serra's film during the Palm Springs International Film Festival at which,
by the end, he was the only spectator remaining in the auditorium (Cummings 2007).
Jonathan Romney, in his survey of contemporary slow cinema in 2010, suggests that one shot from
Serra's El cant dels ocells would suffice to lampoon this “varied strain of austere minimalist cinema that has
13
thrived internationally over the last ten years”: the eight-minute sequence shot that follows the Magi (the
film's few subjects) traversing a desert landscape by staggering into the distance only to return a few minutes
later (Romney 2010: 43). The shot begins with a medium-long framing of the three figures of the Magi,
whose gradual retreat into the distance is observed from a static camera position for five minutes until they
disappear over the horizon [Fig. 2]. Re-emerging atop a ridge that divides earth and sky, they circle back
toward us, and the shot ends only when they have regained half the ground between the horizon and the
camera. The long take here presents us with a vast block of time, clearly directionless and free of dramatic
intent, but with a series of gradual modulations upon which we can train our perceptive capacity. By any
putative standard of classical narration, the shot is purposely limited in its content and simply too long,
meaning that the intensity or indifference of our gaze in contemplating its spatial and temporal aspects
assumes a prominent role. Romney notes that such shots “highlight the viewing process itself as a real-time
experience in which, ideally, you become acutely aware of every minute, every second spent watching”
(Romney 2010: 43).
This rarefied long-take style has been referred to by Adrian Martin as “degree-zero” filmmaking
(Martin 2009a), in a reference to Noël Burch's distinction of the “'zero point of cinematic style'”—one that
derives from a supposedly “erroneous” realist impulse to deny the basic functions of editing (initially in
opposition to the discontinuous découpage of Soviet montage). For Burch, this 'zero point' became desirable
after the introduction of sound film gradually drew attention to the medium's realist potential, and has been
exemplified by certain directors (Luchino Visconti, Michelangelo Antonioni and Alfred Hitchcock, in works
such as Rope [1948] and Under Capricorn [1949]) who “began wondering whether cuts were necessary at
all, whether they should not purely and simply be eliminated or used very sparingly and endowed with a very
special function” (Burch 1973: 11). In Martin's categorisation, the “deliberately drained, stretched-out”
works of El cant dels ocells, Lisandro Alonso's Liverpool (2008) and Lav Diaz's Melancholia (2008)
represent the contemporary exponents of this zero degree, finding themselves “frustratingly self-deprived of
aesthetic resources” after privileging, at all costs, duration over découpage (Martin 2009a).
14
Classifications of slow cinema
Romney's definition of contemporary slow cinema focuses on this degree-zero mode, where “what's
primarily at stake is a certain rarefied intensity in the artistic gaze, whether the images are polished […] or
frugally rough-edged” (Romney 2010: 43). Alongside Romney's useful definition, alternative categorisations
of the basis of slow cinema have included my own tentative formulation 'aesthetic of slow' (Flanagan 2008),
and 'contemplative cinema', a term established on the blog Unspoken Cinema (a website that has served as a
valuable discussion point and resource hub since 2006).4
The construct of contemplative cinema is similar to
the now more widely accepted epithet of 'slow', and has been defined by Doug Cummings as “obviously a
vague term,” but one that clearly signifies “films that resist conceptualization and push beyond words and
thoughts toward silence and meditation.” Engaging with the works of contemplative cinema is, in this sense,
“less about amassing their information and articulating their meanings than sharing their sights, sounds, and
rhythms in deeply experiential ways” (Cummings 2007).
More selectively, de Luca has identified a current of “sensory realism” in contemporary art cinema,
pointing to a grouping of films that are “equally interested in foregrounding material phenomena as enabled
by the long take” (de Luca 2011: 24). Like slow and contemplative cinema, de Luca notes that sensory
realism is “not a structured film movement,” but rather signifies a tendency toward emphasising the sensory
qualities of the filmic image and its subjects (in particular, the physical bodies of actors and material
contours of the world) in preference to overt dramaticity or psychological motivation. His approach focuses
on materiality and carnality in the work of contemporary filmmakers that make visible “geographies of the
earth and geographies of the flesh, and the crossroads at which these geographies overlap,” namely Carlos
Reygadas, Tsai Ming-liang and the 'visionary realism' of Gus Van Sant (de Luca 2010). De Luca suggests
that this tendency extends toward the work of Lisandro Alonso, Apichatpong Weerasethakul, Nuri Bilge
Ceylan, Pedro Costa, José Luis Guerin, Jia Zhang-ke, Abbas Kiarostami and Béla Tarr as well (de Luca
2011: 24).
Other than these varied approaches to aspects of recent slow cinema, however, critical writing on its
configuration has thus far been somewhat piecemeal. Articles have appeared in print and online magazines
4
See: http://unspokencinema.blogspot.fr/.
15
(Fiant 2004; Bíró 2006; Balsom 2007; Flanagan 2008; Bingham 2008; Koehler 2008b; Romney 2010) far
more often than peer-reviewed journals (Schoonover 2012), and attention has otherwise been concentrated in
interviews with filmmakers, blog entries (see, for instance, Bordwell 2011), and one monograph (Yvette
Bíró's Turbulence and Flow, published in 2008). Significant attention has been given to exemplary
filmmakers in the form of book-length studies, and notable English-language texts include Andrew Horton's
The Films of Theo Angelopoulos: A Cinema of Contemplation (1999), Jonathan Rosenbaum and Mehrnaz
Saeed-Vafa's Abbas Kiarostami (2003), James Udden's No Man an Island: The Cinema of Hou Hsiao-hsien
(2009), and the Austrian Film Museum's two exemplary volumes in their FilmmuseumSynemaPublikationen
series on James Benning (2007) and Apichatpong Weerasethakul (2009).
It is perhaps worth mentioning here that slow cinema has received critical attention in curatorial form
too: the 2012 edition of the biennial AV Festival (which took place in Newcastle during the entire month of
March) was titled As Slow As Possible, and proposed “a slower pace and relaxed rhythm to counter the
accelerated speed of today.”5
The programme, directed by Rebecca Shatwell, foregrounded visual and sonic
arts that aim to distend time and perception, and included an expansive film section that showcased work by,
amongst others, Lisandro Alonso, James Benning, Fred Kelemen, Sharon Lockhart, Ben Rivers, as well as a
selective retrospective of Lav Diaz's recent digital cinema curated by George Clark. This broad model of
slowness chosen by Shatwell, which is expressly open to both narrative and non-narrative durational film, is
one that I will retain throughout this thesis.
Similarly, in 2007, Alexander Horwath curated an extensive film-historical programme called
Zweimal Leben / Second Lives for that year's edition of documenta (12) at Kassel. Between 16th
June and 23rd
September, ninety-eight films were screened on fifty occasions (in various pairs and permutations), and
Horwath's programme opens with two narrative features that signify the post-war divergence between
classical and modern cinema (John Ford’s The Sun Shines Bright and Roberto Rossellini’s Viaggio in Italia
[both 1953]), and concludes with a series dominated by durational and observational films: Jia Zhang-ke’s
Platform (2000), Lisandro Alonso’s La libertad (2001) and James Benning’s non-narrative casting a glance
(2007) (for a full listing, see Cherchi Usai et al 2008: 138-140). The structure of Horwath's programme
5
The title is a homage to John Cage's Organ² / ASLSP, a work that was written in 1987 and is currently being
performed at St. Burchardi church in Halberstadt, Germany. This organ performance began in 2001, and is
scheduled to end in 2640. For more information on the AV festival itself, see: http://www.avfestival.co.uk/.
16
attempts to map, with a broad, continuous 'line of communication', the convergences between realist art,
experimental and dominant styles across fifty-four years of modern and recent cinema. Horwath's approach
to the fluidity of aesthetics across film history (in relation to contemporary cinema) echoes the basis of my
extended focus on precedents to slow cinema in Chapters Two and Four.
Other than this curatorial attention and the few print publications cited above, the emergent body of
criticism on recent slow cinema has been predominantly digital in nature, and characterised by its recurrent
presence in forums that have arisen from new online cultures of cinephilia (thanks to the proliferation of
blogs and online journals over the course of the last decade) and a cultic marginality elsewhere
(unsurprisingly, given the limited post-festival visibility and 'underground' digital circulation of the films
themselves). One product of the piecemeal nature of this discourse was a brief debate in the summer of 2010
sparked by a downbeat Sight and Sound editorial by Nick James (see, as representative, Fox 2010; James
2010; Shaviro 2010b), which spilled over from print to an online discussion that centred predominantly on
issues of taste in contemporary art film rather than other aspects (and pleasures) of slow cinema.
One aspect of that brief debate worth noting here, as it reflects a common critique, is the acceptance
of a simple binary opposition between international slow cinema and commercial American film. Whereas
James suggests that slow cinema has represented, for some time, a “clear alternative to Hollywood,” Steven
Shaviro has erroneously argued that slowness has not reflected or embodied the recent social, political and
technological changes of neoliberalism as “a way of saying No to mainstream Hollywood’s current fast-edit,
post-continuity, highly digital style, simply by pretending that it doesn’t even exist” (James 2010: 5; Shaviro
2010b). Slow cinema is here, as often elsewhere, pitched as the exemplification of art cinema's institutional
role as arbiter of the problem of Hollywood, adopting a self-consciously minimalist aesthetic as a direct
rebuke to profligate digital spectacles and intensified continuity style.6
It is true that a clear refusal of the Hollywood model is often expressed in interviews by durational
art and experimental filmmakers predominantly in terms of binary models: fast/slow, more/less, and so on.
For instance, Peter Hutton has spoken of wanting to resist the pace of change in the visual language of
commercial cinema in recent decades:
6
David Bordwell's analysis of intensified continuity in The Way Hollywood Tells It (2006) remains the definitive
interpretation of style in contemporary mass-market film, although Steven Shaviro has recently identified a more
radical current that has now disrupted it: 'post-continuity', which I shall address in Chapter Three (see Shaviro
2010a; 2012).
17
Years ago, when I switched over to an electric camera, I was super-conscious of the fact that the
velocity of the culture was being ramped up by MTV, by the introduction of computers in the home,
by a variety of developments. Even Hollywood films are cut with a shorter and shorter sense of time
and duration; there’s a tendency to try and cram more and more 'information' into a certain concept.
When you feel that your culture is being over-accelerated, there’s a tendency to want to go against
that tide (MacDonald 2008: 218).
More recently, Apichatpong Weerasethakul has expressed a desire to resist Hollywood's sense of spectacle
and return to a more minimalist, primitive aesthetic: “What we think of as a movie now is very complicated.
If we were to show Hollywood films like Avatar [2009] to people thirty or forty years ago, they might find
them impossible to understand. The cut is so quick, the logic is so different; it'd be too mysterious for the
past, too experimental.” Apichatpong has stated that he wanted, when filming Uncle Boonmee Who Can
Recall His Past Lives (2010), “to go back to something very simple, about transformation and saying
goodbye, with simple language” (Gruyaert et al 2010).
This clear retreat from the Hollywood model is misleading, however, if used as a mere definitional
contrast, not least because Hollywood represents only one centre of the cinema that almost all of the
filmmakers under discussion here are working in exile from. As David Bordwell has pointed out, the
intensified continuity style favoured by contemporary Hollywood is in fact “the baseline style for both
international mass-market cinema and a sizeable fraction of exportable 'art cinema',” recasting the dichotomy
on a less easily quantifiable transnational scale (Bordwell 2006: 152). It is true that, as in geopolitics, North
America maintains an aggressive role in the hegemony of global film production, but exaggerating the root
cause of a Hollywood/other binary risks mischaracterising the complexity of the field of slow cinema.
In the context of locating a tendency of sensory realism in contemporary art film, de Luca rejects any
“lumping together” of disparate filmmakers in resistance to dominant cinema, and stresses the importance of
recognising and analysing the “entirely dissimilar traditions” to which contemporary realist projects are
indebted (de Luca 2011: 25, 26). In a similar manner, András Bálint Kovács suggests that, in reference to the
development of modernist art cinema between 1950 and 1980, the reason why it is not possible to speak of
18
Fig. 3: The Sky, the Earth, and the Rain (2008)
one single modernist film style in that period is due to the sheer breadth of styles that derive from “the
diversity of artistic and cultural references modern films use to construct their forms” (Kovács 2007: 122).
The post-war encounter between modern art and different cultural backgrounds in cinema gave rise to a
fullness of style too diverse and fragmented to be boiled down to a stylistic model defined by its oppositional
institutional role, and a similar dynamic is still at work in what I am referring to as slow cinema here. My
aim is not to suggest that the films being discussed as 'slow' can be defined solely by their opposition to
dominant cinema, or to deny that, when considered collectively, they belong to divergent traditions (in
geographic, cultural and economic terms). There is certainly a clear divergence between the production of
duration in contemporary slow and mass-market cinema, and I will address that issue with regard to the
neoliberalisation of contemporary culture in Chapter Three. However, my purpose for employing a collective
differential term here is predominantly to account for the network of convergences in realism and style
across a particularly rigorous type of undramatic art and experimental film, and to understand the
increasingly central role of duration in the modern cinema of the post-war era.
In most of the essays and articles on slow cinema cited above, narrative art film is considered to be
the primary container of undramaticness and durational style in contemporary film. The 'festival film' is
19
perhaps the most populous component of slow cinema, and has come to signify a prescriptive de-dramatised
mode of art cinema rather than a type of film that is exclusively funded and exhibited by an international
network of public and private cultural institutions. In discussion with Simon Field in 2007, James Quandt
suggests that the 'festival film' has emerged as a result of the manner in which prestigious funding bodies (for
example, Rotterdam’s Hubert Bals Fund) and the curatorial leanings of major festival boards have nurtured a
somewhat restrictive aesthetic formula. Quandt picks on José Luis Torres Leiva’s The Sky, the Earth, and the
Rain (2008) as a prime example, and suggests that Torres Leiva's film
sends troubling signals of the emergence of an international arthouse-festival formula, variant from
film to film but adhering to an established set of aesthetic elements: adagio rhythms and oblique
narrative; a tone of quietude and reticence; an aura of unexplained or unearned anguish; attenuated
takes; long tracking or panning shots, often of depopulated landscapes; prolonged hand-held follow
shots of solo people walking; slow dollies to a window or open door framing nature; a materialistic
sound design; and a preponderance of Tarkovskian imagery (Quandt 2009b: 76) [Fig. 3].
This “international arthouse-festival” grouping will no doubt include more than a few of the films and
filmmakers that I have cited so far—as well as, if adapted slightly, a host of other filmmakers (whose
relevant works follow in brackets): Nuri Bilge Ceylan (Uzak [2002], Climates [2006] and Once Upon a Time
in Anatolia [2011]), Benedek Fliegauf (Forest [2003], Dealer [2004] and Milky Way [2007]), Pablo Giorgelli
(Las acacias [2011]), Vimukthi Jayasundara (The Forbidden Land [2005]), Semih Kaplanoglu (for his
'Yusuf's Trilogy': Yumurta [2007], Süt [2008] and Bal [2010]), Cristi Puiu (The Death of Mr. Lazarescu
[2005] and Aurora [2010]), Marc Recha (Pau and His Brother [2001] and August Days [2006]), Carlos
Reygadas (Japón [2002], Battle in Heaven [2005] and Silent Light [2007]), and Andrei Zvyagintsev (The
Banishment [2007] and Elena [2011]).
Although almost all recent slow films feature primary factors that Quandt's formula lists (oblique
narrativity, a tone of quietude and reticence, attenuated takes, long tracking or panning shots of depopulated
landscapes, and so on), I will consider its broad base on more expansive terms here, and adopt an approach
that extends to more austere and rigorous trends in both art and experimental film.
20
Summary of chapters
My aim, in the main body of this thesis, will not be to attempt a fully comprehensive history of slow cinema,
or to analyse the durational style of the entire canon of relevant filmmakers. Such an approach would prove
impossible without spilling over into several volumes, and is a task best reserved for the singular study of
individual filmmakers, national cinemas, or artistic schools (including the few extended studies of
contemporary filmmakers mentioned already). My aim, throughout the following chapters, will be to
contextualise a number of aspects of, and approaches to, the durational and undramatic turn in post-war art
and experimental cinema, and its acute manifestation in contemporary film. I will refer to films (and specific
shots in films) as both concrete instances of style and examples of broader trends, with the aim of giving a
sense of the fluidity of relations that underpin contemporary durational cinema. The breadth of this approach
will, of course, have its limits, but I hope it can serve as an adequate, if potted, map of the rigour and
pleasures of durational style in both contemporary films and their antecedents.
The chapters that follow are intended to contribute to this broad mapping of the evolution and style
of the field of slow cinema. My approach encompasses an analysis of both narrative (art) and non-narrative
(experimental) film, and will attempt to bridge the gap between the two modes: the former will be addressed
predominantly in Chapters Two, Three and the first half of Chapter Four, and the latter will be given
significant attention in Chapter One and the second half of Chapter Four.
My first chapter will outline the most distinctive aspects of the durational turn in contemporary
cinema. This chapter is intended to give a comprehensive sense of the broad dynamics of slow cinema, and
to provide a basis for the film-historical approaches and detailed analyses of individual films that will follow
in later chapters. The first section of Chapter One will focus on recent convergences between elements of
experimental, art, fiction and documentary cinema, and the heightened emphasis on the time and autonomy
of the shot that is evident in new 'hybrid' forms within slow cinema. The second section will address a
concept common to slow films and cinephiliac discourse: an observational attention to 'the wind in the trees',
which has been adopted in certain circles of recent film criticism and scholarship. The third section examines
a cinema that I consider to be particularly significant (and influential) in the context of contemporary
slowness: North American structural film of the 1960s and 1970s, and its embrace of an unmediated
21
durational practice partly inspired by the rediscovery of early films by Auguste and Louis Lumière in that
period. This section will conclude with a reappraisal of Andy Warhol's Empire (1964), and an extended
analysis of a contemporary film: James Benning's non-narrative Ruhr (2009).
My second chapter comprises two aims: firstly, to contextualise the base of modern cinema from
which slow cinema emerges (alongside the experimental tradition addressed in Chapter One) and, secondly,
to return to the criticism of André Bazin as a means of understanding the nature of realism in the field of
slow cinema. My contention here is that Bazin's theory (and model of modern cinema) continues to be
integral to a contemporary understanding of the realist long take, and parts of this chapter will assess Bazin's
continued relevance to contemporary film studies. After a study of realism in modern cinema, the chapter
will close with two sections that focus on the prevalence of a Bazinian (aesthetic) intent on the part of
contemporary filmmakers, and an extended analysis of two Bazinian sequence shots in Jia Zhang-ke's
Platform (2000) and Lav Diaz's Death in the Land of Encantos (2007).
In Chapter Three, I will address the issue of temporality in contemporary culture on terms that most
resemble an oppositional fast/slow binary. The chapter's first section will outline the operation of undramatic
narrativity in contemporary cinema, as well as slow cinema's rejection of the concept of 'central conflict' (a
term formulated by Raul Ruiz in his Poetics of Cinema [2005]). In order to address these tendencies in some
detail, the section will include a few examples of distinctive everyday films: Tsai Ming-liang's Vive l'amour
(1994), Lisandro Alonso's contemporary neorealist work La libertad (2001), and Liu Jiayin's formalist
Oxhide (2005) and Oxhide II (2009). The rest of the chapter will then situate slow cinema in the context of
the production of duration in neoliberal culture, and compare the tendency towards stillness in slow cinema
with the accelerationist aesthetics of recent post-continuity film (see Shaviro 2010a; 2012), and a wider
compression of temporality evident in an era marked by the increasing dominance of financial capitalism.
The purpose of this section is not to propose that slow cinema explicitly adopts a stance of resistance to
accelerated aspects of postmodern culture, but that its unique qualities do in fact reflect on, and are indeed
produced by, the same set of socioeconomic conditions as dominant 'fast' cinema. The chapter will then
conclude with a brief consideration of the endurance of modernist aesthetics in the postmodern era.
My fourth chapter is the broadest in scope, and is divided into four sections. The first section focuses
on Deleuze's concepts of the time-image and the 'still life' in modern cinema, and their basis for two recent
22
films that pay tribute to the use of the latter device in Yasujiro Ozu's work: Hou Hsiao-hsien's Café Lumière
(2003) and Abbas Kiarostami's Five: 5 Long Takes Dedicated to Yasujiro Ozu (2003). The second and third
sections will then be dedicated to two lengthy analyses of these films, which I have selected as representative
of the diversity of form in contemporary durational cinema: Hou's film features an undramatic long-shot
long-take narrational style, whereas Kiarostami adopts a minimalist, non-narrative digital aesthetic for Five.
The final section will address the implications of Kiarostami's adoption of digital technology for a highly
attenuated durational aesthetic, and will conclude with an analysis of Sharon Lockhart's Double Tide (2009)
(as a counterpart to Kiarostami's Five), and a consideration of the differences between analogue and digital
practice in the work of a number of contemporary filmmakers, including James Benning, Pedro Costa and
Lav Diaz.
I have, in part, chosen to analyse certain films here to compensate for the limited attention that they
are likely to receive in other fields of film criticism and scholarship: for instance, James Benning's Ruhr, Lav
Diaz's Death in the Land of Encantos, and Sharon Lockhart's Double Tide. To provide a visual indicator of
the composition of images in these works, which are not always easily accessible by legal means of
distribution, I have added a few screen captures throughout the text, signalled by the notation [Fig.].
23
CHAPTER ONE
KEY CHARACTERISTICS OF CONTEMPORARY SLOW CINEMA
The following chapter consists of three sections that address key characteristics of the field of durational
cinema that I am referring to here as 'slow'. The first section addresses the convergence between different
representational modes (narrative and non-narrative, art and experimental, fiction and documentary) in recent
durational cinema; the second focusses on the notion of 'the wind in the trees' as a means of considering
observational cinema's unique attention to minor or material detail in the cinematographic image; and the
final section looks at the role of duration in North American experimental film of the 1960s and 1970s, and
its relation to contemporary experimental and non-narrative film.
Narrativity and fictionality
One of the most distinctive aspects of the broad base of slow cinema is its diversity of durational style.
Recent durational films are connected by an interplay of formal devices and observational tendencies that, I
would suggest, cannot adequately be accounted for by referring to the standard categories that subdivide
post-war art and experimental film: narrative/non-narrative, fiction/documentary, realism/anti-illusionism,
and so on. Slow cinema is marked by a tendency to actively problematise these divisions, and therefore,
unlike art cinema (following the templates laid out by Bordwell [1979; 1985] and Neale [1981]) its
representative tendencies cannot strictly be tied to narrativity or fiction. It is a common assumption that the
broad base of slow cinema constitutes a narrative cinema (see, for instance, Bordwell 2005; Bingham 2008;
Flanagan 2008; Romney 2010), but, in this section, I will suggest that the slipperiness of recent durational
cinema resists the categories that are conventionally used to address it.
In Steven Shaviro's critique of contemporary slow cinema (see Shaviro 2010b), the basis of recent
durational and undramatic cinema is considered to be a continuation of 1960s and 1970s modernism, in
particular the formative films of Michelangelo Antonioni, Chantal Akerman, Miklós Jancsó and Andrei
Tarkovsky. Shaviro deems the recent post-modernist slow art cinema of Apichatpong Weerasethakul,
24
Lisandro Alonso, Carlos Reygadas and Bruno Dumont as somewhat derivative in the context of this earlier
modernist work, and accuses contemporary filmmakers of visibly lacking the precision, daringness and
provocation of the high modernist tradition. Shaviro observes that, in a judgement habitually made of the
contemporary 'festival film', there is “an oppressive sense in which the long-take, long-shot, slow-cameramovement,
sparse-dialogue style has become entirely routinised; it’s become a sort of default international
style that signifies 'serious art cinema' without having to display any sort of originality or insight” (Shaviro
2010b). In formulating this familiar critique, Shaviro accepts slow cinema as a continuation of a certain type
of narrative art cinema, which perhaps exacerbates the locus of the critique by assigning a calcification of
style to art cinema's most time-honoured methods of de-dramatisation.
Dan Fox, in a response to the cornerstone of Shaviro's critique, takes issue with this assumption, and
points out that ideas of slowness, contemplation, duration and anti-narrative have been in circulation in a
broader sense in film and video art since the 1960s—hence, in parallel with the European high modernism
that Shaviro laments (Fox 2010). Fox suggests that, as sites of exhibition for specialised aesthetic practices
(not limited to narrativity or conventions of art film), the gallery and the museum have been as significant for
slow cinema as the film festival or arthouse cinema, and that the structural cinema of the 1960s—in
particular Andy Warhol’s Empire (1964) and Michael Snow’s Wavelength (1967), as well as the more recent
film works of artists such as Tacita Dean and Matthew Barney (namely Kodak [2006] and the five
instalments of the Cremaster cycle [1994-2002])—indicate that there is a greater convergence between art
and experimental film than has previously been noted. The implication of Fox's intervention is that once one
begins to widen the frame of how slow cinema has evolved, and indeed what it represents, its aesthetic
norms might not look quite so “routinised,” and instead more substantive in scope and operation.
Most contemporary durational filmmakers do indeed work predominantly in a narrative mode, but
the work of some, such as James Benning, Peter Hutton and Sharon Lockhart, cannot be said to belong to art
cinema in either a narrational or institutional sense. Equally, there are feature-length works by Chantal
Akerman (D'Est [1993], Là-bas [2006]), Pedro Costa (Ne change rien [2009]), Liu Jiayin (Oxhide II [2009]),
Abbas Kiarostami (Five: 5 Long Takes Dedicated to Yasujiro Ozu), Raya Martin (Autohystoria [2007]),
Aleksandr Sokurov (for instance: Spiritual Voices [1995], A Humble Life [1997]) and Jean-Marie Straub and
Danièle Huillet (for instance: Trop tôt, trop tard [1981], Itinéraire de Jean Bricard [2008]) that are too
25
resistant to the norms of art cinema to be considered or analysed on such terms. Certain undramatic narrative
films, such as Apichatpong Weerasethakul's Blissfully Yours (2002), Paz Encina's Hamaca paraguaya (2006)
and Lav Diaz's Melancholia (2008), operate at the outermost limits of the form, reaching a point of
convergence with elements of both video installation work and the avant-garde.
In the field of recent slow cinema, the level of experimentation in retardation of narrativity,
combined with the use of long takes for the prolonged study of peopled activity or material phenomena, has
dissolved boundaries between spheres of production that, historically, have not overlapped so directly. In a
review of James Benning's non-narrative RR (2007), which comprises forty-three static shots of freight trains
traversing the American landscape, Michael Sicinski describes a trend in critical and spectatorial reactions to
contemporary experimental film:
The downside to calling something 'avant-garde film' is that too often it sends viewers running in the
other direction. Even normally adventurous cinephiles, attuned to the distended film-time of Ozu and
Tarkovsky or the Cubist editing of Godard, tremble at the thought of losing conventional anchors
like acting and plot. This is a shame, because if we strip away the labels and look at the films
themselves, the space between international art cinema and non-narrative experimentation is actually
narrowing by the day. No current filmmaker exemplifies this hybridisation better than James
Benning. His work simply confounds any easy categorization. Avant-garde? Documentary?
Landscape study? (Sicinski 2010a).
In this context, what binds Benning's recent work to undramatic art cinema is its rigorous observational
realism and attention to landscape, presented in long, still, patient takes. The same attention to place and
duration is a contingency of contemporary slow cinema, and can be achieved with or without the application
of fiction (that is, in the form of minimalist art cinema narration).
Equally, boundaries between art film, documentary and “non-narrative experimentation” are often
complicated within recent durational films themselves, meaning that when observational strategies used by
filmmakers like Benning are employed in narrative works, the operation of narrative is (necessarily) retarded
to a point at which it ceases to operate as the primary structural element. As Laura Mulvey has suggested in
26
the context of cinema's convergence with new media, the films that comprise slow cinema aim to locate new
ways of responding to the world by seeking to “derive images from whatever the camera observed rather
than a narrative aspiration to order and organisation.” Mulvey describes this approach as pertaining to a
“cinema of record, observation and delay,” in which gaps created by the decline of action are filled by
“empty images of landscapes or cityscapes” and elongated shots that enable “the presence of time to appear
on the screen” (Mulvey 2006: 129). In this type of cinema, “the oppositions between narrative and avantgarde
film, between materialism and illusion, have become less distinct, and the uncertain relation between
movement and stillness, and between halted time and time in duration, is now more generally apparent”
(Mulvey 2006: 30). Such a slackening of divisions between narrativity and the avant-garde (that is, between
action and delay), produces a conspicuous fluidity between divergent modes of representation.
James Quandt, in a peerless extended essay that traces the myriad of influences and stylistic
convergences suggested by Apichatpong Weerasethakul's work, provides an instructive sense of quite how
complex and interconnected the field of slow cinema can be in this respect. Quandt suggests that in order to
productively situate Apichatpong's work within existing traditions, the cinephile must look not so much to
other practitioners of so-called 'Asian minimalism' (in particular, Hou Hsiao-hsien, Tsai Ming-liang and
Hong Sang-soo) and everyday realists (principally, Lisandro Alonso), but to key experimental filmmakers
such as Nathaniel Dorsky, Bruce Baillie, and even Andy Warhol (to whom, alongside Tsai and Béla Tarr,
Apichatpong has admitted a profound influence) (see Quandt 2009a: 27-30, 252-253). Michael Sicinski
suggests that this avant-garde turn in Apichatpong's narrative cinema results from the way in which he and
Dorsky both organise their “images to work against expected, consumable orderings, instead responding to
rhythms, tonalities, intensities of light.” The images in Apichatpong's fiction films are often prone to mystery
and illusion, but retain a sense of material aliveness that verifies a beauty and strangeness that is resistant to
the immediacy of narrative containment (see, for instance, [Fig. 4]). Sicinski observes that, in this manner,
Apichatpong and Dorsky's autonomous images “tend to provide an overall sense of shape or an aesthetic
dominant in their initial viewing, but cannot impart their deepest secrets immediately” (Sicinski 2010b).
Tellingly, Apichatpong has acknowledged such connections directly, stating that his experiments with
narrativity and form represent, to some extent, an attempt “to integrate structural film into a Thai context”
(Römers 2005: 43). Throughout Apichatpong's work, the gap between what has been typified as narrative
27
Fig. 4: Uncle Boonmee Who Can Recall His Past Lives (2010)
film and experimental media threatens to disappear, and Albert Serra has suggested that the undramaticness
of Blissfully Yours might represent “the furthest approach towards abstraction without abandoning the
cinema to become video art, and then being subsequently shifted to museums” (Peranson et al 2007: 23).7
Many contemporary slow films also problematise the conventional opposition between fiction and
documentary. In these films, there is a distinct tendency to reverse what Jean-Luc Godard, according to
Jacques Rancière, called 'cinema's original sin': “to take itself for an adult, to devote itself, under industrial
constraint, to the games of fiction, when it was really there to show, to testify to presence” (Rancière 2002:
118). Having already proposed that slow cinema can be considered to comprise a primarily realist cinema, it
is perhaps worth clarifying this division between fiction and the implication of Godard's notion of 'testifying
to presence'. In terms of realism, the opposition between fiction and documentary has conventionally been
divided into the fields of narrative invention acted out by humans, and the disclosure of historical evidence
of the world. As Bill Nichols clarifies, “realism in fiction relates primarily to sensibility and tone: it is a
7
The multi-platform Primitive project (2009) is indicative of this development, consisting of a series of seven
installation videos and two 'detachable' short video works: Phantoms of Nabua and A Letter to Uncle Boonmee (both
2009). The former has generally been exhibited as a standalone installation in galleries, whilst the latter has been
treated as a short film and was initially exhibited online by MUBI and Animate Projects at
http://www.animateprojects.org/films/by_date/2009/a_letter_to.
28
matter of an aesthetic,” whereas “realism in documentary, marshalled in support of an argument, relates
primarily to an economy of logic” (Nichols 2000: 190). Contemporary slow films tend to retreat from the
rhetorical overtones of argument in pure documentary realism, and instead aim for a hybrid disclosure of
'real' images and sounds that are not overtly “persuasive” in their historical argument (Nichols 2000: 188). In
turn, there is a visible softening of Rancière's distinction that “the documentary, instead of treating the real as
an effect to be produced, treats it as a fact to be understood” (Rancière 2006: 158; my emphasis).
This complication of the division between fiction and documentary is particularly prominent in the
films of Lisandro Alonso, Pedro Costa, Lav Diaz, and Jia Zhang-ke.8
Jia, for instance, has worked in the
modes of fictional art film (Platform [2000], Unknown Pleasures [2002]), documentary realism (Dong
[2006] and Wu yong [2007]), and hybrid structures that create deliberate slippages between the two modes
(24 City [2008], I Wish I Knew [2010]). Costa's cinema follows a similar pattern, encompassing art film
(Ossos [1997]), mediated documentary (Où gît votre sourire enfoui? [2001], Ne change rien [2009]) and the
shifting fictionality of Colossal Youth (already embryonic in In Vanda's Room [2000]). The form of Colossal
Youth is representative of realist hybridity in recent durational cinema: a meticulously observed performance
of the everyday, and the product of a particularly rigorous filming process. Over the course of fifteen months,
Costa lived with his actors—the low-paid labourers, immigrants, and heroin addicts of Lisbon's Fontainhas
neighbourhood—to amass 320 hours of footage, often shooting twenty or thirty versions of the same
sequence (by comparison, Costa shot 140 hours for his first hybrid work, In Vanda's Room, and 150 hours for
Où gît votre sourire enfoui?) (see Peranson 2006).
Throughout Colossal Youth, for instance, fictional episodes are integrated into documentary records
of architecture, everyday conversations and vagrant wanderings, guided by the towering figure of Ventura,
who centres what Thom Andersen calls Costa's “attempt to turn the small form into an epic” (Andersen 2007:
60). In close collaboration, Costa and his 'subjects' (here, an entirely inadequate term) create a slight fictional
framework in which the residents of Fontainhas become character actors (Warholian stars, even) in the
recording of their existence. This uniquely modern narrativisation of observational realism, in which Ventura,
Vanda and their friends 'play' rather than 'be' themselves, also extends to three later fictions that supplant the
everyday with myth: Tarrafal (2007), The Rabbit Hunters (2007) and O Nosso Homen (2010) (Andersen
8
See Lim 2010 for a succinct survey of this tendency in contemporary art and documentary film.
29
2007: 58).
In a different manner, Jia has reflected on the hybridity of fiction and documentary in his work by
disclosing that “a lot of people say that my narrative films are like documentaries and my documentaries are
like narrative films, but I feel that both genres have many possibilities for seeking truth.” Jia notes that he
uses narrative to depict subjects or places that he has “observed for a long time […] to tease out the whole
story,” and employs the documentary form for a process of discovery: that is, “to capture the natural state of
life, of the people” (Rapfogel 2008: 46). Jia's documentaries, like Dong and Wu yong, are structured to render
this process of discovery explicitly, moving between social classes, places (from city to country—or, in Wu
yong, from Guangzhou to Paris to Shanxi) and practices of work with an analytical intent that foregrounds
Rancière's distinction of 'understanding' rather than an ordered production of the real.
After Lisandro Alonso's La libertad (2001), the field of slow cinema has also included another type
of realism that Robert Koehler has referred to as “a cinema free of, or perhaps more precisely in between,
hardened fact and invented fiction.” Koehler calls this realist mode “the cinema of in-between-ness,” and
suggests that it is defined by a propensity to slip from ethnographic document to anthropological fiction and
back again, at times in a manner that is undetectable to the spectator (Koehler 2009: 13). This 'in-between'
tendency is centred on exilic films that depict marginal subjects: the everyday activities of recluses, farmers,
drovers, woodcutters and labourers who work on the surface of the earth for either survival or diminutive
profit. The trend comprises a number of distinct characteristics: an undramatic fictionalisation of everyday
life (Uruphong Raksasad's Agrarian Utopia [2008], C.W. Winter and Anders Edström's The Anchorage
[2009]), small-scale, intimate documentaries of individuals subsisting on the margins of civilisation (Peter
Schreiner's Bellavista [2006], Wang Bing's Man With No Name [2009]), and films perhaps best described as
direct ethnographies: Eugenio Polgovsky’s Los herederos (2008), and Lucien Castaing-Taylor and Ilisa
Barbash's Sweetgrass (2009) [Fig. 5].
All these 'in-between' films possess a quite singular discipline in their attention to place, movement
and the activity of labour, and are resistant to any rhetorical form of argument that persists in the
documentary form. Koehler notes that these films function as records not in terms of their mechanical
objectivity (that is, the activation of camera and sound equipment for the purposes of visual and sonic
30
Fig. 5: Sweetgrass (2009)
research), but in their achievement of a radical mode of realism that is more responsive and concrete than the
“visual historiography” that Nichols prescribes of the documentary form (Nichols 2000: 194). These films
are, perhaps, much closer to Fredric Jameson's definition of an explicitly 'scientific' realism, displaying “an
active, curious, experimental, subversive—in a word, scientific—attitude towards social institutions and the
material world” that encourages and disseminates an attitude that extends beyond mimesis, imitation or
transcription alone (Jameson 2007: 205).
The shape of what I am referring to as slow cinema is, at the time of writing, increasingly defined by
two tendencies: that of the episodic long-shot long-take fiction film, and the depopulated anti-narrative
landscape film. The differentiation between these two tendencies has, helpfully, been mapped by Francisco
Algarin Navarro, in relation to the lineage of two of Jean-Marie Straub and Danièle Huillet's key works:
Dalla nube alla resistenza (1979) and Trop tôt, trop tard (1981) (see Navarro 2011). The first strain,
originating with Dalla nube alla resistenza, is represented by the outer limits of art cinema, and incorporates
both textual adaptations and 'games of fiction': Manoel de Oliviera's Inquietude (1998), João César
Monteiro's Vai-e-vem (2003), Costa's Colossal Youth and Serra's El cant dels ocells. Following Trop tôt, trop
tard, the second strain points to non-narrative landscape (and cityscape) films: Chantal Akerman's D'Est
31
Fig. 6: Autohystoria (2007)
(1993), Abbas Kiarostami's Five: 5 Long Takes Dedicated to Yasujiro Ozu, John Gianvito's Profit motive and
the whispering wind (2007), Raya Martin's Autohystoria (2007) and the work of James Benning.9
The stylistic bases of these films do not differ wildly, and most favour static set-ups that fix the
frame as if it were a window or a still life (a strategy that is purely observational in landscape film, yet
resolutely classical and elegant in Oliveira's cinema). Akerman, Martin and Straub-Huillet also use mobile
long takes that are repeatedly decelerated and drawn-out: in Akerman's D'Est, lateral tracking shots glide
through wide streets past lines of patient bodies, whilst Martin, in Autohystoria, executes handheld DV shots
that obliterate their putative meaning with vast swathes of duration, such as in an opening shot that lasts
thirty-seven minutes whilst tracking a man's long walk to an undisclosed destination. In this particular shot,
the (unnamed) character's walk across a number of streets takes twenty-six minutes (during which Martin's
shadow, holding the DV camera, is visible throughout) before the camera is fixed, for another ten minutes,
on the exterior of a house that the man enters after reaching his destination [Fig. 6]. Autohystoria consists of
only fourteen shots in total, of which three are brief excerpts of archival footage of Emilio Aguinaldo's navy
9
In light of this continuum, the eleventh edition of BAFICI (2009) screened Trop tôt, trop tard in a special
programme alongside Benning's One Way Boogie Woogie / 27 Years Later (2005), Casting a glance (2007) and RR
(2007).
32
lasting only a few seconds each (included at the end to echo the film's preceding recreation of Aguinaldo's
execution of Andres and Procopio Bonifacio in 1897, an act that was ordered to quell revolution).
Autohystoria's radical narrative structure, which very sparingly re-enacts the execution of the Bonifacio
brothers, cannot be situated in any existing cinematographic tradition other than the aesthetic dichotomy
suggested by Navarro above. In this context, it is perhaps not coincidental that two extended shots of a bus
endlessly circling an urban roundabout in Martin's film directly recall the opening shot of traffic circling the
Place de la Bastille in its suggested antecedent, Straub-Huillet's Trop tôt, trop tard.
In both strains of narrative art and non-narrative experimental film detailed by Navarro, shots are
consistently structured to accentuate their compositional quality (staging, perspective, light, colour, tonality),
the actuality of their material detail, and the passage of time. Any conventional operation of narrativity
becomes a secondary concern, and content is frequently structured episodically, without recourse to causal
relations. For instance, Straub-Huillet's textual adaptations adopt composite forms (Dalla nube alla
resistenza systematically presents extracts from two texts by Cesare Pavese: six dialogues from Dialogues
with Leuco [1947] and passages from The Moon and the Bonfire [1950]) in a manner that is mirrored by the
fiction films that Navarro mentions: Monteiro's comic, elegiac Vai-e-vem, for instance, is structured solely by
a series of disconnected episodes of gently absurd incidents and erotic conversations, casually connected by
Monteiro's travels through Lisbon on public transport.
In these durational works, narratives are therefore constructed in a similar way to the sequencing of
shots in landscape films—that is, by the selection of what Andrei Tarkovsky described as 'sequential facts'
taken from “'lumps of time' of any width or length,” rather than a cohesive articulation of the chain that holds
them together (Tarkovsky 1987: 65). As these lumps of time are whittled free of causal functionality,
narrative increasingly appears not as a lack (in relation to dominant commercial cinema), but as an addition.
Albert Serra, who selected the extended takes of El cant dels ocells from 110 hours of footage, notes that the
underlying organisation of the final film only reluctantly accommodates the application of narrativity: “I
wanted to make a simple film, like paintings from the Middle Ages. It’s not narrative. It’s one image... stop...
another image. It’s like if you were in a church and saw Middle Age paintings side by side. Very simple”
(Hughes 2009). Here, images are reserved primarily for their aesthetic function, and positioned in the film as
if curated by motif and time.
33
The wind in the trees
The field of slow cinema, above all, attempts to achieve a resonance of time, body or place beyond the
dominant conventions of classical or intensified continuity style (a distinction that I will expand on in
Chapter Three), and dramatic narrativity. This intent is illustrated well by Jonathan Rosenbaum's account of
watching (and listening to) the second section of Straub-Huillet's Trop tôt, trop tard, which was filmed in
mostly rural locations in Egypt, and is accompanied by a non-diegetic reading of Mahmoud Hussein's text
Class Conflict in Egypt, 1945-1970 (1973):
No other film has come even close to making me feel I've been to Egypt, which this film does. A lot
of this has to do with the tempo, rhythm, pacing: the sight and sound of a donkey pulling a cart down
a road towards the camera is recorded in long shots and at leisure, with no sense of either ellipsis or
dramatic underlining according to any principle other than the placement of camera and microphone
in relation to the event. The extraordinary result of this technique is that one almost feels able to taste
these places, to contemplate them—to observe and think about them (Rosenbaum 1983: 194-195).
During many of the bucolic scenes in Trop tôt, trop tard, the only active elements of the shot are the
occasional movement of vehicles, a few people, animals, birds, and the murmuring of the wind [Fig. 7]. This
limited activity resists burdening the spectator with diegetic elements that distract from his or her
contemplation of material detail, a visual strategy that led Serge Daney to suggest that “if there is an actor in
Trop tôt, trop tard, it is the landscape.” For Daney, this actor duly “performs with a certain amount of talent:
the cloud that passes, a breaking loose of birds, a bouquet of trees bent by the wind, a break in the clouds.”
Straub-Huillet's generous attention to the spontaneous activity of nature, unadorned by some form of human
narrativity, makes way for a performance that is strictly “meteorological” (Daney 1982).
Such close attention to place enables the visual articulation of what Siegfried Kracauer refers to as
“things normally unseen” in everyday phenomena, such as “the shadow of a cloud passing across the plain, a
leaf which yields to the wind” (Kracauer 1997: 52). By focussing attention on minor material details, this
mode of observation often recalls D.W. Griffith's statement of 1947, oft-repeated by Jean-Marie Straub:
34
Fig. 7: Trop tôt, trop tard (1981)
“What the modern movie lacks is beauty—the beauty of moving wind in the trees, the little movement in a
beautiful blowing on the blossoms in the trees. That they have forgotten entirely. In my arrogant belief, we
have lost beauty” (Goodman 1961: 11). Straub-Huillet's style itself is, in many ways, devoted to answering
Griffith's plea in the creation of images that frame the material continuum of the visible world: the earth, the
sky, the weather, and the gestures and voices of the people. Tellingly, in an anecdote that echoes Griffith's
statement, Barton Byg notes that Straub once compared “the motion of the musicians' fingers” on piano keys
in Chronik der Anna Magdalena Bach (1967) to the audience's delight at seeing, for the first time, the wind
in the trees in the first films of the Lumière brothers (Byg 1995: 22).10
In Chronik der Anna Magdalena Bach, the motion of Gustav Leonhardt's fingers also signifies the
10
The reaction referred to by Straub here is recalled by Dai Vaughan: “Georges Méliès, a guest at the first Paris
performance [...] made particular mention of the rustling of the leaves in the background of Le Déjeuner de bébé
[1895]—a detail which, as [Georges] Sadoul himself observes, would scarcely be remarked today” (Vaughan 1999:
4-5).
35
film's recording of direct sound—a strategy that, for Straub-Huillet, is as important as the visual
documentation of material detail. With the aim of countering, in Straub's words, “the violent habits of current
recording techniques,” Straub-Huillet located musicians who could play the original baroque instruments of
Bach's era, and filmed live performances solely in uninterrupted takes (Byg 1995: 55). This material
authenticity, visibly demonstrated by the motion of Leonhardt's fingers, also extends to another implication
of Straub's comment on the wind in the trees: that modern cinema signals a return to the latent potential
(rather than the 'essence') of early film (embodied by the work of the Lumière brothers), and to an
observational potential that might have been lost during decades of industrial classicism.
In Kracauer's realist theory, this visual potential stems from cinema's analogue extension of
photography, which gifts film its “outspoken affinity for unstaged reality” (Kracauer 1997: 18). In its earliest
stages, cinema's ability to capture a representation of material phenomena in time produced responses from
audiences that image-sound situations in dominant cinema rarely encourage. Dai Vaughan recalls that what
most impressed the first audiences of the Lumière brothers' films “were what would now be considered the
incidentals of scenes: smoke from a forge, steam from a locomotive, brick dust from a demolished wall”
(Vaughan 1999: 4). These revelations of transient details are often foregrounded in observational cinema,
reminding the spectator that the medium has always been, in Kracauer's definition, “animated by the
chimerical desire to establish the continuum of physical existence” (Kracauer 1997: 63).
This chimerical desire manifests itself in one of cinema's singular ongoing projects: to film the
source of the visible movement of leaves in the trees, and that of smoke from a forge. Daney notes that Trop
tôt, trop tard represents “one of the few movies” that has managed to film this often invisible, everyday
phenomena, at least since Victor Sjöström’s tumultuous melodrama The Wind (1928) (Daney 1982). StraubHuillet's
film enters an informal lineage that spans, for instance, Joris Ivens' Pour le Mistral (1965) and Une
histoire de vent (1988), a number of Chris Welsby's experimental films (in particular, Wind Vane [1972],
Seven Days and Windmill III [both 1974]), John Gianvito's Profit motive and the whispering wind, and the
bucolic images of trees, fields, swaying leaves and flowers in Patrick Keiller's Robinson in Ruins (2010). In a
comment reflective of the continuum between these non-narrative films, Rosenbaum has described Profit
motive and the whispering wind as “a mystical communication with deserted landscapes”—a description that
might equally apply to the purpose of the extended takes in each film (Rosenbaum 2008: 30). Other
36
Fig. 8: Profit motive and the whispering wind (2007)
observational works in slow cinema have also turned their attention to the wind at great length: in, for
instance, Irimiás and Petrina’s determined, forceful walk down the centre of a windswept road in Tarr's
Sátántangó (1994); the exterior tracking shots during an apocalyptic gale in Tarr's A torinói ló (2011); the
gentle movement of cumulus clouds and tidal currents in every shot of Benning's Ten Skies (2004) and 13
Lakes (2004); and the delayed wind shear that hits the trees in the third shot of Benning's Ruhr (2009).11
Gianvito's Profit motive and the whispering wind is a quite unique instance of this trend, and points
to a particular spectatorial practice that is encouraged by key works of contemporary durational film. In
Gianvito's film, images of the wind are not as temporally distended as Tarr's or Benning's (lasting mere
seconds rather than minutes), but their inventory of weathered gravestones, long grasses and buffeted
11
P. Adams Sitney's essay on cinematographic representations of landscape, Landscape in the Cinema: the Rhythms of
the World and the Camera, includes an excellent summary of the film-historical basis of filmmakers' attention to the
wind in experimental and documentary cinema; see Sitney 1993: 119-125.
37
branches perhaps visualise, in an overwhelmingly literal form, what Christian Keathley calls the
“fetishisation of marginal, otherwise ordinary details in the motion picture image” [Fig. 8]. This tendency to
fetishise marginal detail is, as Keathley suggests, “as old as the cinema itself,” and, as in Gianvito's film,
continues to form the basis of one of its most elemental special effects (Keathley 2006: 8).
Film studies has recently embraced this revelatory capacity signified by the wind in the trees in the
concept of the 'cinephiliac moment' (see, in particular, Willemen and King 1994; Keathley 2006). The turn
toward cinephilia in film studies over the last couple of decades has emerged in tandem with the field of slow
cinema under discussion here, and has more recently been symptomatic of the expansion of new digital
communities (across blogs, journals and social media) and film festival networks (in which the exhibition of,
and discourse around, contemporary slow cinema has thrived).12
The concept of the 'cinephiliac moment'
might appear, at first, to reflect contemporary observational cinema's privileging of what Paul Willemen
describes as “the dimension of revelation that is obtained by pointing your camera at something that hasn’t
been staged for the camera,” but it is important to note that it derives, primarily, from a classical rather than
modern schema (Willemen and King 1994: 243).
In the context of this schema, Noel King describes the activity of cinephiliac perception as “the
isolating of a crystallisingly expressive detail” that occurs in excess to the image's primary codification: that
is, a privileged, subjective moment surplus to the order of classical narration (Willemen and King 1994:
227). Keathley states that a cinephiliac moment indicates a detail that is not choreographed for the spectator
to see, “or rather, if it is, it is not choreographed for the viewer to dwell on excessively,” and the five
moments upon which he draws for cinephiliac anecdotes in his valuable book on the subject are notable for
their committed attempts to recover sensuous experience within a highly ritualised framework (see Keathley
2006: 30, 153-177).
For Willemen, the cinephiliac moment's revelation (or excess) is only observable in a dimension that
has been programmed for other means: “if that in itself is the system of the film, as, say, in a Stan Brakhage
film, you don’t have a cinephiliac moment precisely because it’s no longer demarcatable. You can't perform
the critical trawling operation because the whole film tries to be 'it'” (Willemen and King 1994: 238).
12
Alongside Willemen and King 1994, and Keathley 2006, key texts on contemporary cinephilia include Jonathan
Rosenbaum and Adrian Martin's exemplary Movie Mutations (2003), and the extensive dossier on cinephilia
(comprising thirteen essays) in Framework 50 (2009).
38
Although Willemen cites Brakhage's experiments in our limits of perception to this end, it is quite clear that
much contemporary slow cinema, particularly the landscape film, similarly strives to directly present
opportunities for a reanimation of 'it': that is, a liberation of “the repressed materiality of the film image”
from a strictly classical or codified order (Keathley 2006: 53). The spectatorial practice invited by the films
of Straub-Huillet, Gianvito, Benning and Tarr frequently demarcates the whole image as a field in which
revelatory detail abounds, and as a site of 'excessive' material density.
Roger Cardinal, drawing on Roland Barthes's notion of the punctum—a particularly striking detail
that is added to a photograph by a “thinking eye” after being triggered by something that “is nonetheless
already there”—suggests that two divergent strategies of viewing arise from a spectator's attention to the
material registration of detail: a 'literate' mode, which is drawn to the “obvious Gestalt or figure on offer” in
a carefully mediated image, or a 'non-literate' mode that “instead roams over the frame, sensitive to its
textures and surfaces—to its ground” (Barthes 2000: 45, 55; Cardinal 1986: 124). Cardinal also speculates
that the latter (non-literate) mode may be associated with “habits of looking which are akin to habits of
touching,” as the eye roams to collect fortuitous details and “empathetic impressions of touch sensations”
(Cardinal 1986: 124).
This latter mode of spectatorship, of noticing details that “escape directorial sponsorship to take
[their] place before the viewer quite autonomously,” is both encouraged and actively permitted by
contemporary durational filmmakers (Cardinal 1986: 122). In Chapter Four, I will describe at length two
exemplary films that embody, and encourage, such observational attention (Abbas Kiarostami's Five: 5 Long
Takes Dedicated to Yasujiro Ozu [2003] and Sharon Lockhart's Double Tide [2009]), but the tendency is
already illustrated well by Cardinal's account of catching sight of “true glimpses of the natural” in Andrei
Tarkovsky's Mirror (1974): for instance, in “a lingering shot of undergrowth stirred by gusts of wind which
seemed to affect only a small sector of the bushes at a time, brusquely animating and then immobilising
groups of leaves and twigs.” Cardinal notes that Tarkovsky's long takes invite the spectator “to stare
reverently” at “stretches of unkempt woodland (ferns, fallen branches, weeds, stumps)” and other inchoate,
tangled “samplings” of the world by expanding the temporal dimension of the shot (Cardinal 1986: 124-
125). Such displays of material phenomena seem to draw attention to themselves in place of narrative or
human action, and occupy a diegetic function that is “neither informational nor symbolic” but open to a
39
contemplation of things that are simply, “obtusely,” themselves (Cardinal 1986: 122).
The field of slow cinema not only seeks to record material phenomena in this way (spurred by the
motion of the wind in the trees), but invites the spectator to contemplate its activity (or inactivity) for
uncommonly extended periods. Observational films therefore tend to display a preference for images that
contain very little or localised movement (without the distraction of narrativity) in order to free the
spectator's gaze to scan the image's registration of the world over time. In relation to the practice and
experience of watching Trop tôt, trop tard, Rosenbaum comments that
some spectators find this activity tedious; many of the first spectators of Jacques Tati's Playtime
[1967] complained about it in a comparable manner, claiming that 'nothing happens'. Yet the
significant relationships between Straub-Huillet's long shots and Tati's is that something is always
taking place in them, if only the spectator can learn to watch and listen without expecting to be led
by the nose through the sequence (Rosenbaum 1983: 195).
This practice suggests a return, in many respects, to the potential of early cinema implied by Straub and
Vaughan's comments, and to Kracauer's realist designation of the medium's role as an extension to
photography. David Campany, in his book Photography and Cinema (2008), has noted that the “opportunity
to contemplate and interrogate while looking” remains a sensuous experience central to the depictive arts,
regardless of media, as it arises from “the traditional sense of 'presence' typical of art’s materially fixed
media such as painting, sculpture and photography, all of which have valued the depiction rather than recreation
of movement.” In Campany's view, “art’s preference for the slow is motivated by more than the
desire to separate itself from mainstream cinema and spectacle at large,” which would concur with the
sustained macro function of film's extension to photography signalled by both Kracauer and André Bazin's
realist theory (to which I shall return in my next chapter) (Campany 2008: 39). Contemporary films can
certainly be seen in light of this tradition, but I would suggest that there are perhaps more linear filmhistorical
forces at work in the evolution of durational style in post-war cinema. One of these forces arises
from an established current in North American experimental cinema that privileges both the primacy of the
filmic image and a uniquely attentive or contemplative mode of spectatorship. This current extends to
40
contemporary cinema, as I will suggest in the next section.
Experimental cinema and the production of duration
Although post-war art cinema often deviates substantially from classical norms, it is not a paradigm that
rejects narrative in order to accentuate, solely, the capacity of the image (that is, for a real that might be
represented, or an illusion that can be destroyed). Art cinema narration, in David Bordwell's canonical
formulation, always seeks to “pose questions that guide us in fitting material into an ongoing structure,” even
if that structure is loosened of causality or resolutely ambiguous (Bordwell 1985: 210). Bordwell, like Steve
Neale, notes that art cinema has, in part, become “a coherent mode” by defining itself as a stark deviation
from classical narrative since the late 1950s, borrowing concepts from modernism, literature and theatre to
interrupt the authority of objective narrational agency (Bordwell 1985: 228). By contrast, non-narrative
experimental film operates entirely as an externality to classical film, not so much defining itself against the
institutions of commercial cinema as simply taking place outside of them. In the preface to the first edition of
his landmark study Visionary Film (1974), P. Adams Sitney notes that the relationship of avant-garde cinema
to commercial film is one of “radical otherness,” where both modes “operate in different realms with next to
no significant influence on each other” (Sitney 2002: xii). Little has changed in that respect, and the category
of experimental film—often used interchangeably with 'avant-garde' (Sitney's preference) as well as, when
appropriate, 'underground' film—is a fertile site for the exploration of the potential of the image that is not
permitted within the dominant practice of commercial cinema.
Nicole Brenez, in Fergus Daly's documentary Experimental Conversations (2006), suggests that
experimental cinema is, in essence, “the field of investigation of the very modalities of our apprehension, and
particular modes of vision.” The starting point for any experimental filmmaker is not the task of creating
images for integration into a narrational system, but an elemental question: “What use is it to make an image,
or not make one?” Sitney argues that the label 'experimental' detrimentally “implies a tentative and
secondary relationship to a more stable cinema” and adopts 'avant-garde' for the purposes of his study in
Visionary Film, but I prefer Brenez's fuller distinction, and will retain the term throughout this thesis for its
openness and adaptability (Sitney 2002: xii). In Brenez's conception, the image (and its sonic capacity as an
41
extension to the image) is never less than the primary field of research, leading the experimental film to
neglect the secondary function of narrativity. Narrative in experimental film is simply one aspect of what
might be filmed, and a narrative aspect is the concern of dominant cinema. Experimental cinema thus tends to
alight upon singular events (in isolation, series or superimposition) rather than narratives, and its explicit
function is to interrogate both the filmic apparatus and the spectator's perception of those events. This
function is what connects disparate structural or realist works such as Andy Warhol's silent films of 1963 and
1964, Michael Snow's Wavelength (a dual presentation of something that happens in the film and to the film),
the assemblage of experience and memory in Jonas Mekas's diary films, such as Diaries, Notes and Sketches
(also known as Walden) (1969), and the more recent experimental cinema of James Benning, Peter Hutton
and Sharon Lockhart (amongst others).
Post-war art cinema, from which certain aspects of contemporary slow cinema derive, has habitually
drawn attention to the aesthetic qualities of the filmic image, but never to the point of the total elimination of
narrativity. However, Noël Burch observes that Michelangelo Antonioni’s Cronaca di un amore (1950), for
instance, is an early art film that comprises early “proof of a firm bias (in the nobler sense of that word) in
favour of stripping the image of any sort of narrative function. From the point of view of 'storytelling',
practically nothing ever occurs on the screen” (Burch 1973: 75). Cronaca, a melodrama centring on two
violent deaths (and the waylaid plotting of the second), instead sets up a dialectical relationship between
sound and image in which “the complete freedom of the camera is restored” and a certain degree of
autonomy is returned to the image (Burch 1973: 76). Nathaniel Dorsky, when also writing on Antonioni's
film in his important monograph Devotional Cinema (2005), observes that this intermediate release from
narrational agency creates an impression in which “every shot is the narrative and the narrative is every shot”
(Dorsky 2005: 41). For Dorsky, Antonioni's early work is able, in this way, to move beyond treating film as a
“medium for information” in which “the whole absorbing mechanism of projected light—the shots, the cuts,
the actors—is there only to represent a scripted idea.” This is a point at which the medium can, for Dorsky,
embody what he calls 'devotional cinema': the transition from “mere representation” to a “direct experience”
beyond narrative (or intellectual) content (Dorsky 2005: 44). Dorsky's own filmic work, rejecting narrative
for the (entirely silent) presentation of moments of material revelation and aliveness, makes this leap, and I
would argue that an enactment of this transition is a key factor in the formal and narrational operations of
42
Fig. 9: April (2012)
slow cinema (see, for instance, [Fig.9]). Equally, James Benning, when speaking in Reinhard Wulf's
documentary James Benning: Circling the Image (2003), is also clear about foregrounding the total primacy
of the image (at the behest of narrativity) in his work: “There's an essence to an image that isn't even about
narrative, it's about what it is, outside of narrative terms. The frame itself is what tells my story.”
Scott MacDonald has noted that, historically, the attempt to liberate the spatial and temporal
dimensions of the image from narrative has gone hand in hand with the development of 'critical cinema': that
is, the type of film designed to provoke a 'film-critical response' in opposition to the narrative relations of
dominant cinema (MacDonald 1993: 3). Although rooted in the avant-garde practice of the 1920s, the field
of 'critical cinema' developed quite rapidly after the Second World War (as technological and cultural
developments in film, literature, abstract painting and sonic arts took hold in North America, Europe and
Japan), and by the mid-1960s North American non-narrative film had begun to introduce duration (as a
component of the moving image) far more aggressively, and in considerably greater surplus, than the art
43
cinema of the equivalent period. A new cinema of revelation, driven by the availability of cheaper 16mm
camera and projection equipment (as well as new film society networks and filmmaking cooperatives), gave
rise to a representational mode whose influence is integral to contemporary durational style: the structural
film. This cinema adopted as its subject the direct, demystified presentation of light (Peter Kubelka's Arnulf
Rainer [1960] or Tony Conrad's The Flicker [1966]), colour (Paul Sharits' Ray Gun Virus [1966]), and time
(Andy Warhol's silent films [1964-1966] and Hollis Frampton's (nostalgia) [1971]), in turn manifesting
Nathaniel Dorsky's later reminder that “the field of light on the wall must be as alive as sculpture” (Dorsky
2005: 44).
The structural film is defined by Sitney as, simply, “a cinema of structure in which the shape of the
whole film is predetermined and simplified, [so that] it is that shape which is the primal impression of the
film,” and any visible (or audible) content “is minimal and subsidiary to the outline” (Sitney 2002: 348). A
primary characteristic of structural film is the fixed position of the camera and frame, alongside a variable set
of elements that make the material substrate of film unavoidable to the spectator (such as the rapid
alternation of black and white frames for a flicker effect or loop printing for repetition). The “shape” of the
film, in the work of Andy Warhol or Michael Snow throughout the 1960s, became its duration: both the
duration of the object (the content of the image) and the actual time of projection. Structural films are
designed to offer few distractions, or deviations, from those two actualities.
Although this turn toward reduction and duration in experimental film should not be classified as
strictly realist (unlike aspects of post-war art cinema), I would like to focus here on its development of an
integral aspect of more recent slow cinema: the observational capacity of the fixed frame and extended
duration.13
MacDonald notes that the re-discovery of early films by Auguste and Louis Lumière in North
America in the 1960s inspired a number of filmmakers to challenge the ideology (and “system of
identification”) of dominant cinema by adopting a deliberately uncomplicated observational practice (Gidal
1978: 3). Non-narrative filmmakers began to develop what Peter Hutton has referred to as “an extremely
reductive strategy,” firstly by Andy Warhol in his Kiss series (1964), Screen Tests (1964-1966) and durational
works (such as Henry Geldzahler and Empire [both 1964]), then in observational diary or structural land-
13
Structural film also accommodates the fiercely “non-illusionist” practice outlined in Peter Gidal's definition of
Structural/Materialist film, which sets out to minimise any representational content “in its overpowering,
imagistically seductive sense, in an attempt to get through this miasmic area of 'experience' and proceed with film as
film” (Gidal 1978: 2).
44
Fig. 10: Fog Line (1970)
and cityscape films: Hutton's own first films, Larry Gottheim's Fog Line (1970), and James Benning's early
work (such as 11 x 14 [1977] and One Way Boogie Woogie [1977]—a work that Benning re-filmed in 2004
for One Way Boogie Woogie: 27 Years Later, and supplemented with eighteen new shots, filmed in May
2011, for its most recent version: One Way Boogie Woogie 2012) (MacDonald 1998: 247). As a pointer to
this tendency, which originates in yet extends beyond the parameters of structural film, Jonas Mekas'
Diaries, Notes and Sketches (also known as Walden) opens with a title card that boldly declares it is
“dedicated to LUMIÈRE.”
The critical innovation of many of these films, in the context of contemporary observational cinema,
is an application of extended duration to complement the 'window' of the fixed frame. As in the first films of
the Lumière brothers, the visual image is given over to a direct and unmediated representation of reality:
simple or everyday human activities, displays of natural phenomena, and the “incidental” elements of scenes
45
Fig. 11: Sátántangó (1994)
described by Dai Vaughan in my previous section (see Vaughan 1999: 1-8). However, whereas the Lumière
brothers' films would run for just under a minute (occupying fifty-five feet of film cranked by hand at 16
fps), structural filmmakers employ shots that can run for up to eleven or longer than thirty minutes—as in
Warhol's Empire, which was shot with a 16mm Auricon camera that could load 1200ft magazines of Tri-X
negative stock (also pushed to ASA 1000) (see Burch 1990: 15; Angell 1994: 15-18). Larry Gottheim's Fog
Line, for instance, occupies the full length of a single 400ft 16mm reel, and depicts a blanket of fog clearing
above a meadow from a resolutely still camera position [Fig. 10]. As the film begins, only telegraph lines and
the tips of two trees are visible amidst the grey fog, but over the course of eleven (uninterrupted) minutes the
bodies of the trees gradually emerge from the damp air, and colour (a soft green) begins to seep into the
frame.14
The compositional basis of this fixed display of meteorological phenomena is, in slow cinema,
revisited in both narrative and non-narrative works. In Béla Tarr's Sátántangó (1994), for instance, two static
shots of mist enveloping and clearing before a disused building (first in a depopulated frame, then silently
watched by three characters: Irimiás, Petrina and Sanyi) play out according to the principles of observation
14
For an exemplary analysis of Fog Line, see MacDonald 2001: 7-13.
46
set by Gottheim [Fig. 11]. Similarly, in one of the twelve sections of Pine Flat (2006), Sharon Lockhart
stages a sequence shot in which mist rolls into a ten-minute shot, engulfing the restless child sitter before
dispersing as unexpectedly and mysteriously as it arrived. These extended shots, as MacDonald notes of the
films of the Lumière brothers, exemplify a tradition in which
the goal [...] is much the same: to focus attention—an almost meditative level of attention—on
subject matter normally ignored or marginalised by mass-entertainment film, and, by doing so, to
reinvigorate our reverence for the visual world around us and develop our patience for experiencing
it fully (MacDonald 1993: 11-12).
Peter Hutton has described the first films of the Lumière brothers as “a revelation when you see them
in this day and age, because there's a certain kind of innocence to how they were structured. They weren't
designed either to be narrative or documentary” (MacDonald 1998: 246). Hutton's first films, July '71 in San
Francisco, Living at Beach Street, Working at Canyon Cinema, Swimming in the Valley of the Moon (1971)
(a silent diary of commune life shot with a portable 16mm Bolex) and New York Near Sleep (for Saskia)
(1972) directly evoke, in their own spontaneous gathering of images (of places, travels, animals, and
dispersions of light across the surface of interiors and domestic objects), the very subjects of both primary
and “incidental” fascination for audiences of early cinema.15
Hutton has suggested, in conversation with Robert Gardner in Screening Room (with Peter Hutton)
(1977), that the broader adoption of observational practice during the “new period of self-exploration” in the
1960s signified a nascent attempt “not really to try to manipulate things, but just to sort of preserve them.”
After his first city portrait, New York Portrait: Chapter I (1979) (followed by two more of New York [in
1980-1981 and 1990], Budapest [1984-1986], Łódź [1991-1993], and two presently unfinished portraits of
West Berlin [shot in 1980] and Detroit [for 3 Landscapes, a work still in progress at the time of writing]),
Hutton's own films have acquired a rigour in their application of duration that is characteristic of the
15
In Reinhard Wulf's documentary Circling the Image (2003), Benning echoes this by claiming that his intention, in
his 16mm work, has been “to shoot as the pioneers at the dawn of cinema history. In order to film the arrival of a
train or a kiss they simply put up the camera and shot an entire reel. [...] I wanted to go back to the beginning
because I thought filmmaking grew up too quickly. Narratives got introduced and replaced the essence of the image.
So I wanted to go back and do these one take rolls.”
47
Fig. 12: At Sea (2007)
evolution of post-war slow cinema. Hutton appears to choose visual subjects (landscapes, built
environments, or industrial processes) primarily for their revelatory capacity, and, over the course of the last
three decades, has compiled a series of studies of the Hudson River (shot in all seasons and conditions,
between Study of a River [1995-1996] and Two Rivers [2002], the latter of which is paired with footage of
the Yangtze River and very rarely screened); contemplative 'luminist' images of skies, cloud formations and
the moon (accompanied by, in a rare instance of human presence, light scattering across a sleeping child's
face) in Landscape (for Manon) (1987) and In Titan's Goblet (1991); extended takes of sublime Icelandic
landscapes (frozen seas, glaciers, mountain peaks and cavernous valleys) in Skagafjördur (2002-2004); and a
three-part documentary study of the construction, voyage, and dismantling (by scrappers) of container ships
in At Sea (2007) [Fig. 12].
Andy Warhol's structural films represent a hyperbolic extension of this observational strategy, and
48
Fig. 13: Empire (1964)
are, I would suggest, central to the evolution of the field of slow cinema. Some contemporary filmmakers
point to this lineage by directly referencing Warhol's work, such as Pedro Costa (in the compositional
homage to Beauty #2 [1965] in In Vanda's Room [2000]) and James Benning (in the literal Twenty Cigarettes
[2011], which revisits the formal strategy and visual content of both Warhol's Screen Tests and Gregory
Battcock's action at the end of Blow Job [1964]). As P. Adams Sitney notes, Warhol's significance to postwar
durational style can be measured by the rejection of classicism's “myth of compression” in his art, and
his decision to make “a profligacy of footage the central fact of all of his early films” (Sitney 2002: 349).
The 'unmotivated' production of duration in Warhol's silent films of 1963 and 1964 is, as Sitney suggests,
designed to “outlast [the] viewer’s initial state of perception,” so that “by sheer dint of waiting, the persistent
viewer would alter his experience before the sameness of the cinematic image” (Sitney 2002: 351).
I shall briefly reflect on the form and critical assessment of Empire (1964) here, as both the work and
49
its reputation continue to occupy a somewhat monumental position in the context of durational film. Warhol's
film, famously, consists of a single stationary image recorded by a tripod-mounted 16mm Auricon camera
from the forty-fourth floor of the Time-Life building (on Sixth Avenue in New York) on 25th
July 1964 [Fig.
13]. The recording begins at just after eight o'clock in the evening as the sun sets, and ends in total darkness
at about half past two in the morning. The full version of the film, when correctly projected at 16 rather than
18 or 24 fps, runs for eight hours and five minutes, and should only be interrupted by necessary reel changes
that occur after every forty-eight minute period. Empire's visible content is quite minimal and easily
summarised: during the first reel, the exterior floodlights of the building are switched on after the sun fades,
and the resultant image remains constant until the floodlights are switched off in the second-to-last reel, at
which point the outline of the building effectively vanishes.16
Warhol's insistence on decelerating the speed of
projection (by projecting footage at 16 fps) to just over eight hours distends the perceptible duration of this
visible activity and the habitual act of spectatorship—a strategy that, as Peter Gidal recalls in his recent
monograph on Blow Job, has been aptly referred to as 'drugtime' or 'dragtime' (Gidal 2008: 25).
As Callie Angell has noted, Warhol once stated that the point of Empire was “to see time go by,” a
stipulation that is quite literally achieved by a visible clock provided by the light on top of the Metropolitan
Life Insurance Tower—the tall building to the left of the Empire State Building. This light flashes once every
fifteen minutes, and signals every passing hour by blinking the correct number of times (9, 10, 11, 12, 1 and
2). Warhol's strategy of providing a temporal marker within the shot serves to draw attention to the image's
own inexorable production of duration, and Angell has observed that “it is therefore possible to follow the
exact time of shooting throughout the entire film and, quite literally, 'see time go by', [albeit] at a rate that is
one-third slower than real time” (Angell 1994: 16).
Critical discourse around Empire has conventionally focussed on the issue of Warhol's passivity of
form and production (that is, his total indifference to classical narrativity or aesthetics, which is epitomised
by his occasional habit of simply activating the camera and then walking away), and the film's anti-illusionist
exhaustion of representational content (see, primarily, Koch 1991; Shaviro 1993; Margulies 1996). I would
argue, however, that these related approaches tend to sideline a key aspect of Warhol's work that is readily
apparent in more recent cinema: a sustained, contemplative attention to the registration of light and minor
16
For a detailed itinerary of the visible changes in light during each reel, see Angell 1994: 18.
50
detail over time. This tendency is expressed by Apichatpong Weerasethakul, who (without, admittedly,
having seen Empire in full) suggests that “with the superfluity of images today, [Empire] reminds us of the
simple beauty of the film material itself—the patience, the act of watching film, lights being turned on in the
building. […] It's like the camera was seeing for itself” (Quandt 2009: 252).
More often than not, this attention to the “simple beauty” of Warhol's film has been sidelined in
canonical studies. For instance, Stephen Koch contends that the “absolute vacuity” of technique in Empire
comprises a dismissal of “any possible source of visual interest for the audience,” and that the work
constitutes “a massive, absurd act of attention, attention that nobody could possibly want to give or sit
through. Indeed, nothing could possibly tolerate it—and here's the point—but a machine, something that sees
but cannot possibly care” (Koch 1991: 60, 61).17
Steven Shaviro echoes this approach in his analysis in The
Cinematic Body (1993), suggesting that Warhol was only ever concerned with cinema's “most blatant and
superficial effects” that cannot “represent the real [because] what you see is what you get, since the reality of
the image is precisely equal to the reality of the object of which it is an image” (Shaviro 1993: 204, 214).
Hence there is “no latent content at all” in Warhol's art, and the material world is transposed to a form that is
at once banal and sublime, and appears to offer “less a mirror than a black hole” (Shaviro 1993: 207).
Alternative canonical perspectives on the function of reduction and duration in Empire are, however,
evident in Callie Angell's meticulous study of Warhol's body of (cinematographic) work, and in the notable
support for Warhol's films from allies such as Jonas Mekas and Phil Solomon. In recent years, further
attempts have been made to reconsider the purpose of Warhol's durational practice. For instance, Andréa
Picard has argued that Empire must be projected and viewed in its entirety “not so much because of the
image quality, but because of how its image changes and how our viewing is altered during the course of its
protracted screening.” Picard suggests that the complete structure of Empire invokes, in three stages,
romanticism (as the monolithic building emerges from the fog), “expressionistic suspense” (in the “brazen”
flickering of lights for almost seven hours), and a final lulling calm after the floodlights are extinguished in
the penultimate reel. As the last two reels unfold in near complete darkness, Picard observes that the film's
surplus of duration allows “for a sustained meditation on cinema itself, its materiality and the status of
17
Ivone Margulies follows this line by suggesting that “in Warhol, interest is challenged on all fronts: the filmmaker is
absent, the object is banal, the spectator is bored.” This results from Warhol's promotion of “an overvisibility where
it is superfluous” (Margulies 1996: 39).
51
representation,” a space of reflection that is brought about by a “determined frontality and stasis [that]
mounts an assault against the ephemeral” (Picard 2007: 67). Equally, Mekas, in a segment of an audio
interview that is worth quoting at length, describes the purpose of reduction and duration in Warhol's film
(and, by implication, key aspects of non-narrative durational style in general):
Empire is an experience, there are many things that happen. At first, when watching the film with
other people, everybody's sort of sceptical, they think: this will be so boring, that they will be leaving
soon, but at least they have to see something... And then, as time goes by, they begin to relax, to
enjoy, to just watch the screen when nothing really much happens. There's some dust, and then, one
hour or so into the film, the light comes up! This huge, incredible event happens when the building
lights up. So, of course, everybody applauds, it's a great moment. And then, again, you relax, and
you watch, there's some light activity, the building is there, it becomes like a meditation. Those who
stay until the end, they all say it is a very meditative, very relaxing, unique experience: just accept
what's there, don't ask from it anything, because the activity is really the most, most minimal. There
is nothing else like that in cinema (Arnold 2011).18
It is perhaps telling that this critical recuperation of Empire's temporal fullness and accumulation of material
detail has arisen in the most active era of slow cinema. Here, Warhol's production of duration is no longer
pitched as a “stupendously perverse” act, but rather as the formulation of an adequate container for
contemplative spectatorial practice (Koch 1991: 60).
Although I have chosen to focus on a film structured by non-human content here, it is worth noting
that Warhol's 'portraiture' works are equally significant in the context of the evolution of contemporary
cinema. Paul Arthur describes these films, which are usually posed in a studio or vernacular setting, as
consisting of “frontal mid-range compositions in which a subject’s face and hands are privileged foci of
information and/or expression” whilst often engaged in a banal or oral activity (Arthur 2002: 95). In Warhol's
Screen Tests, Eat (1963) and Henry Geldzahler (1964), a series of sitters improvise poses within fixed
18
Mekas' comments here are reminiscent of Henry Geldzahler's reflections on Sleep (1963), in which Geldzahler notes
that “the more that is eliminated the greater concentration is possible on the spare remaining essentials. The slightest
variation becomes an event [and] as less and less happens on the screen, we become satisfied with almost nothing”
(Geldzahler 1964-1965: 13)
52
compositions that emphasise the immediacy of present-tense observation. Henry Geldzahler, for instance,
records the time that it takes for Geldzahler himself to smoke a cigar (eighty-three minutes in total, although
twenty minutes pass before he first drops ash into a decorative tray) in two uninterrupted forty-eight minute
shots (again, projected at 16 fps) that catalogue his bodily gestures: holding a few poses, taking his glasses
on and off (to clean their lenses and chew their arms), and wearily attempting to return the camera and
spectator's gaze. Arthur notes that, of all Warhol's portraits, “ironically, those with the least amount of overt
movement register the strongest impression of temporal flow,” an effect that is replicated in more recent
durational films (Arthur 2002: 109).
Portraiture is a recurring visual strategy in contemporary slow cinema, and is predominantly
employed either as a structural device for an overall stylistic system, or for the purpose of sustained visual
studies at the end of narrative or hybrid films. Compositional portraits appear at the close of Tsai Mingliang's
Vive l'amour (1994) (in a fixed six-minute close shot of Yang Kuei-mei crying and smoking a
cigarette), Lisandro Alonso's La libertad (2001) (in a shot that I will describe in Chapter Three), and in the
final image of Ben Rivers' Two Years at Sea (2011) (in which Jake Williams sits and eats, lit only by the
sparks of a dying fire). Portraiture is also integrated into the aesthetics of documentary or hybrid films as a
means of framing performance, and structures the representation of activities (speech, song and everyday
habits) in Jia Zhang-ke's 24 City (2008), Abbas Kiarostami's Shirin (2008), Pedro Costa's Ne change rien
(2009) and Benning's aforementioned Twenty Cigarettes. Most radically, Wang Bing's Fengming: A Chinese
Memoir (2007) consists of ten shots (of starkly varying durations) that depict He Fengming, an ageing
woman, talking at great length of her experience as a student and journalist after the end of the War of
Liberation and under Maoism. The first take, filmed in Fengming's living room, is a static set-up that records
her oral memories in an unbroken composition for two hours and ten minutes [Fig. 14]. Over the course of
the first hour of the shot, the level of visibility within the room slowly and interminably fades until
Fengming (thankfully) gets up to turn a light on, and the camera's literal attention to the domestic space
means that it even continues to record an 'empty' frame when she takes a toilet break.
In order to further trace the lineage of this observational practice in recent slow cinema, I will now
turn to an extended take in a contemporary film that offers both a visual and temporal homage to Warhol's
53
Fig. 14: Fengming: A Chinese Memoir (2007)
innovation: the final shot of James Benning's Ruhr (2009).19
Ruhr is Benning’s second digital (HD) work,
following his minute-long trailer for the Viennale in 2009, Fire and Rain, and was the first of many features
that he has now shot on the format. The film features seven shots that vary in duration from seven to sixty
minutes over the course of two hours, and is divided into two clear halves: the first six shots are labelled
“1.”, and “2.” is reserved solely for the monumental final shot. Ruhr was initially commissioned and planned
as a year-long observation of industrial processes in the steel mills of the Ruhrgebeit district, but only one
shot (the second) actually takes place inside one of the mills (at HKM Steelworks). The rest of the film
documents sights and sounds that Benning witnessed and gathered throughout the region: the interior of a
concrete tunnel under one of the mills, a stretch of forest near Duisberg Airport, a prayer room in a mosque,
the sight of a masked figure removing graffiti from Richard Serra's sculpture Bramme für das Ruhrgebiet
(1998), and a quiet residential street (for a more detailed description of these six shots, see Flanagan 2010).
The last shot occupies the entire second half of the film, and comprises, like Empire, a static composition
that draws attention to a single monolithic structure: the CSQ quenching tower at Schwelgern coke plant
19
A more literal homage is Phil Solomon's “EMPIRE”, a forty-eight-minute video work that consists of a static shot
of the Empire State Building in “Liberty City” (rather than Manhattan) captured from the video game Grand Theft
Auto IV (2008). Solomon played the game until he found a vantage point from which to record the iconic building,
and his film was completed firstly as an installation work in 2008, and as a fixed video work in 2010.
54
[Figs. 15-16].
Benning has not admitted that this particular shot was directly inspired by Empire, but a fully
conscious reference, as Mark Peranson suggests, is likely (see Peranson 2010: 57).20
Like Empire, the tower
in Ruhr is filmed as darkness sets, and its cyclical production of water vapour repeatedly replicates the sense
of a prolonged emergence from thick fog—that is, the 'romanticism' of Empire to which Picard refers above.
Both Warhol's film and the final shot of Ruhr are, at root, durational studies in light, but Benning's is quite
different (and deeply contemporary) in two respects: in its enhanced, even spectacular, cycle of visual
activity, and in its representational status as what D.N. Rodowick refers to in The Virtual Life of Film as a
strictly 'digital depiction' (Rodowick 2007: 110). Unlike Warhol's analogue transcription of the actual time of
filming Empire (on magazines of Tri-X negative stock that extend to forty-eight unbroken minutes per reel
when projected at 16 fps), Benning's shot is seamlessly edited to compress ninety minutes of original footage
shot on location into an apparently continuous hour (or, more precisely, fifty-seven minutes in its sole
television broadcast on 3sat at 25 fps) in its final version.
I will return to the issue of 'graphic montage' (to adapt Rodowick's description of “graphism” in new
media) in contemporary durational film in my last chapter, but it is worth noting here that I do not consider
the graphic compression of 'actual' time in Ruhr's final shot to raise an ontological issue that invalidates its
impression of continuous duration (Rodowick 2007: 105).21
The primary operation of Benning's shot is to
embody an observational rather than materialist practice—an emphasis that, to a certain degree, I am
attempting to reclaim for Empire above. It is also worth mentioning (again) that Warhol deliberately intended
Empire to be slowed to 16 fps during projection in order to create an effect of 'drugtime' (or 'dragtime'), and
Benning manipulates layers of duration in the same way: as a means of drawing attention to subtle changes
in registrations of light and activity that have taken place in time. Ruhr is perhaps best described, in Mark
Peranson’s words, as “a reality-directed document”—a term that might be equally applicable to other slow
films, regardless of their digital or analogue origins (Peranson 2010: 56).
20
Mark Peranson also suggests that, rather than the Empire State Building, the shot “brings with it associations of
another New York building engulfed in smoke and flame,” but apparently Benning did not think of the World Trade
Center “until sometime during the editing process” (Peranson 2010: 57).
21
Benning uses this graphic strategy in other shots in Ruhr: the first is composed of six seamlessly branched takes, and
Benning cut the appearance of a number of people during the sixth shot to reduce its visible activity. Benning has
stated that “on [analogue] film, if you do a dissolve there’d probably be a slight light shift, and if anybody looks
closely they could perceive it. I have done that with film before, cheating a little bit and people don’t see it, but with
[digital] I know you won’t see it: there’s no indication” (Benning and Nord 2009).
55
Fig. 15: Ruhr (2009)
The tower depicted in the final shot of Ruhr is approximately seventy metres tall (dwarfed by the
Empire State Building's roof height of 381 metres), and designed for an optimised wet quenching system
(CSQ) that was specially developed for the Schwelgern plant in the Ruhrgebeit district. The structure of
activity in Benning's fixed shot is organised by a cyclical process in which super-heated coke is drenched by
over a hundred cubic metres of water, which produces a thick cumulus of vapour that repeatedly surges out
of the filtered sides and top of the tower. This process takes place five times during the shot: firstly, within
the second minute, then after eleven and a half minutes, thirty minutes, forty-one minutes and fifty-one and a
half minutes have passed. Every time a cloud of vapour is expelled from the tower, a period of between five
and seven minutes goes by before the air surrounding the structure clears, and a compositional stillness
returns to the image.
The shot's limited disclosure of visible (and aural) information points to a complex process that is
otherwise unseen, but should be considered as relevant as elements of plot or backstory in a (classical or
modern) narrative film. The procedure depicted in Ruhr takes place thirty-five times a day at the Schwelgern
plant, on every single day of the year. In preparation for the procedure, coal is heated at an intense
temperature (within two batteries of seventy compound ovens) for twenty-five hours. At regular intervals, the
56
super-heated coke is extracted and transported by rail to the bottom of the quenching tower, when a loud
siren signals its arrival. Over a hundred cubic metres of water is then fed into the bed of coke for seventy
seconds, and the resultant mixture of steam and water rapidly cools the coke and propels a vast amount of
vapour up through the tower. After another siren signals that this process is complete, a series of 'baffle
separators' in the tower remove coke dust from the vapour, which is also 'scrubbed' with water and partly
condensed to restrict particulate emissions. Benning describes this unique wet-quenching system as “state of
the art,” and it is apparently both environmentally “cleaner” and more efficient than the dry cooling system
that it supplants (see New Schwelgern 2012; Benning and Nord 2009).
Benning's observation of this process begins during a period of calm that is disturbed only by the
sound of a plane flying overhead, and a low machinic drone that will continue throughout the entire shot.
After a minute, the first siren blares, and a curl of vapour begins to rise from the top of the tower. This
emanation quickly intensifies to form a thick grey cumulus that balloons out of the top and sides of the
tower, climbing the body of the structure in a swirling, helical motion. By the time the second siren is heard
(to signal the removal of the coke car), the entire structure of the tower is engulfed in billowing movement.
After five minutes, the vapour in the atmosphere begins to dissipate, and the tower is restored to full
visibility after another two minutes have passed. At this point, a visible stillness descends upon the frame
again, and the spectator's attention is encouraged to shift to the cloud formations surrounding the tower—
which, by comparison, appear to be eerily motionless. After nine and a half minutes, a bird flies across the
frame, and its distant song is, surprisingly, audible over the industrial drone.
The scale, intensity and colour of vapour emissions varies throughout the rest of the shot. The second
eruption is the most violent, producing a continuous vortex that quickly consumes the entire frame, and takes
the longest time to disperse. Benning's graphic compression of the actual time of his original take then gives
rise to an uncanny dip in ambient light, and the wall of the tower is pressed into shadow for the rest of the
shot. After seventeen minutes have passed, the image is given over to a sustained thirteen-minute lull in
which industrial production is suspended, and activity is limited to a couple of flocks of birds that make their
way across the frame. This extended temps mort is interrupted by a more serene eruption of vapour (the
third), now discoloured by an altered configuration of light between the structure's darkening shadow and the
blue twilight sky. As night begins to fall, the vapour's hue mutates again during the final two emissions: from
57
Fig. 16: Ruhr (2009)
a light grey and acrid yellow to a charcoal grey and steel blue. The last emission, in which vapour
unpredictably rises at first from the bottom of the frame rather than the top of the tower, quietly occurs after
fifty-one and a half minutes. A calm post-action lag is accompanied by the sound of a plane flying overhead,
which provides a cyclical closure to the extended take.
Whereas Warhol's structural films were unconditionally silent, Benning adds a carefully modulated
soundtrack to provide an evenness of tone to this repetition of activity within the shot. In the second half, the
sound of the quenching procedure is gradually yet noticeably softened, so that the noise of siren blasts and
coke cars travelling to and from the tower are quietened to reflect the failing quality of light.
This careful (re)mastering of industrial noise follows Benning's sonic contribution to Sharon
Lockhart's Lunch Break (2008), a structural film that might be thought of as a visual homage to both Michael
Snow's Wavelength (1967) (in its inexorable transformation of the visual field) and Ernie Gehr's Serene
Velocity (1970) (in its systematic focal adjustment of the visible space of a corridor). Lunch Break consists of
one steady continuous tracking shot, recorded on a single reel of 35mm, down a 1200ft corridor in Bath Iron
Works (a shipyard in Maine that builds battleships, destroyers and torpedoes for the United States Navy, and
is the largest employer in the state) [Fig. 17]. The original take was shot during a lunch hour, and depicts a
58
Fig. 17: Lunch Break (2008)
number of workers eating, reading, talking, opening lockers, walking down the corridor, climbing ladders,
and so on. In post-production, the ten-minute duration of the original shot was digitally slowed to an eighth
of normal speed, and runs for eighty minutes in its final version. Like Ruhr, Lunch Break is a 'digital
depiction' that reconfigures time to draw attention to the slightest activity and detail within a concentrated
visual field.
Lockhart distends the time of the shot in Lunch Break with considerable precision. As Benning has
noted, the ten minutes of 35mm film recorded in the corridor captured 14,400 still frames (exposed at 24
fps), which matched the exact length, of 14,400 inches, of the space of the corridor. On transferring the
35mm reel to HD video, Lockhart duplicated each frame (one for every inch of the distance travelled by the
camera) eight times to lengthen the film to its final eighty-minute duration. This means that, upon projection,
each individual frame is quite literally visible for eight times longer on the screen (Benning and Lockhart
59
2010: 101). This strategy, loosely reminiscent of Warhol's 'dragtime' and the slow forty-five-minute zoom of
Wavelength, prolongs the projection of (what was once) actual time, animating still images of objects and
humans for the purpose of a more active spectatorial attention to detail. Lockhart, in conversation with
Benning, describes the effect of watching the shot in its extended duration:
The slowness allows you to get past your first impressions and engage the space. When you see the
shot in real time, you are more likely to succumb to those first impressions and not notice the
subtleties of the space and of the gestures of the people. But in the slowed-down version, you
experience the space more as you would experience it in a photograph or a succession of film stills.
[...] When you’re watching the film, it seems very slow, but then you have these moments in which
you realise that you just missed some detail that you suddenly become aware of and want to see
more closely (Benning and Lockhart 2010: 106).
Although the sequencing of moving images in Lunch Break is actively decelerated, the film's sonic
accompaniment remains in real time, and was mixed by Benning and Lockhart in collaboration with Becky
Allen. Elements of the soundtrack were compiled from a series of in situ walk-throughs of the path of the
camera with composer and sonic ethnographer Ernst Karel (who also worked on Lucien Castaing-Taylor and
Ilisa Barbash's 'in-between' film Sweetgrass [2009]), as well as from a composition written by Allen on a
1960s electronic keyboard. Benning and Allen mixed oscillations from the keyboard with frequencies of
industrial noise recorded in the hallway by Karel in real time, and the resultant composite drone, as Lockhart
notes, ensures that the listener is unable to tell “which tones are 'music' and which are the sounds of the
machines themselves” (Benning and Lockhart 2010: 106; see also Eckmann 2010). Sounds that are audible
during the full film include pitch-shifted machinic noise, snatches of radio broadcasts and unintelligible
conversation, and aural 'pulsations' when the camera passes pipes or large machinery. The sonic effect of this
composite sound, as Benning commented of filming in the Matenastraße Tunnel (an underground tunnel that
stretches for 0.5km beneath HKM Steelworks, and provides the space of the first shot of Ruhr), is “like
living in your eardrum” (Benning and Nord 2009). Throughout Lunch Break, the soundtrack rises and falls in
intensity and volume, and quietens during the final twenty minutes—a symphonic dynamic that is revisited
60
Fig. 18: Exit (2008)
by Benning in order to structure the interplay of light and sound during the final shot of Ruhr.
Lunch Break is part of a composite installation work that also includes three suites of photographs
that depict the everyday lives of the workers at Bath Iron Works, another film, and a newspaper (The Lunch
Break Times) that was published and distributed at the factory whilst Lockhart worked on the project. The
photographic suites focus on three aspects of daily working life: group portraits of the workers during their
lunch breaks (in highly staged poses); still lifes of individual lunch boxes; and documents of the independent
businesses housed at the side of the long corridor in Lunch Break (which are run by some workers to sell
lunch items to others during the work day). Exit (2008), the second film for the project, provides an
observational counterpart to Lunch Break by presenting five eight-minute shots that record the workers
leaving the factory building at the end of the day [Fig. 18].
The structure of Exit comprises a single shot for each day of the working week (divided by intertitles
61
that state “Monday”, “Tuesday”, and so on), and the visible activity in each one is almost constant: workers
initially flood past the camera before thinning out as the minutes pass, at times crossing paths in alternate
directions, and rarely pausing within the visual field. Each day brings with it a different quality of light, and
conditions are sunny and clear on Tuesday and Wednesday (in turn casting areas of the image into shadow),
and noticeably more leaden on Monday, Thursday and Friday. Following the observational tendencies of
1960s and 1970s experimental cinema, Exit returns to the work of the Lumière brothers, principally to Sortie
d'usine (1895): a forty-six-second film that records, without mediation, approximately one hundred workers
exiting a factory through two gates. In Sortie d'usine, as Sabine Eckmann notes, the exiting workers are
filmed by the Lumière brothers from a frontal set-up whereby the camera points toward the factory gate,
meaning that they immediately flood out of the sides of the frame as soon as they appear (Eckmann 2010:
26).22
Lockhart, however, stages a deep-focus sequence shot for Exit that draws attention to the workers' long
walk away from the camera, giving the spectator time to take in the actuality and continuous repetition of
this activity. In Exit, the five autonomous images contain no narrational strategy beyond a fixed, sustained
registration of the everyday, and signify the continuum of an observational strategy actively retained by
contemporary cinema from the durational principles of 1960s and 1970s North American experimental film.
22
For a more detailed analysis of the visual field of Sortie d'usine, see, in particular, Perez 1998: 52-53.
62
CHAPTER TWO
THE MODERNS, BAZIN, AND THE DURATIONAL SEQUENCE SHOT
I have suggested thus far that the field of contemporary slow cinema is contingent on a particular
combination of formal and structural devices evident in post-war art and experimental film. This contingency
can be summarised by the incorporation of three principal characteristics: a minimal, or non-, narrative
structure; the use of the long take, often accompanied by the long shot, to present unspectacular events in
their entirety (including passages of temps mort, or substantial pre- and post-action lags); and an emphasis
upon stillness, both of the frame and of visual content (in order to draw attention to composition and minor,
localised movement). Additional strategies that are employed by these durational films include slow camera
movement (in mobile long takes), the 'evacuation' of the frame (achieved by placing the camera in extreme
proximity to, or distance from, people and objects), the obscuration of the actor (or a visual emphasis on
inexpressiveness of performance) and, occasionally, an abstraction of the image in the form of filters, colour
gradation, animation or digital effects—as in some films by Aleksandr Sokurov (Kamen [1992], Whispering
Pages [1994], A Humble Life [1997] and Mother and Son [1997]) and Jia Zhang-ke (The World [2004] and
Still Life [2006]).
As might be expected, there is a considerable degree of fluidity across this broad stylistic model. For
instance, some divergent films, like those of Sharunas Bartas or Peter Hutton, employ shot durations that do
not signal a durational style as rarefied as other films (like the work of Lav Diaz or Béla Tarr, whose work
exclusively measures ASLs in terms of minutes), and instead exhibit other elements that distinguish them
from canonical art cinema. In terms of shot lengths, Bartas's Trys dienos (1991) and Koridorius (1994) have
ASLs of 20.9 and 27.9 seconds, whilst Hutton's vary throughout his long career: 30.3 seconds in New York
Portrait: Chapter I (1979), 33.8 in Study of a River (1996-1997), 45.8 in Time and Tide (1998-2000), and
44.2 in At Sea (2007). Bartas and Hutton's films remain particularly distinctive, however, for their
combination of other devices: anti-narrative structures, resolutely still framings, as well as a stark despondent
realism (in Bartas' work) and a documentary fixation on landscape and 'minor' material detail (in Hutton's
63
work).23
The field of cinema that I am referring to as slow, then, is changeable and diverse (in terms of form,
narrativity, fictionality, genre and tone), and sufficiently broad categorisations like 'aesthetic of slow' or
'contemplative cinema' have become both appropriate and inevitable (see Flanagan 2008; Cummings 2007).
My own approach is intended to retain this openness to stylistic fluidity, but will also entail another purpose
that extends beyond the simple classification of slowness. In order to consider the durational art and
experimental film under discussion here within a suitably broad historical and theoretical framework, I
believe it will be necessary to consider the contemporary cinema in light of the theoretical basis and mode of
film practice that constitutes post-war modern cinema.
The concept of modern cinema that I will primarily invoke in this chapter is rooted in the
complementary models proposed by André Bazin and Gilles Deleuze. In relation to the evolution of slowness
in post-war cinema, I will remain faithful here to Deleuze's conception of modern cinema as a consequence
of the radical break in image-sound relations following the Second World War, as well as the realist
dimension and emphasis on the long take that is prescribed by Bazin's model. My primary purpose for
addressing the evolution of modern cinema in this chapter will be to provide a film-historical understanding
of the way in which modern films revoke a tendency toward dramatic narrativity in dominant cinema, and
replace that element with other possibilities for image-sound situations, such as observational realism and an
emphasis on the production of duration.
My contention in this chapter is that key recent durational narrative films visibly embody key
elements of Bazin and Deleuze's theories in their realist application of the long take, fullness of duration
within the shot, and artistic positioning in relation to the world. Robert Koehler has suggested that Bazin's
notion of a “truly realist cinema of time” (a phrase describing what Bazin saw as the potential for
undramaticness in Italian neorealism) represents an aesthetic aspiration that “required fifty years to be fully
realised, as it was [in 2001] by Lisandro Alonso with La libertad, which is starting to look like the most
important film of this era” (Bazin 2005b: 76; Koehler 2008b). Koehler, describing the contours of
23
I would rather not consider the boundaries to be too prescriptive here. Some significant contemporary films, such as
Sherad Anthony Sanchez's Imburnal (2008), merge extended takes (still, observational portraits that last five
minutes or longer) with rapid, fragmented montage techniques (such as jump cuts and discursive reframings).
Equally, many unconventional modern or durational strategies are frequently integrated into contemporary 'festival'
films with more diverse styles.
64
contemporary slow cinema, then proposes that the aesthetic and ideological basis of Alonso's enactment of
Bazin's proposition (to which I shall return in my next chapter) “has become a considerable international
movement, embraced wholeheartedly by filmmakers of many different stripes on every continent (Benning
to Weerasethakul, Jia to Costa)” (Koehler 2008b).
In light of this, returning to the implications of Bazin and Deleuze's models of modern cinema will, I
hope, provide a strong foundation for understanding aspects of the origin and theoretical operation of
contemporary slow cinema and its antecedents. This chapter will include a potted history of Bazin and
Deleuze's notions of modern cinema; an extended focus on Bazin's theory of realism and the sequence shot; a
section on Bazinian reflections on the long take by contemporary filmmakers; and two key contemporary
examples of the Bazinian sequence shot. My attention to Deleuze's model will be limited here, as I shall
return to it in more detail (in the context of the concept of the 'still life') in Chapter Four. My motivation for
focussing on Bazin in such detail is twofold: firstly, to reclaim key aspects of his work, and restore its
neglected vitality in line with recent efforts to do the same (see, in particular, Keathley 2006; D. Andrew
2010); and, secondly, to propose that his model of preserving temporal continuum within the long take
continues to comprise the essential operation of contemporary durational style.
Reflecting on the particularity of Bazin's realist theory will also serve, I hope, as a means of
addressing the nature of realism in slow and durational film, and in turn propose that its contemporary
filmmakers substantiate Bazin's notion of the positioning of the artist in relation to the world (as a modern
problematic of realist representation). Equally, returning to Bazin's theory will entail addressing a key issue,
and question, of durational film: why not cut? I will suggest here that Bazin's theory continues to provide a
strong basis for understanding contemporary cinema's resistance to montage, and the utility of its application
of the long take.
Modern cinema
In his two volumes on the cinema (see Deleuze 2005a; 2005b), Deleuze's conceptual history of modern film
charts its evolution between the immediate post-war era and the early 1980s, setting up an illustrative
division between contrasting modes of image-sound production both prior to and during that period. In
65
Deleuze's schema, pre-war cinema is defined by the regime of the movement-image (exemplified by the
action-image system of classical and montagist film), which enters a period of crisis in the 1940s when
confronted by the trauma of the post-war proliferation of “spaces which we no longer know how to describe”
(Deleuze 2005b: xi). In the new post-war environment, the action-image emblematic of the movement-image
regime begins to break down and give rise to the time-image: an uncertain, dislocated set of image-sound
relations in which time “rises up to the surface of the screen” (Deleuze 2005b: xii). The time-image,
unshackled from the sensory-motor schema of the action-image, formulates a direct presentation of time
formerly submerged by the movement-image, which continues its operation regardless. In this schema, the
organic, ordered representational system of the action-image is “shattered from the inside” by the wreckage
of the aftermath of war, and cinema embarks on a new path to reconnect our perception of the world with the
new, uncertain reality of the post-war era (Deleuze 2005b: 39).
As Jacques Rancière makes clear, the modernity of post-war cinema confronts “the classical cinema
of the link between images for the purposes of narrative continuity and meaning with an autonomous power
of the image whose two defining characteristics are its autonomous temporality and the void that separates it
from other images” (Rancière 2006: 107). This confrontation yields an entirely new set of image-sound
possibilities, although the classical utility of narrative continuity still retains its dominance, most
prominently in the enduring, modified classicism of North American and transnational commercial cinema.
In Rancière's classification, modern cinema comprises both a new temporal autonomy of the image
(severed as it is from the need to conform to what Adrian Martin refers to as the “regulated constraint” of
classicism), and the void: a proliferation of gaps, ruptures and dislocations between and sometimes within
images, now that the old system of linking image-sound situations has broken down (Martin 2010). The
former strategy manifests itself in the experiments in duration that lead to contemporary slow cinema (most
notably in the tendency toward long-take styles and 'temporal excess'), and the latter in the elliptical,
synthetic experiments of modernist montage. The latter form, as Rancière notes, has played the most
distinctive role in post-war cinema, establishing a cinematographic modernity in which “each image actually
emerges from the void and lapses back into it, so much so, in fact, that it is the interstice, the separation
between images, that plays the decisive role in modern cinema” (Rancière 2006: 114).
Definitions of the theoretical basis and historical specificity of modern cinema vary amongst its key
66
exponents. As András Bálint Kovács has suggested, modern cinema is not purely connected by formal traits,
and “cannot be considered to be a homogeneous style any more than as a set of incommensurable and totally
unique works of art” (Kovács 2007: 3). There is also a sense in which, as Adrian Martin argues, “the cinema
has always been Modern,” and it should be noted that the post-war period to which I refer here is preceded
by one other major irruption of cinematographic modernity: the spillover of surrealism, Dadaism, cubism
and futurism into the cinema of the 1920s, loosely bookended by the experimental classicism of German
Expressionism and the socialist ecstasy of Dziga Vertov's Man with a Movie Camera (1929) and
Enthusiasm: Symphony of the Donbass (1930) (Martin 2010). The use of the term 'modern cinema' by
Deleuze and Martin signifies more than the return of modernism in the post-war period however, and
incorporates that resurgent trend of the 1960s and 1970s within a broader theoretical and historical
framework.
Helpfully, Kovács draws a distinction between agendas that divide 'moderns' and modernists, and
separates these two sets of theoreticians into groups of 'evolutionists' and 'style analysts': whereas the latter
group makes sure to distinguish between 'modern cinema' and the diverse set of stylistic and narrative
characteristics that constitute modernism, 'evolutionists' do not, preferring to fix instead upon “a single and
synthetic idea about modern cinema”—that is, the belief that modern cinema is always actual and valid in
opposition to classicism (in its historical and reheated forms) (Kovács 2007: 34, 44). Modernist trends such
as the duality between avant-garde and mainstream narrative cinema identified in David Bordwell's
“oppositional program” (drawn primarily from Noël Burch's Theory of Film Practice [1973]) belong to the
latter, and Deleuze's model resides in the former group (see Bordwell 1997: 83-115). My interpretation of the
evolution of key aspects of slow cinema derives primarily from the former: that is, from the (rather
unfortunately titled) 'evolutionism' in terms of its theoretical basis, but also accepting of the distinction
between modernist practice and a form of modern cinema that continues after the demise of modernism in
the late 1970s (an issue that I will return to in my next chapter). It is worth noting that Bazin's own
conception of modern cinema complicates these boundaries, as he implies that a transformation in realist
film style in the 1940s represents as significant a break as the evolutionist perspective on cinema's potential
for an 'art of realism'.
Modern cinema, from the 'evolutionist' perspective, signifies the modern condition of a dislocation in
67
the human relation to the world (one in which we are not ourselves) that is actualised or confronted in film
by a set of formal strategies and innovations. Gabriel Josipovici, in his important monograph on the waning
of modernism in contemporary literature, suggests that this condition constitutes a 'primal time' of doubt and
uncertainty that is always present for us, and one that artists return to, and struggle with, by means of “a
dialogue between us and the world, a dialogue conducted through the medium of our art” (Josipovici 2011:
97). The modernist impulse, for Josipovici, always takes the form of “a response by artists” to a sense of the
'disenchantment of the world', and he proposes a redefinition of modernism as “the coming into awareness
by art of its precarious status and responsibilities, and therefore as something that will, from now on, always
be with us” (Josipovici 2011: 11).
This modern condition manifests itself in art during periods of societal turmoil or decisive change,
such as the period that signalled the emergence of the novel after reigning traditions and tyrannies of the
sixteenth century entered a state of terminal decline. In that era, as Protestantism and Humanism collectively
brought into being a new world, Josipovici notes that “it is no coincidence that the novel emerges at the very
moment when the world is growing disenchanted” (Josipovici 2011: 34). The emergence of the time-image
regime in cinema constitutes another instance of this tumultuous irruption of the Modern, and, following
Deleuze's schema, provides a baseline framework for understanding the decisive turn taken by art and
experimental film after the Second World War. As existing forms of responding to the world no longer
proved adequate in the new conditions of the immediate post-war era, artists turned to new forms actualised
by the instabilities of montage relations and the disciplined, observational realism that forms the basis of the
field of slow cinema. This second strategy, tentatively expressed in stylistic terms in Bazin's model of
modern cinema, enabled film form to relate to the world with a renewed clarity and directness, liberated from
the latent enchantment of dominant forms of classicism.
Deleuze's conception of modern cinema directly takes its cue from Bazin's, whose definition posits
the break as occurring between 1939 and 1941: the point at which Jean Renoir's La règle du jeu (1939),
Orson Welles' Citizen Kane and William Wyler's The Little Foxes (both 1941) herald a stylistic expansion of
extant classicism by developing new strategies based on the application of the long take and focus in depth.
Deleuze draws a line from Bazin's notion of the 'image-fact' to the emergence of 'pure optical-sound
situations' in post-war Italian neorealism (see Deleuze 2005b: 1-12), but, for Bazin, the break occurs at the
68
apotheosis of the “level of classical perfection” achieved by American cinema in the late 1930s, followed by
the introduction of panchromatic film stock within studio production (that enabled an enhanced sharpness of
focus when filming with a smaller aperture) (Bazin 2005a: 30). After the innovations of Welles, Wyler and
Gregg Toland that signify this break, Bazinian modernity is more commonly located in European cinema: in
the location shooting and undramatic narrativity of post-war Italian neorealism, the “almost unbearable”
rigour of Robert Bresson's Journal d'un curé de campagne (1951), and what Bazin referred to as the 'avantgarde'
narration of Roberto Rossellini's Viaggio in Italia (1953) and Agnès Varda's La Pointe Courte (1954)
(Bazin 2005a: 131).24
Modern cinema, in a broad historical sense, undergoes different stages after this opening phase
charted by Bazin. After both neorealism and the emergence of de-dramatised art cinema, by the 1970s a more
radical “constellation” of starkly modernist films had emerged, as Denis Lévy identifies, in the cinema of, for
instance, Chantal Akerman, Jean-Luc Godard, Manoel de Oliviera, and Jean-Marie Straub and Danièle
Huillet (Lévy 2011). After this point, radical modernism began to wane quite rapidly, and the main body of
recent slow cinema begins to emerge. Modern cinema, in this historical interpretation, simply comprises a
continuous order of interruptions of “the system of conditions” of the dominant aesthetic system of
classicism (or, in Lévy's term, 'Realism') (Lévy 2011). If we are to consider modern cinema as signalling a
historical continuum in this sense, it remains important not to sever the formal and realist tendencies of more
recent slow films from their roots. The operative aesthetics of contemporary slow cinema, after decades of
transition, do appear to derive in substantial part from both post-war Italian neorealism (most prominently in
Lisandro Alonso's La libertad, Los muertos [2004] and Liverpool [2008], as I shall suggest in more detail in
my next chapter, and Jia Zhang-ke's Xiao Wu [1997], Unknown Pleasures [2002] and Still Life) as well as the
early art cinema of, for instance, Antonioni's 'great tetralogy' (from L'Avventura [1959] to Il deserto rosso
[1964], to which many contemporary 'festival films' display a strong influence).
As Adrian Martin notes, the modernity of the work of Antonioni, Bergman and Godard had, by the
end of the 1950s, already mapped out “all the explorations and experimentations that define modern cinema”
in embryo, in their rejection of classicism and adoption of radical approaches to narrative, performance, body
and gesture. In the history of modern cinema to follow, Martin notes that some of these elements are duly
24
For a more detailed description of Bazin's 'evolutionist' conception of modern cinema and its relation to the
emergent avant-garde, see D. Andrew 2010: 99-110.
69
obliterated altogether, producing a cinema capable of operating with “no stories, no bodies, no linear time, no
discernible real world” (Martin 2010). This evolution is fulfilled in more recent slow cinema by the
experimental practice of, for instance, Sharon Lockhart, James Benning and Peter Hutton, which tends to
dispense with narrative and human presence where necessary.25
Hence, in Martin's suitably expansive definition, modern cinema incorporates both art and
experimental, as well as narrative and non-narrative, film. Just as Deleuze's canon prioritises key late
modernist filmmakers (Godard, Duras and Straub-Huillet), modern realists (Jean Rouch, Pierre Perrault) and
late classicists heralded by Bazin (Orson Welles), contemporary slow cinema derives from an equally diverse
base. It is clear that, echoing Kovács's observation regarding late modernism, that the diversity of the postwar
time-image regime has enabled “every possible manner of transgression of classical narrative style”
(Kovács 2007: 207). In this respect, the set of formal devices employed by recent slow cinema represents one
such potential outcome from the core elements of modern cinema, and I believe its formal transgressions
must be understood in the context of earlier innovations in both theory and practice.
Bazin's art of realism
In both Bazin and Deleuze's schema, Orson Welles' Citizen Kane is the key text for cinema's emergent
modernity, representing “the first occasion on which a direct time-image was seen in the cinema” (Deleuze
2005b: 102). For Deleuze, Citizen Kane unlocks a new structure of the image, one that is able to manifest a
continuum of regions of time—that is, of the narrative's past and present—on the surface of the screen.
Welles' combination of the long take and deep focus within the shot presents a succession of states between
action and recollection (“regions of the past”), that lead to an actualisation of how cinema thinks the world it
represents (Deleuze 2005b: 103). For Bazin, however, the modernity of Citizen Kane is primarily realist in
an analogue sense: in the devices employed by Welles, cinema is able to outgrow its status as a 'realistic art',
and return to the continuum of reality itself. This is in line with Bazin's prescription for the medium's
aesthetic capacity as a whole: that it “will be fulfilled when, no longer claiming to be an art of reality, it
becomes merely reality made art” (Bazin 1992: 119). The transition to a modern cinema in Bazin's schema
25
As Benning himself has recently quipped, “I think people are crazy, and consequently my latest work features few
or no people” (Benning and Lockhart 2010: 102).
70
derives from the “profoundly aesthetic” way in which reality is disclosed within the shot—or, less ideally,
within a carefully constructed sequence of shots (Bazin 2005b: 25). The aim of what we might think of as the
Bazinian model of the disclosure of reality in the shot comprises “a way of considering reality as if it were
homogeneous and indivisible, as if it had the same density of all coordinates on the screen” (Bazin 1997:
235).
Bazin’s realist system derives from his belief that the quest for a realism of the image has constituted
“the essence of film from the very start (one might even say as early as its seed took root in the inventors’
imagination)” (Bazin 1997: 88). This essence, for Bazin, initially arises from, but is not beholden to,
cinema’s indexical function: the process by which a camera automatically formulates an image of the real by
impressing light on a sensitised photographic plate, before transferring it to photochemical film. The
objectivity of this machinic function leads Bazin to suggest that cinema’s profound vocation “is to show
before it expresses, or, more accurately, to express through the evidence of the real” (Bazin 1997: 90; my
emphasis). Bazin’s theory can, as Brian Henderson has noted, then be divided into two clear systems: an
ontological concern, regarding cinema’s indexical function, and a historical system that serves to evaluate,
and privilege, this position on the level of aesthetics. It is here that Bazin “downgrades any kind of form
except that subservient to the form of the real,” shaping in advance the direction of his model of modern
cinema (Henderson 1980: 28). Throughout his writing, Bazin’s schema privileges a certain type of fiction
and documentary film: one that is resistant to an overtly 'forceful' presentation of its content or material
record; patient and observational in its style; and respectful of the duration of the impression of reality that it
captures (Bazin 2005a: 32).
The notions of objectivity and indexicality upon which Bazin builds his realist theory have been
subject to severe scrutiny by film scholars in many different fields of study, but it is important to note that, as
I have previously mentioned, certain aspects of his work (and its perceived naiveté) have recently been
salvaged by theorists such as Paul Willemen (1994), Christian Keathley (2006; 2011) and Dudley Andrew
(2010). To elaborate upon this trend, Philip Rosen has noted that “Bazin-bashing” has been something of an
established tradition in film studies since the 1970s, and has been particularly active in critiques of Bazinian
realism, ontology, subjectivity, and phenomenological intentionality (see P. Rosen 2001: 8-14). Alongside
Siegfried Kracauer's starker realist theory (see Kracauer 1997), Bazin's model of realist style has, rightly or
71
wrongly, more often than not been subject to criticism for its assumed naiveté, reductionism or misplaced
medium-essentialism, an approach that is most prominent in Peter Wollen's Signs and Meaning in the
Cinema (1969); the 1970s Screen inquisition (in particular, Colin MacCabe's essay “Theory and Film:
Principles of Realism and Pleasure” [1976]; for a comprehensive overview refer to P. Rosen 2008); Noel
Carroll's exasperating Philosophical Problems of Classical Film Theory (1988) and Theorising the Moving
Image (1996); Victor Perkins' Film as Film (1972); and, more sympathetically, Brian Henderson's A Critique
of Film Theory (1980) and Gilberto Perez's The Material Ghost (1998). Tag Gallagher, in a footnote in his
biography of Roberto Rossellini, has reflected on the most counter-productive aspects of these approaches:
Bazin’s enthusiasm repeatedly inspired him to poetic exaggerations that, unfortunately, American
academics have accepted as well considered statements of film 'theory'. The notion, for example, that
neorealism is styleless. Or the notion that cinema is an imprint of reality, which academics have
glossed with reams of analytic disputation, as though it were intended as the foundation of cinematic
ontology rather than one among many poetic insights into cinema’s experiential problems (Gallagher
1998: 736).
Some film historians, such as David Bordwell and Kristin Thompson, have appropriated the virtues and
limitations of Bazin's historicisation of genre and style with considerable openness (see, most notably,
Thompson 1988; Bordwell 1997: 46-82). However, I would suggest that Bazin's influence has been felt, and
respected, most strongly in work on photography and its ontology (regardless of the implication of
Gallagher's critique): in Stanley Cavell's The World Viewed (1979), Roland Barthes' Camera Lucida (1980),
and Susan Sontag's On Photography (1977), although it is interesting that Barthes mentions Bazin by name
only once (in passing), and Sontag does not cite Bazin at all.26
Only recent work on cinephilia (Willemen and
King 1994; Keathley 2006) has taken Bazin’s ontological inquiry very seriously, in turn rejecting a decadeslong
tradition of 'Bazin-bashing' to signify a resurgence of interest that is strongest in Andrew's What
Cinema Is!: Bazin's Quest and its Charge (2010) and his co-edited collection (with Herve Joubert-
26
Barthes only mentions Bazin in the following passage: “Yet the cinema has a power which at first glance the
Photograph does not have: the screen (as Bazin has remarked) is not a frame but a hideout; the man or woman who
emerges from it continues living: a 'blind field' constantly doubles our partial vision” (Barthes 2000: 55-57).
72
Laurencin), Opening Bazin: Postwar Film Theory and Its Afterlife (2011).
My contention here, in line with these recent interventions (most forcefully by Andrew), is that key
aspects of recent durational film bear out the potential, rather than the strict historical or formal actuality, of
Bazin's model of modern cinema. This approach is rooted in Bazin's proposition of a durational style centred
on the long take, and is perhaps equally appropriate when we consider the contemporary films under
discussion here primarily as responses by artists to the world: that is, instances of a process by which artists
'filter' reality and re-create “the sense of the world coming to life as we look at it and live in it,” as Josipovici
writes of the landscape painting of Paul Cézanne (Josipovici 2011: 96).
The principal difference between much of the cinema of which Bazin wrote (the cinema, necessarily,
of his time) and more recent slow cinema is that the latter presents this sense of the world in a deeply
rigorous form designed for contemplative spectatorial practice. Many of the particular formal innovations
that are evident in the field of slow cinema are unthinkable in line with the classical aesthetics of the
Bazinian model, and are better accommodated by Deleuze's model of the time-image (which poses problems
at the level of 'the mental' rather than at the level of the real). What Bazin's theory helps us to understand
above all, however, is the level of investment in the realism of the image common to much contemporary
slow cinema, in all its necessary states of artifice.
It is important, in this context, not to see Bazin's vision of realism as deriving directly from an
assumed objectivity of cinema's technological function, but rather as a perpetual 'striving' for realism on
behalf of both the artist and spectator. Philip Rosen, in his decisive reading of Bazin in Change Mummified
(2001), suggests that Bazin's notion of realism originates not from a belief in an “absolute concrete
objectivity that cinema can somehow make immediately available,” but rather from the assertion that
photography and cinema, by drawing on their indexical function, “are discoveries that satisfy, once and for
all and in its very essence, our obsession with realism” (P. Rosen 2001: 14; Bazin 2005a: 12; my emphasis).
This obsession with realism, for Bazin, is a psychological preoccupation that has defined the history of the
plastic arts, and is characterised by the concept of the 'mummy complex': the 'embalming' of time that
compels artist and spectator to invest “an unprecedented credibility in the image” in order to affirm the
prolonged existence of their subjective being (P. Rosen 2001: 23).27
Made possible by what D.N. Rodowick
27
See Bazin 2005a: 9-16. Bazin's starting point here is that cinema's indexical function forces us “to accept as real the
existence of the object reproduced, actually re-presented, set before us” (Bazin 2005a: 14). The notion of the
73
refers to as the mechanical “automatic analogical causation” of photography, the evidence of the real
prevails as a principal element of the filmic image, enabling “a subjective striving, a subjective investment in
the image precisely as 'objectivity'” (Rodowick 2007: 49; P. Rosen 2001: 14).
Bazin assumes that artists respond to this anxious, searching condition of the human subject through
style and narrativity, which “necessarily embody an attitude to the world” (P. Rosen 2001: 23). This basis of
subjective intentionality provides the foundation for the more conventional definition of realism, summarised
by Kristin Thompson as, simply, “an effect created by the artwork through the use of conventional devices”
(Thompson 1988: 197). Although sets of realistic devices change over time, filmic realism frequently
prioritises the two constituents that inspire subjective investment in the image: an attention to indexical and
minor detail and the function of time, both of which come to the fore in contemporary slow cinema.
In the broader context of modern cinema, Denis Lévy observes that after the innovations of the
Nouvelle Vague in the 1960s, modern films have persistently complicated this subjectivity by embodying an
awareness that “we do not have to seek adequacy between ideas and things, and that the object and the image
are representations; which is to say, imaginary, and provide only a limited access to reality.” Hence, at the
peak of the modernist era in the 1970s, Lévy argues that modern cinema principally sought to display a
“consciousness of the fact that the real is unrepresentable” (Lévy 2011; my emphasis). This complication of
realism is apparent in many contemporary films, and I would argue that it must be considered an extension to
the notion of Bazinian realism, rather than an ineluctable destruction of it.
In line with certain interventions that have been made with regard to Bazinian theory, I accept here
the axiom held by the editorial team of Cahiers du Cinéma during its militant phase in the late 1960s and
1970s, later summarised by Serge Daney: “L'Axiome Cahiers: c'est que le cinéma a rapport au réel et que le
réel n'est pas le représenté—et basta,” which is translated by Dudley Andrew as “the Cahiers axiom is this:
that the cinema has a fundamental rapport with reality and that the real is not what is represented—and that’s
final” (Daney 1993: 301; D. Andrew 2010: 5). This axiom, following Bazin's implication that cinema is the
'mummy complex' derives from the religious practice of the ancient Egyptians, who, suffering from a psychological
need to resist death (the “victory of time”), believed survival to be possible if the corporeal body could endure.
Forms of keeping the body “in the hold of life” besides mummification were later adopted, such as the
representational model of terracotta statuettes, but proved inadequate (Bazin 2005a: 9). In later centuries, the
question of survival after death ceased to be a function for art, giving way instead to the artistic creation of “an ideal
world in the likeness of the real.” Hence the psychological desire for likeness was usurped by an aesthetic one, and
only the invention of photography and cinema have been able to restore “the ontological identity of model and
image” (Bazin 2005a: 10).
74
art of the present, is designed to resist what Dudley Andrew refers to as the “epiphanic view of cinema” to
which traditionalists adhere: that of holding up the camera as “a unique device that captures the visual
configuration of a given moment, perhaps revealing its truth” (D. Andrew 2010: 5). This critique of
ontological traditionalism is commonly levelled at Bazinian realism with variable levels of severity—the
most devastating being Annette Michelson's preface to Noël Burch's Theory of Film Practice, in which
Michelson deems Bazin's criticism to be “fundamentally antipoetic” and “resolutely antimodernist,”
assuming, perhaps unfairly, that he argued exclusively “for a cinema of 'transparency', for a style grounded in
mimesis” (Michelson 1973: xi, viii). Daney's Cahiers axiom is designed to counter such accusations that
designate realism as involving little more than an objective transcription of a material reality, and instead
signifies a deeply modern recognition of the limits of objectivity that is upheld in the contemporary
durational films under discussion in this thesis. In order to rehabilitate a theory of realism that can adequately
reflect on the durational style of contemporary art and experimental cinema, Daney's axiom might be
rephrased as: the real might not be what is represented, but the cinema can (re)connect us to it.
It is worth noting, then, that Bazin, in his short essay on Roberto Rossellini's Europa ‘51 (1952)
suggested that realism “calls for the most sophisticated stylisation possible,” laying the foundations for a
cinema that is “the very opposite of a realistic one 'drawn from life'.” Instead, Europa ‘51 represents “the
equivalent of austere and terse writing, which is so stripped of ornament that it sometimes verges on the
ascetic.” The future of modern cinema would be composed of “reality through style, and thus a reworking of
the conventions of art,” and it is in this sense that Bazinian realism retains its modernity (Bazin 1997: 139).
Bazin's notion of realism here should be considered in light of Denis Lévy's contention that modern cinema,
despite rupturing classical cinema's 'Realist' system of conditions, has consistently been defined by a regime
in which “the art of realism was contrarily maintained.”28
In a comment that is particularly suggestive when
considered in light of the Bazinian model and the evolution of realism in contemporary cinema, Lévy
concludes that “the moderns are not merely heirs of the classics, they are in fact the only heirs of classical
art” (Lévy 2011).
28
The 'Realism' of classicism to which Lévy refers here is based on the diegetic operation of fiction, rather than the
more material purpose that I am ascribing to it. In this fictional use, as Bill Nichols explains, the operation of
Realism is simply “to make a plausible world seem real.” Hence Realism depicts an imaginary “life as lived and
observed,” and equally functions as “a vantage point from which to view and engage with life” (Nichols 2000: 188,
189).
75
Fig. 19: Sisters of the Gion (1936)
The modern sequence shot
The central stylistic device of modern cinema, for Bazin, is the sequence shot: a combined use of the long
take and depth of visual field. Bazin attributes this combination firstly to Jean Renoir, whose primary
innovation was to uncover “the secret of a film form that would permit everything to be said without
chopping the world up into little fragments, that would reveal the hidden meanings in people and things
without disturbing the unity natural to them” (Bazin 2005a: 38). The collaborations between Welles, Wyler
and Gregg Toland in 1941 (of which stylistic precursors are already clearly visible in Toland's earlier work,
such as the chiaroscuro photography of John Ford's The Long Voyage Home [1940]) extend this dynamic in
particular shots. The use of depth of field at particular points in Citizen Kane means that, for the first time in
classical cinema, Welles' camera is able to take in “with equal sharpness the whole field of vision contained
76
Fig. 20: The Best Years of Our Lives (1946)
simultaneously within the dramatic field.” For Bazin, this means that “it is no longer the editing that selects
what we see,” and instead the spectator “is forced to discern, as in a sort of parallelepiped of reality with the
screen as its cross-section, the dramatic spectrum proper to the scene.” In conjunction with the respect for
duration enabled by the long take, Bazin argues that this strategy (the principle of the sequence shot) restores
to reality “its visible continuity” (Bazin 2005b: 28).29
Bazin's account of the development of the sequence shot (concentrated in two essays: “The
Evolution of the Language of Cinema,” a composite of three artic1es written between 1950 and 1955, and
“The Virtues and Limitations of Montage,” originally published in Cahiers du Cinéma in 1953 and 1957; see
Bazin 2005a: 23-52) omits an important precedent (inevitably, due to the films' pre-war lack of distribution
29
Bordwell, Staiger and Thompson summarise Toland's own description of this realism as “realism of space, because
the eye sees in depth,” and “realism of time, because cuts call the audience’s attention to 'the mechanics of picturemaking'.
In all, realism in the name of both continuity and concealed artifice” (Bordwell et al 1985: 348; also 344-
352).
77
in the West): Kenji Mizoguchi's experiments with long-shot long-take style in the 1930s. Mizoguchi's style
is, I would suggest, even more radical than Welles', most notably in the extended takes and geometric
compositions of Osaka Elegy (1936), Sisters of the Gion (1936) and The Story of the Last Chrysanthemum
(1939). Sisters of the Gion, for instance, contains a deep-focus shot that is remarkably similar to what
Deleuze refers to as the famous “diagonal which joins the back to the front” in the barroom scene of Wyler's
The Best Years of Our Lives (1946), in which the image is divided into three planes as one of the characters
(Fred Derry) makes a phone call at the rear of the visual field, hence drawing attention away from the
character interaction taking place in the image's primary plane (Deleuze 2005b: 104) [Figs. 19-20].30
Mizoguchi (alongside his contemporaries Hiroshi Shimizu and Yasujiro Ozu, to whom I will return in
Chapter Four) developed a nascent modern aesthetic much sooner than North American filmmakers, and his
early staging strategies retain a spatiality and plenitude of duration that is otherwise entirely uncommon in
pre-war narrative cinema.
Mizoguchi, Welles and Wyler's intention of preserving a 'visible continuity' of reality through the
long take is a constant in contemporary slow cinema, and more prevalent as a defining aspect of the sequence
shot than depth of field—which is employed as a means of expanding the spatiality of the image by a few
contemporary filmmakers, most notably Béla Tarr, Lav Diaz, Hou Hsiao-hsien and Tsai Ming-liang. In
Mizoguchi, Welles and Wyler's cinema, however, the temporal continuum of reality is, of course, mediated
through the machinations of dramatic image-sound situations. Although the stylistic experimentation of their
work signifies the emergence of a form that can accommodate a more autonomous temporality of the image,
it remains subject to what Deleuze calls an “excess of 'theatricality'”: that is, the tendency to use all narrative
means to bring an “additional” layer of dramatic 'Reality' to the screen (Deleuze 2005b: 105).
Wyler and Welles' classical schema do contain indicators of strategies that are later adopted by a far
more austere cinema that otherwise bears very little formal, narrational or ideological relation to their work.
For instance, in The Best Years of Our Lives (which has an ASL of 15.8 seconds), Wyler and Toland construct
a dialogue scene between two principal characters, Al and Fred, in an unconventional manner. The scene is
bookended by two long shots that display Fred's entrance and exit in full, as well as two standard shot /
30
For an exemplary analysis of Mizoguchi's use of the long take, see Bordwell 2005: 83-139, in which Bordwell also
adds Japanese woodblock prints and the “aggressive foreground” dynamics of late Soviet montage cinema to Bazin's
incomplete history of deep-focus long-take style.
78
Fig. 21: Evolution of a Filipino Family (1994-2004)
reverse shot framings of each actor. However, the central component of this otherwise strictly classical
sequence is an unconventional static planimetric two-shot of both characters that lasts a minute and fifteen
seconds, disclosing Al's warning to Fred as the dialogue hastens and toughens.31
This basic static set-up is
replicated often in, for instance, Hong Sang-soo's films (in their many scenes of eating, drinking and
conversation, featuring two actors or more), or in individual instances such as a sequence shot in Lav Diaz's
Evolution of a Filipino Family (1994-2004) that maintains an identical composition (in a conversation
between Fernando and Dahila) for eight and a half minutes [Fig. 21].32
Alongside an emphasis on the 'visible
31
Planimetric compositions display a background perpendicular to the lens axis, against which characters stand or sit
in profile. For a summary of this device, which is far more common in modernist film, see Bordwell 2005: 167-168.
32
Also, the next scene in The Best Years of Our Lives (a dialogue scene between Peggy and her mother, Milly) is
staged as a full sequence shot (with a fixed frame) that runs for almost two minutes. The shot's action plays out in a
confined space (a kitchen), in which Peggy sits in the foreground, distractedly shelling peas at the table, before
leaving in a fit of emotion after her mother approaches to sit with her. Here, the action builds to a carefully staged
melodramatic climax, but the sensibility of a style committed to the fixed, extended observation of space and human
activity over time is apparent.
79
continuity' of space and time in the sequence shot, another aspect of the Bazinian model that has implications
for the contemporary cinema lies in Bazin's identification of the potential for a 'democracy' of the image, and
the development of a mode of spectatorial practice unique to modern rather than classical cinema. Bazin
suggests that, in our perceptual relations with the everyday world, “we are free to create our own mise en
scène,” selecting what to focus on by scanning our surroundings at the level of attentiveness that we choose.
When watching a film, however, the classical operation of 'analytical cutting' deprives us of “a freedom that
is well grounded in psychology and that we give up without realising it: the freedom, at least the potential
one, to modify at each instant our method of selection, of 'editing'” (Bazin 1997: 8). The activity of montage,
in Bazin's theory, is not necessarily a task for the artist, but rather for the spectator. Bazin offers many
definitions of artistic realism throughout his criticism, and one is particularly appropriate here: an act “to
force the mind to draw its own conclusions about people and events, instead of manipulating it into accepting
someone else’s interpretation” (Bazin 1997: 124).
When considered in light of this ideological basis, the imposition of analytical cutting entails, for
Bazin, “not only a dramatic, emotional or moral choice, but also, and more significantly, a judgement on
reality itself” (Bazin 1997: 8). What occurs in modern cinema, however, is a reluctance to 'fragment' reality
purely on the level of narrative action (the incitement of drama or overt emotion), as well as an acceptance
by the artist that “the spectator [must] make his own choice” (Bazin 2005a: 92).
Two shots in Welles' The Magnificent Ambersons (1942) are instructive in this respect. The first, a
mobile long take in which George and Lucy stroll past shopfronts on the town's boardwalk, is staged with
sufficient distance between camera and action to allow us to observe the physical shifts in their relations
during conversation. As the shot unfolds we notice that slight aspects of Lucy's behaviour appear to betray
her dialogue: an indirect revelation of character that is not asserted or signalled directly by editing, but
simply plays out for the spectator to locate within a clear, unbroken spatial and temporal field. Victor Perkins
notes that the restraint of this scene demonstrates “the subtlety that can be attained when the camera is held
back from any display of psychological awareness and discrimination,” displaying a refusal of “the easy
rhetoric of emotional and psychological exposure that analytical editing makes available” (Perkins 1999: 58-
59). During this otherwise quite ordinary long take, we are free to turn our attention to evidence of deed and
gesture, and, true to the utility of classical style, in narrative terms it pays off.
80
Fig. 22: The Magnificent Ambersons (1942)
The second shot from The Magnificent Ambersons (1942) is one about which Bazin writes at length
in his monograph on Welles: the lengthy kitchen scene consisting of a dialogue between Fanny, George and
Jack (see Bazin 1991: 74-81) [Fig. 22]. Throughout the main body of the shot, which lasts four and a half
minutes, the camera remains in a fixed position, and the dramatic structure of the scene slowly builds to
Fanny's emotional outburst, which is followed by the men's inadequate reaction and a reflective conclusion.
The shot's staging is, in effect, undramatic: the camera is uncommonly still, the lighting low in depth, and the
dialogue is delivered deliberately, accompanied by the rhythm of persistent rainfall. The frame is cluttered
with inanimate objects, leading Bazin to note that “during the whole scene, objects, outrageously irrelevant
to the action yet monstrously present (cakes, foodstuffs, kitchenware, a coffee-pot), solicit our attention
without a single camera movement conspiring to diminish their presence” (Bazin 1991: 80). Before the
drama unfolds, our gaze is free to scan the frame and examine these objects as things-in-themselves, and it is
81
perhaps worth noting that Welles' fixed staging would seem as unusual today in the context of commercial
cinema as in 1942. Since the 1980s, contiguous with the gradual emergence of slow cinema, Hollywood and
mass-market film has jettisoned the “unbudging long take” that formed part of its stylistic inventory (as
deployed by Wyler in the example above) in the immediate post-war era. Hence, as David Bordwell
suggests, “today perhaps the most radical thing you can do in Hollywood is put your camera on a tripod, set
it a fair distance from the action, and let the whole scene play out” (Bordwell 2005: 29-30). When adopted
by Welles, however, this basic strategy works to validate Bazin's assertion that, under the sustained gaze of
the long take, we are permitted to perceive “the ontological ambivalence of reality directly, in the very
structure of its appearances” (Bazin 1991: 80).
What Welles achieves in this shot is a critical step toward what Lévy refers to as the 'emancipation of
the gaze' in modern cinema. Before the shot's dramatic structure fully reveals itself, there is no immediate
“seduction or manipulation” by camera movement or editing, and a muted abstention “from effects that
dominate the viewer’s gaze.” For a brief period, the shot allows “the full enjoyment” of one’s spectatorial
freedom (Lévy 2011). This effect is only temporary, however, as the dramatic locus of the action gradually
overwhelms other components of the shot, charging the scene until it “produces a spark that all the action has
been directed towards” (Bazin 1991: 68).
The method by which Welles carefully composes this fixed field with reduced attendant movement
finds an unlikely correlate in Chantal Akerman's Jeanne Dielman, 23 quai du Commerce, 1080 Bruxelles
(1975). Welles' strategic staging plays out, in retrospect, like a primitive classical version of the hyperrealist
long takes in Akerman's film, which are designed to accentuate the psychological instability of Jeanne's
character as the narrative itemises her everyday activities. As Ivone Margulies has noted, after the central
'event' of the burning of potatoes in Jeanne Dielman's narrative, “the extended duration of the shots no
longer points to a nondramatic action but starts to function as the locus par excellence of the drama's
eruption” (Margulies 1996: 77). The stillness of the camera's extended gaze sets up a spatial and temporal
field in which tension slowly, inevitably, builds, eventually reaching a point of psychological crisis.
In recent slow films, the employment of the realist sequence shot is simpler. Whilst conforming in
principle to the tendency to construct a scene as “a complete unit in time and space,” recent durational
cinema's modification of early modern form lies in the total ejection of dramatic purpose from the shot
82
(Bazin 1991: 68). The stylistic devices employed in the contemporary cinema remain not dissimilar—a fixed
frame, long take and undramatic staging—but the liberation from dramatic narrativity tends to invert the
function of the shot. In slow cinema, 'insignificant' details, objects and events become the image's operative
subjects, and a 'surplus of theatricality' is replaced by a gradual amassing of duration. Hence, in the more
recent cinema, the classical container of the Bazinian sequence shot is, in total form, de-dramatised.
The elimination of editing
Roberto Rossellini, when interviewed by Bazin in 1958, is quite clear about one direction sought by modern
cinema in the post-war period:
I don’t know whether, today, montage is so essential. I believe we should begin to look at the cinema
in a new way, and to start with abandoning all the old myths. The cinema at first was a technical
discovery; and everything, even editing, was subordinated to that. [...] From the silent cinema we
have inherited this myth of montage, though it has lost most of its meaning. Consequently, it is in the
images themselves that the creative artist can really bring his own observation to bear, his own moral
view, his particular vision (Bazin 1958: 27).
Bazin suggested that this vision was embodied by Rossellini's early post-war films (between Rome, Open
City [1945] and Stromboli, terra di Dio [1950]), in which successions of 'image-facts' take on “a moral the
mind cannot fail to grasp since it was drawn from reality itself” (Bazin 2005b: 36). In the modern era,
cinema makes way for this vision by abandoning its old myths, stripping away “all expressionism” and
embracing “the total absence of the effects of montage” (Bazin 2005a: 37).
Bazin’s insistence on the superiority of analysing the dramatic field in time (“a positive action the
results of which are far superior to anything that could be achieved by the classical cut”) once led him to
suggest that “the average number of shots in a film diminishes as a function of their realism” (Bazin 2005a:
34; 1997: 10). This notion was denounced by one of his fiercest critics, Gérard Gozlan—writing for Positif at
the height of that journal’s territorial tussle with Cahiers du Cinéma—by taking it to its logical extreme:
83
“Reality has more than one level; it is not ambiguous, but complex and deceptive; and any attempt in art to
explain or criticise it demands a search, an analysis, and an organisation. Bazin does admit the need for
organisation, but only as a pis-aller, for in his opinion the ideal film would consist of only one shot” (Gozlan
1968: 52).
After the introduction of Cinemascope in 1953, Bazin restated his position: “It is not true that cutting
into shots and augmenting those shots with a whole range of optical effects are the necessary and
fundamental elements of filmic expression, however subtle that expression might be. On the contrary, one
can see that the evolution of film in the last fifteen years has tended towards the elimination of editing”
(Bazin 1997: 90). The teleological implication in Bazin's work is that the evolution of “dramatic realism”
would produce a cinema that regarded the analytical operation of montage as increasingly inessential, yet
only the emergence of slow cinema's pared-down aesthetic, in which the operative element of the diegesis is
the image and its duration, accords something prophetic to Bazin's predictions (Bazin 1991: 80).33
It is perhaps telling, then, that many contemporary slow filmmakers speak of practice, editing and
spectatorship in a manner deeply reminiscent of Bazin's criticism. When formal issues are addressed by
filmmakers in interviews or articles (Jia Zhang-ke, James Benning and Raya Martin, for instance, are
particularly accomplished essayists), technique is frequently discussed with an indirect bias that echoes both
Bazin's and (in a slightly lesser manner) Siegfried Kracauer's realist theory.
Reflections on the long take by contemporary filmmakers
The reflections of two filmmakers who actually studied Bazin are especially relevant here: Jia Zhang-ke and
Lav Diaz. Jia studied Bazin at Beijing Film Academy, and has spoken of how Bazin's work, alongside other
texts (notably, Andrei Tarkovsky's Sculpting in Time [1987], and collections of interviews with Hou Hsiaohsien
and other filmmakers), directly shaped the realist direction of his work: “naturally there are some
directors who neglect reality in their work, but my aesthetic taste and goals don't allow me to do that—I can
never escape reality” (M. Berry 2005: 186, 193). Diaz, mostly self-educated and extremely well-read, has
33
In line with the emergence of the current 'mature' stage of slow cinema, Peter Matthews notes that “on his death an
obituary notice in Esprit cited Bazin as predicting that: 'The year 2000 will salute the advent of a cinema free of the
artificialities of montage, renouncing the role of an “art of reality” so that it may climb to its final level on which it
will become once and for all “reality made art”'” (Matthews 1999: 25).
84
referred to Bazin as one of his “great heroes”: “He gave me the kind of freedom to seek the truth, to see that
it’s all about aesthetic fulfilment. It’s not about anything else: you just have to work on the domain of
aesthetics, of truth, of beauty” (Ruiter 2012).
It is worth noting, firstly, that the observations of contemporary slow filmmakers tend to assume
particular importance in the critical appropriation of their work (as, more often than not, other forms of
coverage or criticism are regrettably sparse or thinly distributed). David Bordwell has commented that this
trend is common in modern cinema, and interviews with contemporary filmmakers are usually as generous
and self-reflective as those with earlier modernists such as Michelangelo Antonioni and Andrei Tarkovsky
(see Antonioni 2007 and Tarkovsky 1987).34
Theo Angelopoulos, for instance, was frequently one of the most
astute commentators on his own style—a status that he sought, as Bordwell notes, as a result of the need to
establish a reputation on an expanding festival circuit, secure international funding for production and
distribution, and aid spectators in the interpretation of a potentially unfamiliar high modernist style (see
Angelopoulos 2001; Bordwell 2005: 144-145). Contemporary filmmakers that adopt unconventional or
rarefied aesthetics continue to face this set of problems as a consequence of their marginality, and to a
considerable extent continue to shape critical discourse around their work.
When discussing Platform (2000) in an interview with Stephen Teo in 2001, Jia reiterates the
Bazinian prescription of permitting the spectator to make their own choice:
I use a lot of long shots. If the audience can see things in there, that’s good, if they can’t, so be it. I
don’t want to impose too many things onto the audience. [...] I want to give them a mood and within
that mood, you can see things that you want, or you can’t see things. My films are not very clearly
stated to the extent where the audience can see clearly the objects they want to see, this pen or this
watch. If they don’t notice it, they don’t notice it (Teo 2001).
This ambiguous, restrained style is, for Jia, reflective of a subjective striving for the real, rather than an
imagined objectivity: “I am imparting a director’s attitude, how he sees the world and the cinema. What I
34
Bordwell also points out that this is far from new: apparently, a close inspection of papers and journals of the time
reveals that Bazin's insights into Welles, Wyler and Toland's innovations in depth of field were deeply influenced by
the stylistic observations made by Toland and his directors in studio publicity materials (see Bordwell 2007b).
85
mean to say is that it’s only an attitude because you can never be absolutely objective. [...] There is no
absolute objectivity, there is attitude, and through this attitude, there is an ideal” (Teo 2001).
The realism of Jia's work, then, arises from the artist's necessary response to the world: a world, in
Platform, that is perpetually in transition—post-socialist China between 1979 and 1989. Many interviews
with contemporary slow filmmakers echo Jia's emphasis on the subjectivity of imparting an aesthetic
“attitude.” Tsai Ming-liang, for instance, has described his own adoption of a long-shot long-take style as the
only method that, necessarily, enables the “duration required by an author in his expression.” Tsai states that:
“The purpose of the long takes is not to force the audience, but rather stems from my own need. For me, the
duration of the take is based on the need of the creator, and it is up to the audience to respect it” (Lim and
Hee 2011: 186).35
The sequence shot becomes a device for the artist to observe and mediate the world, as
well as an attempt to permit the spectator to recreate that experience.
The film that Jia discusses here, Platform, is constructed in a long-shot long-take style that preserves
separate events and interactions as individual blocks of time (in shots that, on the whole, tend to run for over
a minute each). Jia places the camera at a substantial remove from his actors, and employs conscientious
staging to organise the field of action in place of analytical editing or the use of off-screen space. Long shots
are staged with a depth that is attentive to both architecture (the empty halls, brickwork, and propaganda
hoardings crumbling after Deng Xiaoping's first wave of market reforms) and group interactions of the
characters (that is: time-wasting, singing, dancing, and, tellingly, distractedly watching films). Unlike the
earlier, grittier Xiao Wu and Unknown Pleasures, Platform was shot on 35mm (rather than 16mm or DV),
and its expansive, nostalgic grandeur is reflective of its format.
Jia's Xiao Wu, Unknown Pleasures and Still Life (to which I will return in Chapter Four) are perhaps
best considered as instances of a contemporary neorealist cinema, and feature a few basic principles that
Bazin applied to that immediate post-war cinema: accidental and episodic narratives that are performed by
non-professional actors, and shot on location with portable or handheld cameras (here, in streets, bars, stores
and cramped domestic interiors) (see Bazin 2005b: 93-101). The portrait of a lone pickpocket in Xiao Wu
was filmed on the hoof in only twenty-one days, and represents, in its visual and social rawness, what Thom
Andersen has referred to as the “necessity” of contemporary neorealism: the manifestation of a “cinema of
35
For a detailed study of authorship in Tsai's work, and his position as a 'paratextual' auteur, see Lim 2007.
86
despair” that endeavours, and fails, to posit “the possibility of human agency in a time when we have lost
faith in that possibility” (Andersen 2009: 27).
After Platform and The World (2004), Jia's more recent films explicitly complicate the realist
boundary between fiction and documentary, taking as their subject an intersection between craft
(performance, painting, textile design), China's rapid development of a market economy, and the country's
resultant changing rural and urban environment (centred on Chengdu in 24 City [2008], and Shanghai in I
Wish I Knew [2010]). Jia designs image-sound situations in which portraits of people (from bourgeois
fashion designers to rural miners and peasants) reflect on fragments of history, the politics of representation,
and the inexorable activity of the market over the last three decades (that is, the mass production and export
of consumer goods, a struggle between artisanal and factory production, and an increased precariousness of
labour in both city and country).
Jia's realist “attitude” depicts changes in post-socialist China through the durational observation of
minor or undramatic human activities within casual, expansive or experimental (hybrid) narratives. In a
strategy that has remained broadly consistent throughout his work, Jia opts to stage these activities in
extended sequence shots, and reveals that
what I like most in a long take is that it preserves real time, it keeps time intact. [...] If I were to cut
the scene into pieces, there would be a lot more subjective things that I put in. In Platform, the
characters have a relationship with time. You see two people smoking and talking aimlessly for a
long time. Nothing happens plot-wise but at the same time, time itself is kept intact. In that long and
tedious passage of time, nothing significant happens, they are waiting. Only through time can you
convey this. If I were to break up that scene which lasts for six or seven minutes into several cuts,
then you lose that sense of deadlock. [...] Everybody experiences the monotony of time passing
where nothing that is noteworthy occurs (Teo 2001).
Here, two contingencies of form in contemporary slow cinema are plainly evident: the combination of a void
in narrativity (a suspension of linear action for the purpose of imparting a sense of tediousness and
aimlessness), and the extended take necessary to maintain an unbroken duration of 'nothing noteworthy'. The
87
latter strategy, above all, constitutes a modification of the Bazinian model: an analysis of the undramatic
field in time, intended to maintain the “sense of deadlock” that arises from a dislocation of the everyday, and
the characters' own experience of time within the diegesis.
Jia's commitment to conveying the passage of time of undramatic events recalls Bazin's description
of technique in Visconti's La Terra Trema (see Bazin 2005b: 42-43). In both Visconti and Jia's films, the
principal function of the image is to contain both an action and a meaning that it expresses fully, regardless
of its 'noteworthiness'. With regard to the duration of shots, Bazin observes that Visconti “seems to have
wanted, in some systematic sense, to base the construction of his image on the event itself. If a fisherman
rolls a cigarette, he spares us nothing: we see the whole operation; it will not be reduced to its dramatic or
symbolic meaning, as is usual with montage” (Bazin 2005b: 43). Here, an action as everyday, and banal, as
the rolling of tobacco must be accorded its fullest expression: only in this sense, as in Visconti's later work,
does La Terra Trema display an inclination toward 'the epic'.36
Such actions in Visconti's film are rendered in
what Bazin refers to as a “quasi-documentary” style, shot on location with a variable depth of focus and
consistency of chromatic scale.
Although there is a stateliness and rigour to the static set-ups and slow pans across land- and
seascape in Visconti's film that points to its later modern descendants, Bazin lamented that “the aesthetic of
La Terra Trema must be applicable to dramatic ends if it is to be of service in the evolution of cinema”
(Bazin 2005b: 45). This has often been taken to indicate that the Bazinian model is incompatible with
undramatic cinema (for instance, de Luca is rightfully sceptical: see 2011: 23-24). However, I would argue
that Bazin's concern about the “austere 'entertainment'” of La Terra Trema is not so much a judgement on the
aesthetic itself, but rather its potential marginality to the “need” of a public that cinema, in Bazin's view, is
tasked to embody and represent.
For Bazin, the source of cinema's moral vision (embodied in the post-war period by Visconti's
“concern with things themselves, as in life,” or Vittorio de Sica's humanism) derives from its relation to a
public that has been “devoured by the mechanised aspects of work, and standardised by political and social
restraints” (Bazin 2005b: 41; 1981: 36). In a striking essay written during the Second World War, “For a
36
Interestingly, there is no shot in La Terra Trema that foregrounds the action of rolling a cigarette that Bazin
describes. Bazin's essay contains further inaccuracies (such as the suggestion that some shots last three or four
minutes), so it is perhaps best to consider his description as representative of an everyday event rather than strictly
actual (Bazin 2005b: 43).
88
Realist Aesthetic” (1943), Bazin argued that “cinema, like any naissant art, must be analysed in its concrete
complexity, in the totality of its relations with the social milieu apart from which it would not exist.” In line
with his privileging of psychological subjectivity in relation to filmic ontology, Bazin argues that “the
function of cinema, above even its artistic function, is to satisfy the immutable collective psychic needs that
have been repressed.” The implication of this public function, more succinctly, is that “the cinema aesthetic
will be social, or the cinema will have to do without an aesthetic” (Bazin 1981: 37).
Bazin's reflections here evoke, perhaps consciously, Walter Benjamin's contention that photography
and cinema's capacity for “historical witness” signified “the first truly revolutionary means of reproduction”
that could accelerate the “orientation of reality toward the masses” (Benjamin 2009: 236, 235). In his
seminal essay “The Work of Art in the Age of Mechanical Reproduction” (originally published in 1936),
Benjamin stated that, throughout Western Europe, “capitalist exploitation of film bars modern man’s
legitimate claim to be reproduced from being considered,” and that, “given such circumstances, the film
industry has every interest in arousing the participation of the masses by the means of illusory presentations
and suggestive speculations” (Benjamin 2009: 247). In Benjamin's contention, the task of the filmic medium
would be to further “a revolutionary critique of traditional notions of art,” but for Bazin, cinema's popular
function would be simpler: an acceptance of a certain degree of “illusory presentation,” as long as a sense of
“the genuineness of a thing [and] its material duration” is preserved (Benjamin 2009: 245, 233).
In an essay entitled “Défense de l'avant-garde” (published in L'Écran français in 1948), Bazin also
suggested that “being destined to appeal to a very wide audience is both the weighty burden and the unique
opportunity of cinema. [...] Cinema, by its very nature, is destined to serve the masses all over the world.”
Hence “any attempt to create an aesthetic from the perspective of a narrow audience is, in the first place,
historically inexact and doomed to fail: it leads up the garden path” (qtd. in Bickerton 2009: xv). Bazin
assumed that the “filmic monster” of ascetic style signified by La Terra Trema would remain “inaccessible to
the general public,” and thus fail to occupy the position of a truly social art (Bazin 2005b: 45). Regardless of
the validity of these concerns (which, I would be inclined to suggest, are misplaced, and as conservative as
Annette Michelson's critique, cited above, would suggest), the basis of Bazin's rejection of a nascent
undramaticness is at least consistent with his prevailing vision of cinema's social function.
I will return to Jia's Bazinian tendencies in Platform shortly, after considering Lav Diaz's reflections
89
on the long take and an example of an undramatic sequence shot from his work. Diaz's style, which I shall
return to in more detail in Chapter Four, is one of the most radical in contemporary narrative film. Diaz
repeatedly ascribes his use of the long take to an attempt to convey the “real time” experience of “the
struggle of the Filipino people” (or what he also refers to as “the Filipino psyche”), and in an interview with
Brandon Wee (conducted after the completion of the final version of his first digital work Evolution of a
Filipino Family [1994-2004]), Diaz summarises his technique:
I avoid close-ups when treating the characters I create in my films. I prefer long and oftentimes static
takes, just like stasis—long, long takes in real time. My philosophy is [that] I do not want to
manipulate the audience’s emotions. I want them to experience [Evolution of a Filipino Family] the
way they would experience their normal lives (Wee 2005).37
Diaz comments that if and when his characters walk, work, eat, converse or display emotions (such as
boredom, sadness or hysteria), those experiences must be conveyed to their fullest duration—an approach
that leads Diaz to speculate on forms that might be necessary to accentuate the actuality of duration: “How I
wish I could do a fifteen-hour film about walking characters.” In Heremias Book I: The Legend of the Lizard
Princess (2006), Diaz actually comes close to achieving this ideal by producing a nine-hour film that depicts
the solitary travels on foot, and by ox and cart, of a rural salesman for vast swathes of its full duration.
Equally, long walks (necessitated by either exodus or mania) take up entire sections of Evolution of a
Filipino Family and Century of Birthing (2011), and can last up to half an hour at a time. The dilatory
structure of Diaz's films are also marked by their total abstention from dominant cinema's reliance on
montage or non-diegetic music to supplement situations that are either undramatic or melodramatic, which
he refers to as “Hollywood’s paradigm of drama, where resonance is manipulated and pathos is contrived.
[…] I want the audience to see the truth and to discover their truths by experiencing the realities that I am
presenting or re-presenting” (Wee 2005).
37
The most comprehensive interview with Diaz yet published remains Alexis Tioseco's for Criticine 2, conducted in
2006 and divided into several sections at http://www.criticine.com/interview_article.php?id=21. See also Wong
2005.
90
The Bazinian sequence shot in contemporary cinema
At the risk of relying on an excessively schematic example, I shall now turn to a shot in Diaz's work that
represents an actualisation of the observational principle described by Bazin and Jia above. Jia's model of a
plotless scene in which characters smoke and talk aimlessly “but time itself is kept intact” whilst “nothing
significant happens” is epitomised by the thirteenth shot of Death in the Land of Encantos (2007) (Teo
2001). This routine sequence shot occurs towards the end of the first hour of Diaz's film—which, in total,
runs for nine hours and five minutes.
Death in the Land of Encantos is a 'hybrid' work that integrates both fiction and documentary
material (a form described in my first chapter). Most of the film was shot in Padang, near Legazpi City in the
Bicol region of the Philippines, in the aftermath of Typhoon Reming (known internationally as Typhoon
Durian). Typhoon Reming struck in the last few days of November 2006, and triggered lahars from the
slopes of Mount Mayon (a volcano that looms over Padang), burying homes and drowning families in the
town and surrounding area. Diaz stages a fictional narrative in the ravaged landscape of the region, and also
films a series of interviews with residents and survivors in the same location during the same period. These
two modes intersect throughout the final film, and are accompanied by fixed long takes and subjective point
of view shots (filmed with a handheld digital camera) that record the devastation to Padang for protracted
durations.
In the narrative component of Encantos, Benjamin Agusan, a poet, returns to Padang from Kaluga (a
town near Moscow, Russia) to confront his past. Benjamin wanders the landscape and passes time with
friends (Teodoro, a fisherman, and Catalina, a sculptor), before slowly descending into insanity and paranoia
—a condition that, by the end, is revealed to be induced by political surveillance of his everyday activities.
The thirteenth shot of Encantos is preceded by two extended takes. Both are extreme long shots
(what might be called 'cosmic' long shots, as I will detail in Chapter Four), in which Benjamin and Teodoro
traverse a deserted landscape on foot. The first shot lasts a minute and twenty seconds, and depicts Benjamin
and Teodoro's complete walk from a distant point of the visual field toward the camera. The next shot, a
reverse angle, shows them walking away from the camera and toward the sea (which takes another minute).
Both are archetypal sequence shots, in the sense that they present an action in its entirety: a walk from one
91
Fig. 23: Death in the Land of Encantos (2007)
visible point to another, in which a cut occurs only when the spatial and temporal representation of the
journey is complete. Extended takes of undramatic actions such as this are a cornerstone of Diaz's technique,
and Encantos has a total ASL of 196.1 seconds.
The next shot lasts almost six and a half minutes, and provides a literal actualisation of Bazin's
dictum regarding the fisherman's action in La Terra Trema: to base the construction of the image on an
undramatic event (the rolling of a cigarette) in which “we see the whole operation; it will not be reduced to
its dramatic or symbolic meaning, as is usual with montage” (Bazin 2005b: 43) [Fig. 23]. After Benjamin
and Teodoro arrive at the shore, Teodoro sits and leans back against a rock, lighting a cigarette as soon as the
shot begins. Benjamin slowly makes his way across the shredded branches and detritus that litter the beach,
and takes just over a minute to reach Teodoro. A plastic bag flutters at Teodoro's side, and the wind buffets
the camera's unprotected microphone—the only other diegetic sound that is heard throughout the shot (Diaz
is habitually unfussy about removing technological imperfections that point to the real). Benjamin then sits
near to Teodoro, who continues to smoke as they talk of memories, friends and family: he asks whether
Catalina's character has changed in his absence, and Teodoro tells of his parents and relatives who were
92
buried alive by the lahars that followed the typhoon. Teodoro talks, gesticulates and laughs (in a subdued
manner) more than he drags on the cigarette, and dismissively flicks ash into the sand. After five minutes
have passed, Teodoro stubs out his cigarette, and the two men pause, silently, to continue to look out to sea
without fixing their gaze on anything in particular. Teodoro then gets up, covers his head with a knit cap, and
leaves the frame after another minute. Benjamin watches him go, and Diaz cuts about twenty seconds later to
an extreme long shot of Mount Mayon, its peak shrouded in mist.
The Bazinian conception of the de-dramatised sequence shot common to slow narrative cinema is
ably illustrated by this example. Diaz subtly structures the activity of the image by privileging a minor,
habitual procedure: the smoking of a cigarette, which must be performed in full (an activity that tends to be
completed in everyday life, but rarely when subjected to the compression of classical narration).38
The shot is
activated when Teodoro lights the cigarette, and its end is signalled when he finishes, not least because it
provides a motive for Teodoro to break off the hesitant conversation. Throughout the shot, Benjamin and
Teodoro's poses are, respectively, studied and placid (despite Teodoro's mild gesticulations), and their
interactions are subject to contemplative lapses into silence. Both men continuously gaze into the distance,
and take their time when they speak: a patience that Diaz's minimal application of style does not attempt to
hasten or interrupt.
David Bordwell, in his analysis of Theo Angelopoulos' modernist style in Figures Traced in Light,
addresses the dynamics of bodily posture in undramatic long takes such as this. In the context of
Angelopoulos' work, Bordwell identifies a particularly prevalent “three-quarter dorsality” posture that is
adopted when figures turn away from the camera's gaze, minimising the spectator's access to their reaction to
the world (Bordwell 2005: 162). This opaque bodily staging occurs frequently in the cinema of Michelangelo
Antonioni, Theo Angelopoulos and Béla Tarr, and constitutes a prime signifier of modernist technique. Diaz
adopts a similar strategy at other points in his work (although not often), and, in this shot, opts for a more
frontal pose that nevertheless achieves a similar effect to that which Bordwell describes.
In the thirteenth shot of Encantos, Benjamin and Teodoro's postures are coordinated to mark out their
presence in the landscape, but their positioning also tends to coax the spectator's attention back to the
38
A more literal instance of this representational content would be Henry Geldzahler's smoking of an entire cigar in
Warhol's Henry Geldzahler (1964), a film that I described briefly in my previous chapter.
93
Fig. 24: Platform (2000)
surrounding visual field and the off-screen projection of their gaze. As Bordwell suggests, in these types of
shots “we are obliged to study the body or, more accurately, study the body's relation to the larger field into
which it's inserted,” through staging strategies that “encourage us to pass from figure to environment”
(Bordwell 2005: 164). The complete representation of an everyday action provides an unobtrusive container
for this observational strategy, which is signalled by Bazin's conception, rather than his positive validation,
of the function of the undramatic sequence shot.
Having used this routine sequence shot from Diaz's work as a representative example of a broader
trend, I will now turn to a similarly 'ordinary' long take from Jia's Platform. This particular shot takes place
towards the end of the first hour of the film, and lasts for three minutes and forty seconds. Its primary
purpose is to depict a character interaction that is integral to the overall narrative (which details the decadelong
conversion of the Peasant Culture Group of Fenyang into the “All-Star Rock 'n' Breakdance Electronic
Band” through the lives of its members), but over half of the shot's duration is wholly surplus to the linear
development of plot.
The shot begins with a basic planimetric composition, as Yin Ruijuan (Zhao Tao) and Cui Mingliang
(Wang Hongwei, playing a character named in homage to Tsai Ming-liang) are framed against a visibly
94
Fig. 25: Platform (2000)
crumbling wall (which is, typical for a planimetric staging, perpendicular to the camera's position and lens
axis) [Fig. 24]. Ruijuan tells Mingliang, in guarded terms, that she wants to end their relationship, in part due
to pressure from her father. As she talks, Ruijuan fidgets and draws her ponytail across her mouth, whilst
Mingliang looks down toward his broken wrist (in plaster), and at the ramshackle brickwork next to him.
Visibly upset, he leaves after just over a minute has passed, and the camera tracks his movement until he
climbs a few steps to exit the frame.
Ruijuan calls for Mingliang to return (to which he does not reply), and continues wandering
aimlessly for a few yards, preoccupied. She fiddles with her hair, loosening her ponytail and tying it back
again, and eventually rests against a wall. By this point, after three minutes have passed, the camera's gradual
panning motion has revealed the space to be the precise location where Ruijuan and Mingliang were first
seen together in the film, in an awkward initial meeting. Jia lingers on this space, which carries with it both a
subjective and material history, as Ruijuan clasps her hands by her waist and remains immobile until the end
of the shot [Fig. 25]. A thin cloud of dust rises out of the bottom of the frame, and after forty seconds of this
minimal movement and barely visible activity, the shot ends.
Ruijuan’s actions tell us as little about her subjective motivation as the dialogue itself, which is
95
undemonstrative and brief. The extended take is not intended to disclose psychology or accentuate a feeling,
but to spend time with the characters, to wait as they talk and think for a few seconds or a few minutes, in
turn conveying a sense of actuality that grounds the narrative's broader compression of history.
Jia's lingering on Ruijuan's presence after the shot's ostensible narrative purpose has ended is typical
of a technique that has been characterised as the 'post-action lag'. This term, which I have already referred to
without clarifying in detail, originates in Ben Singer's extended analysis of the 'interrogative' form of Chantal
Akerman's Jeanne Dielman, 23 quai du Commerce, 1080 Bruxelles (1975) (see Singer 1989). Singer notes
that, in Jeanne Dielman, Akerman forcibly defamiliarises classical norms of shot transition by employing a
“trope of standardisation” that signals and draws attention to every impending edit. The narration in Jeanne
Dielman achieves this by following a familiar structural pattern: after Jeanne has completed an everyday task
in its entirety (for instance: bathing, folding laundry, making the bed, polishing shoes, grinding coffee,
washing dishes, preparing and eating dinner, and so on), she immediately moves off-screen to another room
or exterior space, and turns off the lights (if necessary) soon after leaving the frame. Singer notes that
in many shots this process is followed by a post-action lag in which the camera lingers on inanimate
objects for about six seconds before the cut occurs. Any semblance of Hollywood's invisibility of
editing disappears in this standardised, repetitive, self-conscious system of prefacing cuts (Singer
1989: 59).
Singer also comments that Akerman's combination of the post-action lag with a literal representation of
mundane tasks (in their entirety) results in the impression of an “excessive duration and emptiness,” in which
the spectator is “induced into feeling time” (Singer 1989: 58).
The device of the post-action lag therefore primarily derives from late modernist cinema. In his book
on Jean-Marie Straub and Danièle Huillet's first four films (which are, for some reason, credited only to
Straub), Richard Roud makes note of a narrational strategy that has been evident in Straub-Huillet's work
from the very beginning: “while [Straub-Huillet] cut out all connecting links, suppressed explanation and
exposition, [they have] nevertheless held on to shots after the characters have left the frame, training the
camera on the 'empty' screen” (Roud 1972: 52). In Straub-Huillet's work, the necessity of a post-action lag
96
arises from their materialist attention to direct sound—that is, the sonic fidelity that I referred to briefly in
my previous chapter, in relation to Chronik der Anna Magdalena Bach (1967).
Roud illustrates this tendency with an example from Straub-Huillet's second film, Nicht Versöhnt
ode Es hilft nur Gewalt wo Gewalt herrscht (1965). At the close of a shot that depicts Heinrich Fähmel and
his secretary leaving an office, the camera lingers on the depopulated frame until an outside door slams and
the characters' footsteps have audibly dissipated. In this post-action lag, sound is amplified by the reduction
of other diegetic elements, isolating a “concrete reality” that “at the least, if not at the most, exists by and for
itself,” as Bazin described the direct sound in the classical en plein air cinema of Marcel Pagnol (Bazin
1997: 54). In later films by Straub-Huillet the post-action lag is subject to a more hyperbolic application of
duration, and the second shot of the third section of Dalla nube alla resistenza (1979) (a dialogue between
Oedipus and Tiresias that is staged on a moving cart pulled by two oxen) is left to run for an excess of just
under four minutes in order to preserve an attractive birdsong that had been recorded by chance (occurring
just before the end of a reel) during filming (see Costa and Gorin 2009).39
Other durational filmmakers use the post-action lag solely for the purpose of providing time for quiet
reflection, and record a surplus period of temps mort after the primary operation of a sequence shot has
passed. Theo Angelopoulos, in an interview with Francesco Casetti from 1977, likens this strategy of
including post-action “dead moments” to an evocation of “musical pauses”: “After the last note, there is a
moment of silence, allowing the viewer to grasp the sense of the entire sequence. Normally, shots are cut
when the action is over, or the last sound is heard. Emptiness, the dead moment, is the impression you have
when there is nothing more to show or to hear” (Angelopoulos 2001: 26).
The post-action lag in the shot in Platform described above embodies this contemplative strategy
particularly well. Jia allows for a period of reflection after the sequence's dialogue is over, when Ruijuan and
Mingliang are left with nothing to say or to hear. After they separate and Ruijuan is left alone, the image then
retreats to a starker condition of immobility, as if opting for a moment of temps mort to delay an uncertain
future.
Slow cinema's tendency to employ the post-action lag in this way (for the purpose of lingering
39
This intriguing anecdote was told by Pedro Costa during a talk with Jean-Pierre Gorin at the Tate Modern in London
on 3rd
October 2009. For a transcription of the talk (in Spanish), see Costa and Gorin 2009. The dialogue between
Oedipus and Tiresias in Dalla nube alla resistenza is adapted from Cesare Pavese's Dialogues with Leuco (1947).
97
reflection or materialist practice) privileges the temporal rather than spatial aspects of the Bazinian sequence
shot. Bazin's spatial rule—that montage must not be adopted if a “simultaneous presence of two or more
factors in the action” is “essential” to the image—tends toward the representation of scenes of conflict (such
as a bullfight, as Bazin describes in his essay “Death Every Afternoon” [1958]) or the 'realistic' depiction of
illusion or fantasy (as Bazin writes of Albert Lamorisse's Le Ballon rouge [1956] in “The Virtues and
Limitations of Montage”), both of which are either de-dramatised or excluded altogether by contemporary
durational filmmakers (see Bazin 2005a: 50; Bazin 2003).
Slow cinema, then, returns to what we might call the Bazinian root of durational style, displaying a
clear acceptance that film adequately registers an impression of the world as it exists in time, and a subjective
striving for a realism of duration. Robert Koehler, in the context of Albert Serra's El cant dels ocells (2008)
(mentioned briefly in my Introduction), describes this Bazinian root as a “faith in the camera’s power to
convey and transform one's sense of time, duration, and position on the earth below the sky.” This faith,
manifested in the spiritual belief of the Magi and their quest in Serra's film, entails a “choice to express real
time” in order to “specifically locate the viewer inside the spiritual labours [of the characters] as they are felt
and experienced” (Koehler 2008b: 55). In light of Jia's reflections on the undramatic long take, and of
Bazin's dictum that the realist image will be based on the duration of the event itself, it might be said that the
sequence shot's expression of 'real time' obliges the spectator to wait and observe events as we would in the
physical world. In slow film, the experience of an event (and its observation) is often conveyed to its fullest
duration, in turn prolonging and sharpening our capacity to apprehend the world, and our patience within it.
98
CHAPTER THREE
UNDRAMATICNESS, TIME, AND THE NEOLIBERAL ECONOMY OF CONTEMPORARY CINEMA
The following chapter comprises a survey of the narrativity and aesthetics of slow cinema in the context of
other modes of contemporary film, and the production of culture in the neoliberal era. The structure of the
chapter is divided into three broad sections: an elaboration of the concept of 'undramaticness' (in terms of
narrativity and time) in slow cinema; a focus on the production of duration in dominant cinema and culture
of the neoliberal era; and, lastly, a speculative attempt to account for the historical emergence of the
aesthetics of slow cinema.
Undramaticness
In terms of narrativity, André Bazin's model of the application of the sequence shot to 'action' in modern
cinema, as discussed in my previous chapter, is not particularly conducive to the contemporary cinema under
discussion here. Writing of Jean Renoir's Boudu sauvé des eaux (1932), Bazin once argued, tellingly, that
“drama is necessary—that is what we go to the movies to see” (Bazin 1992: 32). However, Bazin also noted
that the “dramatic value of a scene” in Renoir's work often serves as a pretext for opening up aspects of the
shot, and gaps in storytelling, for everyday or material revelations. In Boudu sauvé des eaux, Bazin recalls a
scene (in which seemingly “nothing happens”) that is followed by a quintessential cinephiliac moment: “at
the end of the pan, the camera picks up a bit of grass where, in close-up, one can see distinctly the white dust
that the heat and the wind have lifted from the path. One can almost feel it between one's fingers” (Bazin
1992: 85). Here, a slackening of dramatic development can, at least, make way for a testimony to presence.
This notion of a fleeting liberation from causal relations leads to the concept of undramaticness that is
evident in the field of slow cinema, one that stems from the implication (again, rather than the narrational
actuality) of Bazin's analysis of Vittorio De Sica's Umberto D (1952): “The narrative unit is not the episode,
the event, the sudden turn of events, or the character of its protagonists; it is the succession of concrete
instants of life, no one of which can be said to be more important than another, for their ontological equality
99
destroys drama at its very basis” (Bazin 2005b: 81; my emphasis).
Taking Bazin's phrase literally, this notion of the destruction of drama serves as a pretext for three
principal undramatic effects. In slow films, when dramatic narrativity is ejected from image-sound situations,
other aspects of the image tend to come forward to fill, or deepen, the void: vision (in the form of extended
observation), the depiction of everydayness, and the production of duration. In Deleuze's schema of modern
cinema, the post-war regime of the time-image gives rise to this new cinema of vision through a compulsive
performance of looking, enacted by both the artist and the protagonist of fiction. This cinema of the 'seer'
occurs firstly in neorealism's pure optical and sound situations, where characters, no longer themselves, react
to a transformed world by striving to see (in a new way, or for the first time) the reality of their surroundings.
In this cinema, as Deleuze describes, “vision is no longer even a presupposition added to action, a
preliminary which presents itself as a condition; it occupies all the room and takes the place of the action”
(Deleuze 2005b: 124; my emphasis).
I will return to this concept of the cinema of the 'seer' in my consideration of Deleuze's model of
modern cinema in Chapter Four, but it is important to note here that this cinema is marked, on the level of
both character action and the shot, by a depletion of movement—which, in Deleuze's phrase, “tends to zero”
(Deleuze 2005b: 124). Both the artist responding to the world, and the fictional character in art film, find
their vision immobile. This tendency is evident in the interplay of seeing and stillness (or inaction) in many
slow films, particularly in the depiction of characters who do nothing but look (either intently or idly), and
fixed shots that, themselves, stare for as long as they like. This replacement of action with observation is
narrativised by characters such as the Doctor in Tarr's Sátántangó, who compulsively watches and records
the events of the film's narrative from the perspective of his desk; the mute, lost characters that do little but
walk, sit and watch each other (in, of course, depopulated cinematheques) in Tsai Ming-liang's Goodbye,
Dragon Inn (2003) and Alonso's Fantasma (2006); the portraiture of patient children in Sharon Lockhart's
Pine Flat (2006); and the wanderers of Albert Serra's two fiction films, simultaneously confounded and
overwhelmed by the beauty of nature that surrounds them.
The second aspect of undramaticness, everydayness, is manifested in contemporary cinema by the
literal depiction of events familiar to the spectator from quotidian life: walking, washing, labouring,
preparing food, eating, drinking, conversing, lazing, having sex, masturbating, sleeping, and so on. Key films
100
that present (and structure their narratives by) such activities at unusual length include Liu Jiayin's Oxhide
(2005) and Oxhide II (2009), Aleksandr Sokurov's reclusive, almost mystical, documentary A Humble Life
(1997), and Tsai's Vive l'amour (1994) and I Don't Want to Sleep Alone (2006) (as well as stretches of his
intervening work). These films place considerable emphasis on tasks performed by necessity (eating,
drinking, bathing), as well as two flipsides of the everyday: the physical effort of work (craft and skilled
labour in Oxhide and A Humble Life) and its inverse (idleness, purposelessness and displacement in Vive
l'amour).
I am invoking the notion of everydayness here in its broadest, most literal, sense. In Everyday Life
and Cultural Theory, Ben Highmore comments that of the many guises in which 'the everyday' circulates in
critical discourse and Western culture, “one difficulty becomes immediately apparent: [that] 'everyday life'
signifies ambivalently.” The term, in its broadest sense, points in two directions: to the habitual and repetitive
actions, movements and inhabited spaces that make up our day to day life, but also to another “special
quality” that is less quantifiable. Highmore suggests that this second aspect of everyday life is “the everyday
as value and quality,” hence everydayness. The notion of everydayness can extend beyond actuality and
humdrum routine, and signify a state in which “the most travelled journey can become the dead weight of
boredom, the most inhabited space a prison, the most repeated action an oppressive routine” (Highmore
2002: 1). Both these aspects of the everyday are visible in the work of the contemporary filmmakers that I
have mentioned here (for instance, the films of Liu Jiayin embody the former aspect, whilst certain films by
Tsai Ming-liang appear to evoke the latter), and are expressed in a variety of narrational forms that aim to
attend to the everyday simply by making it, in Highmore's term, “noticeable” (Highmore 2002: 23).
Liu Jiayin's films adopt an overwhelmingly literal approach to ordinary aspects of the everyday by
meticulously disclosing isolated instances of commonplace activity. Her first feature, Oxhide, comprises
twenty-three shots (one for each of Liu's years of age at the time of filming) of family life in a cramped
Beijing apartment, filmed in a 2.35:1 ratio that deliberately restricts, rather than widens, the scope of the
image. Liu commits a single sequence shot to each concrete instance of the quotidian, and a repetition of
routine activity forms a dominant figure in the visual itinerary of mealtimes, instances of familial bickering,
bored jokes, and the physical craft of working with leather (for the family business of designing handbags).
Oxhide and Oxhide II were both filmed, after a number of rehearsals, with Liu's parents, and the familiarity
101
Fig. 26: Oxhide II (2009)
and volatility of actual familial relations (what Liu refers to as the “conflict and rawness” of Oxhide)
underpins the seeming 'ordinariness' of domestic scenes (Rist 2009).
Oxhide II, the second part of a proposed trilogy, is quite strictly a structural film, and submits the
preparation, steaming and eating of a meal of dumplings to a formal conceit that foregrounds the shape of the
film alongside the minutiae of the activity itself. The film's structural organisation consists of nine re-framed
shots that rotate by forty-five degrees after each edit, gradually revolving full circle around a thick
workbench during what appears to be the actual duration of the task from beginning to end. Each member of
Liu's family performs a separate task (for instance: Liu chops garlic and chives whilst her father kneads
dough and prepares pork for the filling), and the preparation of the meal takes two hours and thirteen
minutes. The nine sequence shots are divided fairly equally across the activity's full length, and seventy-three
dumplings in total are prepared (in the first two shots), rolled, flattened and filled (in the next five shots),
cooked (in the second to last shot) and eaten (in the final shot [Fig. 26]) (for a comprehensive video analysis
of the film's sequencing and formal organisation, see Bordwell and Lee 2011).40
As in the hybrid form of
Alonso's La libertad and Costa's Colossal Youth, Liu orchestrates a meticulous performance of the everyday
(in which the frame is at times simply a narrow container for hands performing tasks), and notes that Oxhide
was based entirely “on my family’s real experiences: we re-enacted real-life events. The film was born from
40
With regard to the actual continuity of duration, Liu notes that only “three or four of the nine shots” in the film were
captured in a single take. The rest of the shots were taken three or four times. Liu adds: “we tried to shoot everything
in strict continuity. The dumplings that are being boiled in the eighth shot and the ones we are eating in the ninth
shot are exactly the same. And they didn’t taste good at all” (Rist 2009).
102
a desire to preserve memories” (Edwards 2010).
In his valuable book Disclosure of the Everyday: Undramatic Achievement in Narrative Film (2000),
Andrew Klevan defines undramaticness as the mode by which cinema discloses 'ordinariness' in narratives
that “genuinely acknowledge the everyday” rather than set out “to avoid it or transform it” (Klevan 2000:
30). Klevan's definition of undramaticness is somewhat panoptic, and incorporates a set of films that range
from Hollywood classicism (for instance, Leo McCarey's The Awful Truth [1937]) to canonical post-war art
cinema (the cinema of Robert Bresson and Eric Rohmer). Klevan's definition groups films by conformity on
three levels: a refusal not to 'dramatically transform' scenes, but instead disclose their significance “by the
non-energetic arrangement of body, environment and object”; the presentation of information, or provocation
of emotion, not primarily for plot progression (although intricate narrativity is fundamental); and a refusal to
create visual compositions that might be “pleasurable or gripping” in themselves—hence, autonomous of
narrativity (Klevan 2000: 48-49, 209). The task of undramatic cinema, for Klevan, is not tied to a specialised
set of formal concerns, but is rather “to find a visual style and credible metaphorical scenarios to express less
vehement emotions and states of mind, those not broadly associated with events of crisis” (Klevan 2000: 64).
The notion of undramaticness that I am addressing in this thesis, then, extends far beyond Klevan's
distinction. Contemporary slow films share an emphasis on the 'acknowledgement' of the everyday and
refusal of “events of crisis” that Klevan describes, but do not limit their mode of undramaticness to active
narrativity, or heed the conventional classical resistance to exploring the autonomous (aesthetic or
observational) capacity of the image (at the behest of causal relations). If applied to contemporary cinema,
Klevan's approach might best be suited to the de-dramatised narrativity of what David Bordwell (perhaps
problematically) calls 'Asian minimalism', in reference to the Taiwanese New Cinema of the 1980s and more
recent films by, for example, Ann Hui, Hirokazu Kore-eda, Hong Sang-soo and Ho Yuhang that emerged
during and after the 1990s. Bordwell notes that films such as Kore-eda's Mabarosi (1995), Hong's The
Power of Kangwon Province (1998), On the Occasion of Remembering the Turning Gate (2002) and, more
recently, Ho's Rain Dogs (2006), share a similar set of minimalist narrative tendencies by focussing on
relatively few characters and mundane activities:
In place of the earth-shattering conflicts we see in more mainstream entertainments, these films
103
present everyday and intimate human dramas often embedded in routine activities—riding a train or
a bus, walking through a neighbourhood, eating and drinking with friends and family. While the
situations may recall the problems of love and duty we associate with melodrama, the characters tend
not to burst into grand emotional displays. Instead, their feelings tend to be muted or stifled,
repressed rather than expressed (Bordwell 2007: 19-20).
The field of slow cinema does broadly adhere to these principles, but I would argue that many of its key
films' distinctive qualities intentionally extend beyond, or exhaust, both the operation of narrativity and a
suppression of character emotion. In this respect, it might be said that the prevailing indifference to narrative
primacy in contemporary durational film gives the impression that artists are less likely to be concerned
about its operation, or inclusion, than their critics or potential audiences.
Undramatic narrativity, 'dailiness', and contemporary neorealism
Some contemporary filmmakers reject narratives derived from 'events of crisis' by turning to highly
unconventional methods of depicting the everyday in a direct, linear form. In a comment concerning the
focus of his first film, La libertad (2001), Lisandro Alonso describes the starting point for this tendency:
I chose not to show the extraordinary or spectacular accomplishments that occur every two or three
or five years in somebody's life, such as to have a child, a serious accident, or something out of the
blue. One lives half of their life in a routine that no-one pays attention to. I wanted to register these
minimal moments so that when we see them, we'd rethink what we are doing with our own lives
(qtd. in Falicov 2007: 126-127).
In his 'outsider trilogy' of La libertad, Los muertos (2004) and Liverpool (2008), Alonso narrates the aspects
of life “that no-one pays attention to” in three studies of individuals who subsist at the very margins of
Argentina's late capitalist economy. The lone protagonists of Alonso's trilogy are united by their retreat, in
isolation, from urban society to nature: in La libertad, a depopulated forest is Misael Saavedra's home and
104
source of income; in Los muertos, Vargas travels through remote jungle locations by river to visit his
daughter after being released from prison; and in Liverpool, Farrel returns to an isolated community (after
working at sea) before departing, without warning or narrative signification, into a forbidding wilderness.
I shall focus briefly on La libertad here, and its undramatic depiction of a single day in the life of
Misael Saavedra: a woodcutter who performs his actual everyday routine for the film in collaboration with
Alonso. Misael lives and works in the middle of a forest, felling trees and chopping timber in order to make
enough money to afford a loaf of bread and packet of cigarettes after a day's work. In Alonso's film, a few
aspects of Misael's life are fictionalised, but the sequencing of his daily activity is broadly accurate in
relation to its everyday reality (see Anderson 2001). Alonso extensively prepared the structure of La libertad
with Misael before shooting (travelling from Buenos Aires to the country to talk with him every fortnight for
eight months), and the film was shot in nine days for only $70,000, of which Alonso's father lent him
$40,000 to shoot on 35mm rather than video. At the end of the film, Alonso includes a title that simply reads
“Argentina—2001,” reminding the spectator of the film's position in contemporary history: a portrait of
marginal labour filmed at the height of Argentina's socioeconomic crisis of 1999-2002 (that is, a sovereign
debt crisis and deep recession from which the Argentinian economy has not fully recovered a decade later).
La libertad's hybrid aesthetic belongs to Robert Koehler's definition of the contemporary 'cinema of
in-between-ness' described in my first chapter. However, its unique attention to the 'ordinariness' of a man's
life also recalls Cesare Zavattini's concept of 'dailiness' in his reflections on the moral vision of post-war
Italian neorealism. In the post-war period, as Bazin has noted, Zavattini's “dream” was “to make a whole
film out of ninety minutes in the life of a man to whom nothing happens” (Bazin 2005b: 82). Although Bazin
suggests that Zavattini's idealism was fulfilled by “two or three sequences in Umberto D [1952]” that “give
us more than a glimpse of what such a film might be like,” the neorealist cinema of the late 1940s and 1950s
—not least Zavattini's own collaborations with Vittorio De Sica and Luchino Visconti—fell short of this
ideal cinema of 'dailiness' (Bazin 2005b: 82).41
This was due, understandably, to a combination of artistic
compromise and a surreptitious embrace of populism, and Zavattini, after completing Umberto D, lamented
that that particular film was too firmly “set within the framework of traditional narrative. We have not yet
41
Bazin also notes that Ladri di biciclette (1948) is designed to turn “reality into a spectacle,” and suggests that “until
the realisation of Zavattini's dream of filming eighty minutes in the life of a man without a cut,” it will represent
“without a doubt the ultimate expression of neorealism” (Bazin 2005b: 67).
105
come to real neorealism” (Zavattini 1978: 73).42
However, in a similar manner to Bazin's reflections on durational style, it is worth considering
Zavattini's model of neorealism in relation to its belated actualisation in Alonso's cinema—which is quite
literally rooted in “things as they happen day by day in […] their longest and truest duration” (Zavattini
1966: 220). Alonso's films display a rigorous attention to detail and routine that recalls Zavattini's singular
obsession with the minutiae of the everyday in his scripts. Tag Gallagher, in his biography of Roberto
Rossellini, has documented how “days were wasted” in script meetings between Zavattini, De Sica and
Sergio Amidei (for Sciuscià [1947] and Ladri di biciclette [1948]) deliberating details such as the colour and
shade of an apple that a character might bite into, or the precise word (“Unità”) that would be visible in “a
baguette of bread and mortadella wrapped in a newspaper” (Gallagher 1998: 293). As a result, Marcus
Millicent notes that Zavattini, Amidei and De Sica spent six months working on the script for Ladri di
biciclette, but too little of this singular attention to the everyday made it into the final film (Millicent 1986:
57). Alonso, by contrast, wrote a script of only five pages for La libertad, mostly to itemise activities for the
shoot, and instead 'encounters' the everyday (in Zavattini's term) entirely through observational practice.43
La libertad consists of sixty-two shots that last seventy minutes in total. After a bracing blast of
techno during the opening titles (which are displayed in reverse) and a glimpse of the film's final
composition, the first seventeen shots of La libertad detail Misael's morning routine: scouting trees to fell,
collecting logs, squatting to relieve his bowels (in an exemplary sequence shot that lasts a minute and a half),
clearing a tree's roots, chainsawing a felled trunk, and stripping it of dead branches and bark.44
In the
following three shots, Misael eats leftovers of the previous night's meal for lunch, and smokes an entire
cigarette. When Misael retreats to his tent for a nap, Alonso's camera detaches from its subject for the first
time, and, in a series of weightless panning movements, drifts through the forest in nine brief unmotivated
42
In his writings on neorealism, Zavattini recorded his desire to “leave the studio in search of direct contact with
reality” and create “a system of production which will bring with it the freshness of collective awareness.” Zavattini
was aware that this collective neorealist practice would not be achieved without overturning the existing industrial
mode of production, and stated that “the number of films we make also plays an important part. If we make only
three [films a year, rather than a hundred], we will have to submit to the conditions of production as they exist
today” (Zavattini 1978: 70-71). Hence conformity with the existing system, and its aesthetics, was both the likely
and actual outcome.
43
Alonso wrote his longest screenplay for Liverpool: an epic twenty-four pages that outlined the film's few locations
and characters.
44
Alonso has said that the anomalous inclusion of loud techno at the very beginning is intended to evoke the film's
distance from “the city” (Anderson 2001: 36).
106
Fig. 27: La libertad (2001)
takes. After this interlude, Misael gets up, washes, puts on a clean t-shirt, and departs from the dwelling until
the end of the film.
The next sixteen shots detail Misael's excursion to sell timber to a wholesaler (in a difficult deal that
reduces his revenue), as well as making a short phone call and buying bread, cigarettes and petrol. Misael
travels to the wholesaler's yard by hitching a lift, and returns to his tent partly on foot. This part of the
journey is rendered in eight long shots (both fixed set-ups and pans) that are composed to emphasise the
openness of the rural landscape, and the bright, sporadically clouded expanse of sky under which he passes
[Fig. 27]. In the final eight shots of the film, which take place before and after dusk, Misael kills an
armadillo by striking its head with iron tongs, and cuts the throat of the animal to prepare it for cooking (by
scraping its scales, cutting its limbs, and salting the raw flesh). Misael works quickly and methodically
throughout this unflinching sequence, which is reiterated by Alonso at length in Los muertos when Vargas
slaughters and guts a goat by the side of the river.
Misael uses the dead branches that he stripped earlier in the day to light a large fire, and as night falls
a gathering storm is signalled by rolling thunder. Illuminated only by the amber glow of the fire and remote
flashes of lightning, Misael shreds and eats the cooked armadillo whilst looking toward, and seemingly past,
107
Fig. 28: La libertad (2001)
the camera (in a composition that directly recalls the final seven-minute take of Delphine Seyrig's stare
toward the static camera in Chantal Akerman's Jeanne Dielman, 23 quai du Commerce, 1080 Bruxelles
[1975]) [Fig. 28].45
Jens Andermann notes that Misael's furtive looking at the camera here “briskly cuts off
any empathetic affective investment, re-conducting attention to the documentary terrain of the opening
section.” A precarious balance of reality and fiction is formed between the camera, Misael and his look that,
as Andermann suggests, reflects the equally precarious manner in which “the film shows labour and nature to
be always already subjected to exchange value and the market”—an aspect of the everyday that the central
section of the film's narrative makes emphatically noticeable (Andermann 2012: 87-88). Alonso holds this
final composition on Misael for four minutes, only briefly panning down to display the carcass of the
armadillo and reframing slightly in response to Misael's movements. After the shot ends, the sound of distant
rolling thunder is replaced by torrential rainfall, and the image returns to black.
Here, then, we confront a cinema of 'dailiness' in which, as Zavattini wrote, “awareness and contact
must be direct. A hungry man, a downtrodden man, must be shown as he is, with his own first and last
names” (Zavattini 1978: 72). Zavattini's essays on neorealism frequently echo Bazin's emphasis on the
45
This shot from Jeanne Dielman also appears, in isolation, in Akerman's video installation Woman Sitting Down After
Killing (2001).
108
subjective striving for realism in modern cinema (which Zavattini refers to as “its hunger for reality”), and
point to an “ethical impetus” that extends beyond technical or stylistic considerations (Zavattini 1966: 218).
Marcus Millicent observes that neorealism is primarily a moral statement ('una nuova poesia morale') whose
purpose “was to promote a true objectivity—one that would force viewers to abandon the limitations of a
strictly personal perspective and embrace the reality of the 'others', be they persons or things, with all the
ethical responsibility that such a vision entails” (Millicent 1986: 23). Hence Zavattini's concept of 'dailiness'
might not, even if actualised in Alonso's feature-length rendering of how a man lives, be an aesthetic code at
all, but rather an ethical one: the neorealist “dream” signifies not only the coming into being (in film) of a
renewed social awareness, but an extension of solidarity to fellow men and women in the form of images of
the ordinary world.46
Realisms of behaviour and time
Other contemporary narrative films that are not quite as singular or rigorous as La libertad continue to turn
to the observational disclosure of human activity and gestural detail, as well as a portrayal of what Klevan
refers to as “the inscrutability of undemonstrative behaviour” (Klevan 2000: 155). In Vive l'amour, Tsai
Ming-liang enacts this latter tendency in an ordered series of image-sound situations that focusses on three
characters in possession of a key to the same, empty apartment in Taipei (pending sale in a speculative
residential development symptomatic of Taiwan's pre-1997 economic boom). The lives and movements of
these characters—an estate agent (Yang Kuei-mei), street pedlar (Chen Zhao-rong), and columbarium seller
(Lee Kang-sheng)—overlap in a slow, coincidental chain of cause and effect; a distended form of what
Bordwell refers to as the 'network narrative' (of which Béla Tarr's Sátántangó [1994], Gus Van Sant's
Elephant [2003], and Lav Diaz's Evolution of a Filipino Family [1994-2004] and Melancholia [2008] are
also key contemporary equivalents) (Bordwell 2006: 110). Tsai observes the disaffected listlessness and
psychological strain felt by these characters in the neoliberal city, a conurbation that is depicted in his work
46
This is probably, in retrospect, the most productive way of looking at Zavattini's own collaborative work. Tag
Gallagher is scathing about the schematic nature of Zavattini and De Sica's collaborations in comparison to the
modernity of Rossellini: “Zavattini’s actual scripts [...] are so melodramatically edited and pointed that they often
resemble lachrymose comic strips in slow motion. Unlike in Rossellini’s movies [...] mood comes first in every De
Sica-Zavattini film, before people, who are the puppets of that mood” (Gallagher 1998: 298).
109
Fig. 29: Vive l'amour (1994)
as a 'symbiotic' space that combines the popular and the marginal, the conspicuous and the invisible. Emilie
Yueh-yu Yeh and Darrell Davis have noted that, in Vive l'amour, “instead of panoramic establishing views of
[Taipei], Tsai shows anonymous cubical spaces where people spend most of their time, boxed in,” such as the
empty, cold apartment where the characters of Vive l'amour meet and exchange sex, and a visual urban
geography of strip-lit malls, neon arcades, fast food restaurants, dank hotels, and other (non-)places of
transition, labour and leisure (Yeh and Davis 2005: 224).47
Within this spatial network of impersonal or concealed locations, Tsai's camera gazes at bodies and
their actions from a distance, setting up a series of revelatory portraits of modern alienation and its
expression in private behaviour. In one particularly odd sequence in Vive l'amour, Lee Kang-sheng, assuming
that he is alone in the unoccupied apartment, kisses a watermelon (not timidly), carves three holes into it, and
rolls it into a wall as if bowling a strike. As the melon splits open on impact, Lee crouches on the floor and
slurps at the fruit's flesh and seeds, rubbing a moist shard of the shell across his cheeks and face [Figs. 29-
30]. This activity is swiftly interrupted by Chen Zhao-rong's presence in the apartment, but until that point its
depiction comprises a rare observation of the strangeness of personal behaviour: a testimony to the folly of
47
For an exemplary analysis of Tsai's style and its representation of space in the neoliberal city, see de Luca 2011: 117-
203.
110
Fig. 30: Vive l'amour (1994)
solitude, and an emancipation born of aloneness. As Kent Jones comments, this scene depicts “a kind of
private activity that's common in American cinema, usually [as] part of a moment of self-liberation in a
musical montage. But with Tsai, such a moment loses its free-floating vagueness. Idleness and boredom are
so thoroughly delineated in [Vive l'amour] that you get a glimpse of something rare in cinema: an incomplete
ego in action” (Jones 2003: 49). Tsai modifies the dominant representation of idle activity by constructing
Vive l'amour's narrative around the very insignificance of its characters' activities, and their risk of boredom:
such idle moments are elevated to a structural level within the film's content, rather than a complement to (or
pause within) a more active, emphatic dramatic structure.
In departing from Klevan's model of undramaticness, recent durational film embodies a tendency
that is better expressed by his distinction of the 'melodrama of time': that is, a cinema “where time is not
subservient to the dictates of action, but becomes a subject in itself.” Contemporary slow films that
habitually remove dramatic convention and causal relations from the shot opt to produce duration in its
place, and transform extended takes into 'limit situations' where “the visual patterning and composition are
forceful, even though the plotting is not” (Klevan 2000: 45). Hence the concrete instants of everydayness
that Bazin identified as central to neorealism (and which are visible in the structure of Alonso's La libertad)
111
become concrete instances of duration: either of actual time (a linear temporal record), or the assertive
impression of its passage (a hyperrealism of duration). As Denis Lévy notes, modern cinema's liberation of
duration from the regulated constraint of dramatic standards means that the structure of modern films can be
built from “successive blocks torn from actual time (the time of the take) in a cumulative logic [where] time
is not linked together, it is amassed” (Lévy 2011; my emphasis).
Ivone Margulies, in her book on the hyperrealist everydayness of Chantal Akerman's work (entitled
Nothing Happens [1996]), makes an important distinction between this notion of a realist and hyperrealist
production of duration: whereas the former tends to serve an analysis of the undramatic field in time, the
latter points to a “peculiar sandwiching of literalness and representation” that, additionally, produces a
surplus of the real (Margulies 1996: 84). Margulies suggests that the key durational filmmakers of post-war
modern cinema, Warhol and Akerman, adopt an overtly literal approach to temporal realism that deliberately
exhausts the 'iterative' representation of the everyday in neorealism. Neorealism's 'iterative' operation derives
from its aforementioned moral positioning as an art of social awareness and implication, in which imagesound
situations concerning a single character or location may signify an entire class, or, in the immediate
post-war period, any number of devastated spaces in war-ravaged Italy and other parts of Europe. The
minimal hyperrealism of Warhol and Akerman's work, however, replaces this mimetic representation with a
literal depiction of place and (idle or everyday) activity, centred on the use of the fixed frame and extended
takes to amass a fullness of material record. This strategy, as Chris Berry implies in a comment on Tsai's
Vive l'amour, “performs its realism so excessively as to draw attention to itself, making it a limit case
realism,” and produces “a sense of plenitude, similar to that produced by the standard realist film and yet
different” (C. Berry 2005: 89, 94). In what Margulies refers to as the 'minimal-hyperrealist' image in
Akerman's work, a sense of concreteness is returned to objects and places, and a rendition of “literal time
robs [the image] of the possibility of standing for something other than that concrete instance” (Margulies
1996: 37).
This surplus of realism often foregrounds the literalness of duration by imparting a heightened sense
of its passage. At times the effect is augmented by familiar devices such as an unnatural acting style, as Berry
suggests is present in the “hyperbolic realism” of the three characters' blank enactments of loneliness and
boredom in Vive l'amour (C. Berry 2005: 89). Akerman, in Jeanne Dielman, 23 quai du Commerce, 1080
112
Bruxelles, carefully distends the impression of time by subtly manipulating diegetic elements that the camera
otherwise records with utmost literalness. Akerman states that, in many sequences in the film, “time was
totally recomposed, to give the impression of real time,” and has spoken of directing Delphine Seyrig's (in
her performance as Jeanne) gestures to affect their duration, such as requesting that Seyrig prepare schnitzel
more slowly by delaying the downward movement of her arm, then pick up the sugar pot slightly more
promptly, and so on. Akerman discloses that her directorial method was actively hyperrealist in intent: “I
didn't want to manipulate [Delphine]. I showed her afterward and said to her, 'You see, I don't want it to look
real, I don't want it to look natural, but I want people to feel the time that it takes, which is not the time that it
really takes'” (M. Rosen 2004: 125).48
In modern cinema, the foregrounding of a literal depiction of duration is also exemplified by Andrei
Tarkovsky's work and writing, which combines an almost spiritual disposition to materialism with the
imprinting of 'time-pressure' “in the frame” itself (a combination that Tarkovsky refers to as 'time-truth')
(Tarkovsky 1987: 117, 120). The influence of Tarkovsky's work and prose on later durational cinema cannot
easily be overstated, and the lineage from his and Miklós Jancsó's modernist cinema to the mobile long-shot
long-take styles of Béla Tarr and Fred Kelemen (who also worked as Tarr's cinematographer for A Londoni
férfi [2007] and A torinói ló [2011]) is a key component of contemporary slow film. Writing of Stalker
(1979), Tarkovsky's de-dramatised adaptation of Arkady and Boris Strugatsky's Roadside Picnic (1972),
Tarkovsky drew attention to the literalness of his approach to duration in representing the quest for the Zone:
“I wanted time and its passing to be revealed, to have their existence, within each frame. […] I wanted it to
be as if the whole film had been made in a single shot” (Tarkovsky 1987: 193-194). In Tarr and Kelemen's
cinema, sequence shots emphasise their durations by becoming tests of the endurance of existence, imparting
conditions that slow the physical actions of humans trapped in them (in maelstroms of punishing wind, rain
and ice, a morass of dense mud, and the perennial void of pitch black nights). Rosenbaum has described Tarr
and his cinema, appropriately, as a “despiritualised Tarkovsky,” and the same notion applies, in particular, to
Kelemen's grim, forbidding Frost (1997)—a film composed of deathly sequence shots shorn of all causal
operation except the persistence of the handheld camera's ceaseless crawl (Rosenbaum 2004: 51). There is
48
Sami Frey's documentary Autour de 'Jeanne Dielman' (1975), edited by Agnes Ravez and Akerman herself, provides
a valuable document of Akerman's instructions to Seyrig regarding de-dramatised performance. Akerman
consistently refuses to disclose the “secrets” or psychological motivation of Jeanne's character to Seyrig, who at one
point replies: “I get the impression it's a matter of speed.”
113
little belief in the world, or a future, in Tarr and Kelemen's films, and the everyday merely appears to signify
an unbearable boredom, and inexorable duration, that can only be survived. At one point, Tarkovsky wrote
that he wanted the filmic frame in his work to be unlimited in its visual depiction by serving as “a pointer to
life”—in light of this, it might not be an exaggeration to suggest that Tarr and Kelemen take such a striving
for the infinite to its most literal, atrophic conclusion (Tarkovsky 1987: 117).
Central conflict
Bazin's notion of a literal destruction of drama, cited at the beginning of this chapter, also points to a more
elemental dynamic in contemporary slow films: the disruption of 'central conflict' on a structural level. The
concept of central conflict was devised by Raul Ruiz in the first volume of his Poetics of Cinema (2005), and
signifies the “all-encompassing narrative and dramatic theory” that provides a backbone to plot in classical
and dominant cinema. In Ruiz's definition, central conflict theory affirms that “a story begins when someone
wants something and someone else doesn’t want them to have it,” and binds all narrative elements to this site
of perpetual contestation (Ruiz 2005: 11). Ruiz suggests that central conflict mostly produces linear,
possessive forms of narration, and “forces us to eliminate all stories which do not include confrontation and
to leave aside all those events which require only indifference or detached curiosity, like a landscape, a
distant storm, or dinner with friends—unless such scenes punctuate two fights between the bad guys and the
good guys” (Ruiz 2005: 11). This resistance to non-narrative eventfulness originates in North American
classical film, and constitutes a normative ideological system consistently upheld by international intensified
continuity style—that is, by, as Ruiz half-jokes, its “theologians, inquisitors, and police force” (Ruiz 2005:
15).
The critical operation of central conflict, for Ruiz, is that it proposes a “strict equivalence between
stories of conflict and everyday life,” and cannot incorporate either slow cinema's “privileged moments” of
boredom and ennui, or Ruiz's own poetic, “cartographic” narration that folds layers of reality, fiction and
hallucination until one layer becomes indistinguishable from another (Ruiz 2005: 15). Central conflict has,
therefore, been most definitively renounced only by select types of art and experimental film that reduce it to
its most naked essence: in, for instance, the mere shard of fiction in Snow's Wavelength (1967) (where Amy
114
Taubin's phone call to report the discovery of a body is obliterated by the camera's inexorable zoom), or the
metaphysics of duration in Tarkovsky's Stalker (1979) (in which the quest to reach the Zone functions as an
allegorical struggle of faith).
Central conflict serves as the basis for the organisation of duration in contemporary commercial
cinema, and classical continuity style remains its benchmark. I have thus far referred to the operations of
classicism in broad terms as the inverse of the Modern, but it is important to clarify that classical continuity
style continues to form the basis of the narrational system of international dominant cinema that the marginal
contemporary filmmakers under discussion here mostly operate in exile from. David Bordwell notes that
intensified continuity style (international cinema's dominant visual language) has now existed for almost
fifty years, as long as the Studio-era continuity system in which Welles, Wyler and Toland tested the stylistic
genesis of modern cinema (Bordwell 2006: 189). In terms of its default content, little has changed: in
contemporary dominant films, narratives present characters that we are encouraged to care about, in
narratives that are framed by the resolution of central conflict's site of contestation. Bordwell has noted that,
until recently, it has not been necessary to posit a 'postclassical' cinema, as intensified continuity in its most
common form continues to conform to classical precepts. Bordwell does note, however, that “in its wilder
reaches, [intensified continuity] presents a boisterousness in tune with the edgier examples of innovative
narrative,” but, largely, it is a style that “prefers to parade virtuosity while remaining within the ambit of a
stable system” (Bordwell 2006: 189).
Contemporary cinema, fast and slow
In recent years, roughly analogous with the continued rise of slow cinema, Hollywood's application of
spectacle to intensified continuity has further intensified itself (in terms of both speed and affect), giving way
to a style that we might call 'hyper-intensified' continuity, or, in Steven Shaviro's term, 'post-continuity' (see
Shaviro 2012). This new style has, understandably, often been placed in stark opposition to slow cinema, and
exhibits an almost total disregard for the visual continuum of action and geography (particularly in its most
spectacular sequences) by bathing the spectator in an incessant flow of images and sounds that, as Shaviro
notes, “do not interact dialectically” and instead only form allusive, indirect relations with each other. In a
115
manner reminiscent of anti-narrative structures found in certain slow films, the images of post-continuity
films “seem to be linked together merely in the manner of bricks or building blocks,” in turn breaking
continuity rules by refusing to organise the narrative operation born of central conflict “in such a way as to
maximise narrative flow and impact” (Shaviro 2010a: 72, 73).
Many critics, as previously mentioned, have pointed out the ever-expanding gap between this
emergent mode of contemporary 'fast' and slow cinema. Robert Koehler, for instance, sets up an antagonistic
contrast between dominant “CGI fantasy cinema,” which involves “a hyperplasticity and comic-book
framing that celebrates its own one-dimensionality,” and slow cinema's de-dramatised, explicitly Bazinian
realism (Koehler 2008b). Post-continuity style is only evident in a minority of the former “CGI fantasy”
trend—for instance, Michael Bay's first three Transformers films (2007, 2009, 2011), and Neveldine/Taylor's
Gamer (2009)—but its emergence in tandem with an increasing rigour of undramaticness in the slow art,
experimental and 'festival' film is, I would suggest, more than coincidental.
Although post-continuity films do not eradicate central conflict, they do disrupt its standard
containment of duration. In both Studio-era classical and intensified continuity style (from which postcontinuity
style emerges), the main purpose of narration is to render events that are integral to the narrative
and its creation of emotional and psychological effects, whilst skipping any intervals that would threaten to
inhibit “an orderly flow of action” (Bordwell et al 1985: 44). Duration is only experienced on the level of
narration, as the regulated constraint of central conflict forbids extended shot digressions (so as not to risk
boredom), and the flow of action fixes on both a search for meaning and the projection of future events or
outcomes. Bordwell notes that, in order to satisfy the default demands of character-based fiction, classical
films tend “to plunge [characters] into a swirl of cause and effect, goals and obstacles, conflicts and
resolutions, appointments and deadlines,” creating anticipation for the development and resolution of conflict
(Bordwell 2006: 107). Hence, as Bordwell, Staiger and Thompson note, duration can only become “a process
to be investigated on its own” rather than a vehicle for causality when this operation is broken down and
“'dramatically meaningless intervals'” are allowed to attain a level of prominence in excess to the narration's
prescribed structure. It is this dynamic that gives rise to the foundational stricture in classical Hollywood
cinema that “a walk without dialogue is 'dead' or wasted time,” in comparison to its more privileged role in
post-war modern cinema (Bordwell et al 1985: 47).
116
As Theodor Adorno and Max Horkheimer describe in their conception of the culture industry, this set
of strictures and 'swirl' of conflict and resolution can produce a “rigidly invariable” conformity that
effortlessly reproduces its own relentlessness: “nothing remains as of old; everything has to run incessantly,
to keep moving,” even if like a machine rotating on the same spot (Adorno and Horkheimer 1997: 125, 134).
Hence, unlike contemplative film, orthodox classical cinema risks leaving “no room for imagination or
reflection on the part of the audience,” meaning that “sustained thought is out of the question if the spectator
is not to miss the relentless rush of facts” (Adorno and Horkheimer 1997: 126-127). Duration is therefore
excessively patterned, and defined primarily by what the film leaves out (that is, the undramatic aspects that
are reclaimed in modern cinema), and the dramatic appointments and deadlines that the narration sets.
Post-continuity cinema and the temporality of neoliberalism
Post-continuity film, like the field of slow cinema, has begun to pull against this foundational operation of
classicism. Shaviro notes that, in the affective stylisation of Richard Kelly's Southland Tales (2006) and
Olivier Assayas' Boarding Gate (2007), “continuity rules are used opportunistically and occasionally, rather
than structurally and pervasively. Narrative is not abandoned, but it is articulated in a space and time that are
no longer classical” (Shaviro 2012). Like the sequenced blocks of time in durational film, the looseness and
arbitrariness of montage in Southland Tales is “in fact the very point of the movie,” and signifies an active
refusal to “coalesce into any sort of higher, synthetic unity” (Shaviro 2010a: 74). Rather unlike the
contemporary slow film, however, duration is produced to enact a series of “hypermediated, heightened and
intensified Nows” that embody a sense of the accelerated, imagistic, free-floating matrix of late capitalism,
where all that is solid melts into air (Shaviro 2009).
Shaviro identifies a quite unique operation in the image-sound situations of Boarding Gate, and
proposes that its rapid editing, unpredictable alternations in perspective, furiously drifting camera, and light,
sleek stylisation enact the spaces and flows of neoliberalism: the flexible motion of capital that, as David
Harvey describes, “emphasizes the new, the fleeting, the ephemeral, the fugitive, and the contingent in
modern life” (Harvey 1989: 171). Boarding Gate centres its narrative conflict on the movement of people
and illicit capital under globalisation, but its form also depicts an entirely relational space: a global 'world
117
system' driven by the proliferation and decentralisation of finance capital, enabled by modern technological
innovation and the international expansion of speculative market activity. Shaviro notes that this space
“cannot be 'represented' in any form accessible to the human senses” (as it can only be defined
computationally), but Assayas' achievement is to translate the “impalpable flows and forces of finance into
images and sounds that we can apprehend on the screen” (Shaviro 2010a: 36).
One noticeable aspect of discourse around contemporary durational films is the common assumption
that their tendency toward 'slowness', in terms of temporality, style and tone, ignores or blocks out a sense of
the totality or lived experience of modern life under neoliberalism (see, for instance, Bíró 2006, Romney
2010 and, unfortunately, Shaviro 2010b). Whilst Shaviro's valuable recent work on film tracks the changing
mediascape of late capitalism through dominant cinema's production of images and sounds, it is worth
considering that recent slow cinema might similarly reflect on the changing temporality of modern life.
Whilst it is certainly true that many contemporary durational films signal a retreat from narrativising the
accelerated pace or instability of life in affluent or rapidly developing economies, this is mostly intended for
the purpose of sustained observation, often in the form of a de-dramatised Bazinian realism. The distinctive
aesthetics of slow films tend to emerge from spaces that have been indirectly affected or left behind by
globalisation, most notably in the films of Alonso, Bartas, Jia, Costa and Diaz. Alongside the contemporary
neorealism of La libertad, many individual works by these filmmakers turn their attention to marginal
peoples (low-paid manual labourers, poor farmers, the unemployed and dispossessed, petty criminals and
drug addicts) subsisting in remote or invisible places, and depict the performance of (waged or unwaged)
agricultural and manufacturing work that is increasingly obscured by the macro volatility of finance-capital's
huge speculative flows.
Whereas slow cinema undertakes the sustained process of (re-)seeing and thinking the 'tatters of this
world' (in Deleuze's phrase), Shaviro argues that the response of post-continuity style is one of acceleration;
in turn suggesting that, following the internal logic of late capitalism, “the only way out [of late capitalism]
is the way through” (Shaviro 2009). Shaviro proposes that post-continuity cinema, like contemporary slow
film, belongs to Deleuze's model of the post-war time-image regime: that is, a cinema where time rises to the
surface of the screen in response to a dislocated relation between humans and the world (see Shaviro 2010a:
118
59-60).49
Whereas slow films tend to produce time in the form of surplus and excess (in contrast to
movement-image cinema), post-continuity film is marked by its unswerving refusal of a material density of
time, producing only “a thin sliver of pure present, without any thickness of duration” (Shaviro 2009).
Neoliberalism has altered our cognisance of temporality on the level of both capitalist circulation and
its impact on culture. The intensification of the market as a world system, combined with innovations in
communication technologies, has given rise to a world system in which, as Fredric Jameson describes,
“capital transfers abolish space and time and can be virtually instantaneously effectuated from one national
zone to another,” often with incalculable long-term results (Jameson 1998: 143). These inexorable,
accelerated capital transfers primarily circulate in the form of (either expedient or arcane) financial
instruments that link geographically dispersed markets and open, by speculative force if necessary, “local
sites of difference to the flow of financial power” (LiPuma and Lee 2004: 125).
Over the course of the last two decades in particular, this financial power has principally asserted
itself through the device of the derivative: an instrument that structures and manages duration in relation to
capital's exposure to risk and volatility. As Edward LiPuma and Benjamin Lee note, for the derivative “the
mitigation of circulatory risk depends on the compression or neutralisation of time,” and increasingly this
mode of temporal compression provides the very means for finance-capital to accumulate or extract a
prevailing source of profit and reproduction. Hence, in order to minimise risks and maximise connective
utility, speculative capital “must continually turn itself over in the shortest time possible” in an unforgiving
cycle of competitive pressure (LiPuma and Lee 2004: 129). Late capitalist temporality is, in this way,
relentlessly compressed to a transactional point of milliseconds—a stage at which, as Donald MacKenzie
suggests in the context of high-frequency trading (HFT) that takes place in mere milliseconds, only “the
speed of light is an insuperable barrier” (MacKenzie 2011: 16).50
This new sense of temporality creates, in
seemingly immaterial form, the impression of a world that Shaviro describes in Southland Tales as
49
Shaviro objects, with validity, that the distinction between the movement-image and the time-image has exclusively
come “to be equated, somewhat unfortunately, with that between mass-market movies and art-house cinema”
(Shaviro 2010a: 60).
50
Felix Salmon, in a post that features an astonishing visualisation (in .gif form) of the present state of HFT activity in
the stock market between 2007 and 2012, describes the market today as “a war zone, where algobots fight each
other over pennies, millions of times a second,” and that “the HFT world is so complex that its potential systemic
repercussions are literally unknowable” (Salmon 2012). Alongside Donald MacKenzie's work, Salmon and Jon
Stokes' feature article in Wired is a solid primer to the compression of duration signified by HFT (see Salmon and
Stokes 2010).
119
“accelerated to the breaking point” (Shaviro 2010a: 90).
It is, equally, often said that contemporary commercial film reflects the acceleration and dislocation
of modern life, at least in the affluent West, through its own increasing abundance of speed and spectacle.
With this in mind, Jameson describes dominant cinema under neoliberalism as collections of “narrativised
image-fragments” that persistently offer “the creative transformation not of riches into dead leaves, but rather
of banalities into elegant visuals self-consciously offered for the eye's consumption” (Jameson 1998: 161,
157). Late capitalist cinema has thus far primarily fragmented time and space through the acceleration of
editing, meaning that the standard 'analytical cut' of classicism is now tasked with far more than a
communication of dramatic action with Fordist efficiency.
Bordwell has observed that some intensified continuity films “flirt with shot lengths reminiscent of
late 1920s Soviet silent montage,” and the relentlessly disorientating, frenzied post-continuity editing of
Michael Bay, Neveldine/Taylor and Tony Scott's recent work deliberately embodies a “twitchy, agitated
interpassivity” emblematic of the visual language of advertising and new media in contemporary culture
(Bordwell 2006: 122; Fisher 2009: 24). Mark Fisher ascribes this hyperactivity to a pathology born of the
condition of 'capitalist realism', and argues that such visible symptoms of attention deficit hyperactivity
disorder arise from “being wired into the entertainment-control circuits of hypermediated consumer culture”
(Fisher 2009: 25).51
Bay, Neveldine/Taylor, and Scott would, quite rightfully, take this allusion as a
compliment, and I do not intend it to constitute a wholly unfavourable judgement here.
The disparity between the production of duration in recent slow films and intensified and postcontinuity
style clearly represents two extremes of contemporary film. However, I would like to suggest here
that this very disparity produces a dynamic uniquely reflective of temporality in neoliberal culture, one
proposed by Jean-François Lyotard in his concept of 'acinema'. As the aesthetics of post-continuity film
directly reflect the intensifying conditions of neoliberalism (that is, the totality of its transformation of the
capitalist world system and the temporal organisation of aspects of everyday life in the affluent West), the
equally polar radical aesthetic of slow film should also be considered to have emerged, quite directly, from
51
Fisher also makes a direct connection between style and late capitalism's cultural anxiety in relation to Paul
Greengrass' The Bourne Supremacy (2004) and The Bourne Ultimatum (2007): “Bourne's transnational nomadism is
rendered in an ultra-fast cutting style which functions as a kind of anti-memory, pitching the viewer into the
vertiginous 'continuous present' which Jameson argues is characteristic of postmodern temporality” (Fisher 2009:
58).
120
this same set of conditions.
Aspects of post-continuity and slow cinema might be considered to represent, respectively, Lyotard's
distinction between the disposition toward “extreme mobilisation” and “extreme immobilisation” that can
thwart the superstructural function of the film industry under late capitalism (Lyotard 1986: 356; my
emphasis). In the first instance, post-continuity style produces an extreme mobilisation of hyperactive motion
and affective fragmentation, which is met by its opposite disposition in slow cinema's immobilised images of
cardinal stillness and blocks of temps mort. The immobilisation in the tendency toward stillness in slow
cinema serves to form the basis of a style that cannot be mobilised or accelerated, and instead works to
produce an antithetical impression of duration in relation to dominant commercial cinema.
These two extremities of form arise from Lyotard's contention that the commercial film industry can
only value image-sound situations “following the cyclical organisation of capital” (Lyotard 1986: 353). For
Lyotard, film form constitutes the inscription and ordering of movement, and the style of commercial film
must be synthetically composed “like a unified and propagating body” in order to maximise profit through its
production of images and sounds (Lyotard 1986: 352). The task of the commercial filmmaker, then, is to
shoot and select only commodifiable movements that conform to the basis of central conflict and the
operation of continuity style, displaying clear representational images free of errors, abstractions or
impulsive movements. Lyotard's distinction is similar to Adorno and Horkheimer's conception of the culture
industry (representative of the “ruthless unity” of resolutely classical form under modern capitalism), in the
sense that what Adorno and Horkheimer referred to as the “constant sameness” of mass-produced continuity
permits “all sorts of gaps, jolts, postponements, losses, and confusions” to occur, but only if they “no longer
act as real diversions or wasteful drifts” (Adorno and Horkheimer 1997: 123, 134; Lyotard 1986: 353). This
commodification of cultural components continues in the postmodern era, where, as Jameson describes,
“'culture' has become a product in its own right,” and “the market has become a substitute for itself and fully
as much a commodity as any of the items it includes within itself” (Jameson 1991: x).
Hence only an extreme mobilisation or immobilisation can disrupt “the synthesis of good movement”
that actualises capital in film form, and disrupt the illusion that permits its superstructural function. Slow
films can achieve an effect of immobilisation by depicting “useless expenditures” ('false' movements, to use
Deleuze's term) that cannot be subsumed by capitalist circulation, and by employing the formal strategy of
121
the tableau vivant, both of which point to a resolute stillness (and seeming 'unproductiveness') that renounces
the propulsion of circulatory logic.
These 'useless expenditures' (of either image-capital or time) occur in the attention to particular types
of undramatic activity in slow cinema, and its tendency to produce a surplus of duration within the shot. In
the first instance, useless expenditures are manifested by the depiction of mute, subdued or private
behaviours, like the activities of Lee Kang-sheng in Vive l'amour, or the depictions of wasted everyday time
in Oxhide, or, for instance, Apichatpong Weerasethakul's Blissfully Yours (2002). The 'disproportionate' shot
duration given over to wasteful activities in these films perhaps recalls Boris Groys' description of the
function of “suspended time” in time-based art: that is, a literal representation of duration that “captures and
demonstrates activities that take place in time, but do not lead to the creation of any definite product” (Groys
2009). In marginal slow films, unproductive activities abound that are comprehensively rejected by the
economy of the commercial film industry, and in this sense it could be argued that, in slow cinema, time is
not invested 'properly' in the formation of images for narrative efficiency, so cannot turn (or maximise) a
profit.
Slow films instead point to an “unproductive, wasted time” that is symptomatic of the aftermath of
modernism—a concept used by Groys to describe the problem of “truly” contemporary art, which is stranded
in a time not absorbed by historical process, or the expectation of a vanishing future (Groys 2009). Under
capitalist realism, as Shaviro points out, time is produced in a manner that is frequently reductionist, yet only
'wasted' for the purpose of relentless circulation: in late capitalist culture, duration has “imploded,” and
shrunk “to a dimensionless, infinitesimal point” where “time is emptied out, or whittled away” (Shaviro
2010a: 90). Groys argues that time-based art, in its minimal, literal, or repetitive modes, can resist this
implosion, and demonstrate “the 'bad infinity' of wasted, excessive time” that cannot be absorbed by either
capitalist circulation or the (Modern) procession of historicity. In this way, the mode of durational film
represented by slow cinema can transform “a scarcity of time into an excess of time” (Groys 2009; my
emphasis).
Before returning to the implications of contemporary slow cinema's relation to post-continuity film
and neoliberalism's compression of duration, I would like to briefly address two significant precedents for
what is ostensibly an opposition between commercial and marginal (durational) aesthetics here. The first
122
concerns Serge Daney's reflections on the 'anti-consumerist' cinema of Jacques Rivette, and the second
concerns the political cinema of Jean-Marie Straub and Danièle Huillet. Both cinemas are deliberately
ineffective at reproducing the 'synthesis of good movement' that Lyotard refers to as actualising capital in
film form, and are worth discussing before coming back to the principal issue at hand.
Images that don't sell anything
Speaking in Pierre-André Boutang and Dominique Rabourdin's documentary Serge Daney: Itinéraire d’un
ciné-fils (1992), Serge Daney points to a conception of cinema that appears to resist preserving the cyclical
organisation of image-capital. Following Jameson's observation that “aesthetic production today has become
integrated into commodity production generally,” Daney draws attention to the increasing convergence
between cinema and advertising under late capitalism: its formulation of “a realm of images” that are “sold
into prostitution” by shaping aesthetics entirely on terms of monetary transaction (Jameson 1991: 4). For
Daney, the locus of this realm of images is demonstrated by Jean-Jacques Annaud's L'Amant (1992), which
adopts a form that traps the spectator in a “feeble status as consumer-decoder.” Daney, in a scathing critique
of Annaud's film, argues that the filmmaker merely presents an assemblage of thinly disguised
advertisements, logos and kitsch, creating a series of “orphan images which must, one by one, be seen,
recognised and, so to speak, ticked off by the spectator-consumer” (Daney 1992: 16).52
Modern cinema has, however, as Daney suggests in Itinéraire d’un ciné-fils, produced adversaries of
this vulgar aspect of late capitalist cinema: that is, “filmmakers who make images that don’t sell anything.”
Daney cites Jacques Rivette in this context, as a “pure cinéphile who lives outside of consumerism” and
observes, with “an intense curiosity, the life of his contemporaries.” Deleuze describes Rivette's cinema as,
simply, “the art of light,” an appropriate summation that is echoed by Daney: “When you see Le Pont du
Nord [1981], to take one of his most beautiful films, there isn’t a single shot in the film that could sell
anything: one that could sell the actress who plays in it, the quality of the sun, nothing. Because it’s used for
something else, it’s used for building time” (Deleuze 2005b: 11).
Before Le Pont du Nord, Rivette filmed Paris s'en va (1981), a non-narrative short film that consists
52
In a similar context, Dudley Andrew applies Daney's critique to a more contemporary film: Jean-Pierre Jeunet's Le
fabuleux destin d'Amélie Poulain (2001). See D. Andrew 2010: 17-25.
123
Fig. 31: Paris s'en va (1981)
of glimpses of events, and documents of locations, that appear in its companion film. Paris s'en va was
designed as an embryonic version of the fictional structure of Le Pont du Nord, but is particularly noteworthy
(in retrospect) for isolating the aspects of Rivette's cinema described by Daney. For Paris s'en va, Rivette
strips out narrativity and replaces action with a series of concrete images of place and dérive-like movement:
bursts of lens glare as the camera circles the Lion of Belfort statue in Place Denfert-Rochereau; light
shimmering across the placid Canal Saint-Martin [Fig. 31]; 'empty' images of barren streets and sites of
industrial wasteland; and still lifes of laundry hanging from tenement windows (a traditional signifier of the
everyday in modern cinema that I shall return to in my next chapter). Like Peter Hutton's short films, or
Gianvito's Profit motive and the whispering wind, the content of Paris s'en va is the sensuous capacity of the
filmic image to render documentary record by replacing central conflict with 'real diversions' of impressions
of light and time.
124
Although I would hesitate to classify Rivette's work as 'slow' or strictly durational, his films of the
last decade do reflect the gradual turn toward increasingly undramatic tendencies in contemporary art film:
Histoire de Marie et Julien (2003) and Ne touchez pas la hache (2007) are perhaps his most sombre works,
consisting of haunted and opaque image-sound situations, and his final work, 36 vues du Pic Saint-Loup
(2009), is almost naturalist (drawing on a gently undramatic situation, a cast of few characters, and images
that are open to the wind in the trees and the moon hanging above the mountains). Deleuze suggests that
Rivette's games of fiction, in their rendering of visions that compulsively slide between past and present (and
the mental and the physical), are compelled to create a sense of the real “through the force of visual
description” (Deleuze 2005b: 11; my emphasis). This dynamic reflects a key principle of modern cinema,
and, as Daney suggests, signals a retreat from the market economy of images that is represented by L'Amant.
For some modern filmmakers, resistance to the style of late capitalist commercial cinema is a
deliberate political stance. The work of Jean-Marie Straub and Danièle Huillet is perhaps most prominent in
this respect, and many commentators have drawn attention to what Martin Walsh describes as the “political
necessity” of their anti-illusionist “restructuration of cinematic language” (Walsh 1978: 14, 15; see also
Fieschi 1980; Byg 1995; Perez 1998).53
Straub-Huillet's aesthetic project is deeply rooted in the high
modernism of the 1960s and 1970s, and their uncompromising commitment to an ostensibly modernist style
after that period (presently maintained by Straub himself, after Huillet's death in 2006) has remained doubly
resolute as dominant cinema has itself altered and accelerated. Straub has argued that a political cinema can
be defined as the total avoidance of “what keeps capitalism alive, such as inflation,” and has suggested that
“if, at the aesthetic level, you practice the same inflation which fuels capitalist society as well as the world
we live in, then there's no point; you're just grist for the mill” (Albera 2004: 42).
On a formal level, Straub-Huillet retreat from 'inflationary' cinema in two ways: by reducing the field
of the image to a depiction of the fullness of minimal gesture and place; and, as Walsh notes, by rejecting
“the unifying repetitions of classical narrative” for a narration that organises each segment “according to a
particular logic of its own” (Walsh 1978: 27). As Barton Byg observes, both strategies feed into a process
that attempts to “simplify each shot to the point that it conveys one idea clearly,” as “only an 'empty' frame
53
Not that it is necessarily important to note, but at one point in Costa's Où gît votre sourire enfoui? (2001), Straub
pledges allegiance to rebuilding communism as “the only thing that will ever save the earth, and the children of the
earth.”
125
can capture the invisible textures of the surface of the world” (Byg 1995: 21).
Straub's defiant position is clear in his memorable critique of what he calls dominant cinema's
“paratroopers”: filmmakers “who simply fall from the sky somewhere and, boom, the camera is running
already. They film something they have never even seen.” Straub argues that patient observation is the only
necessary response to the modern world, a position that reflects his and Huillet's insistence on filming and refilming
the world in order to adequately depict a sense of the (fleeting and historical) continuum of material
reality. Thom Andersen recalls that during the filming of Klassenverhältnisse (1984), Straub-Huillet worked
with a small crew of “only six or seven people” and a limited budget of $250,000, yet insisted on shooting
twenty or thirty takes of each shot on 35mm (Andersen and Costa 2007: 43). Andersen comments that this
made Straub-Huillet, in their own description, “the richest filmmakers in the world,” as “they had found a
way of working which enabled them to do what needed to be done, to do as many takes as they thought were
necessary, by simplifying other aspects of production” (Andersen and Costa 2007: 44).
This insistence on the repetition of recording is central to Straub-Huillet's practice, and, when valued
against the economy of commercial cinema, constitutes deliberately unproductive activity. The practice of
waiting, looking and listening (with the camera) illustrates Straub-Huillet's singular rigour and attention to
the real diversions (in Lyotard's term) located in Rivette's work: the body of the actor, the quality of the sun,
and so on. In the end, it is only the smallest material detail—an actor's blink, the movement of a branch in
the wind, a birdsong on the soundtrack—that dictates Straub-Huillet's selection of a final shot from a surplus
of takes.54
As Straub states in Costa's documentary Où gît votre sourire enfoui? (2001), they are filmmakers
who respond to the world with rigorous observation, rather than circulate 'image-fragments' “for the sake of a
so-called wealth of imagination.”
54
The DVD edition of Klassenverhältnisse released by Edition Filmmuseum in 2007 includes a valuable document of
this process: Klaus Kanzog and Klaus Volkmer's Work in Progress. Genetic Analysis of the Sign Structure and
Rhetoric in the Opening Sequences of the Film 'Klassenverhältnisse' on the Basis of Unused Takes (2007). Kanzog
and Volkmer compile multiple takes of the same subject to illustrate the degree of subtlety in Straub-Huillet's
selection of final shots. Overwhelmingly, the only elements that differ between takes include the position of the
microphone (to record direct sound), the speed of the wind, and the quality of ambient light.
126
Contemporary slow cinema and neoliberalism: a historical evolution
In linking the disparity between the aesthetic extremities of dominant cinema and slow cinema here, I am
proposing that both poles of film form are equally symptomatic of late capitalism's transformation of cultural
logic. Following the tentative chronology that I set up in my Introduction, the field of slow cinema under
discussion here gradually emerges at the same time as dominant cinema's intensification of speed and
spectacle in the 1980s and 1990s (in line with the speculative evolution and acceleration of late capitalism),
and expands further at roughly the same time as commercial film 'jumps the shark', in Shaviro's phrase, to
form post-continuity (Shaviro 2012). I concur with Robert Koehler that the present 'mature' stage of slow
cinema starts, just after the turn of the millennium, with Alonso's La libertad, and has continued most
prominently in the work of Alonso and his contemporaries: Apichatpong Weerasethakul, Lav Diaz, Jia
Zhang-ke, Liu Jiayin, Raya Martin, and Albert Serra. Hence post-continuity style emerges in Hollywood at
the same time as key works of durational and undramatic cinema are first screened in festivals and arthouse
theatres: for instance, Apichatpong's Syndromes and a Century (2006), Diaz's Heremias Book I: The Legend
of the Lizard Princess (2006) and Death in the Land of Encantos (2007), Jia's Still Life and Dong (both
2006), Liu's Oxhide (2005), Martin's Autohystoria (2007) and Now Showing (2008), and Serra's Honor de
cavalleria (2006) and El cant dels ocells (2008).
There is, in terms of the relationship between culture and capitalist circulation, a significant
historical precedent to this movement toward two extremities of contemporary cinema. I have previously
argued that slow cinema can be considered to be a continuation of 'modern cinema', in Bazin and Deleuze's
open-ended distinction, after the specific period of post-war modernist film practice outlined by Bordwell
(1997), Kovács (2007), and Lévy (2011). Kovács defines this era of post-war modernism as beginning in the
early 1950s and ending in the mid- to late 1970s in the face of the onset of postmodernism—that is, late
capitalism's cultural and ideological 'structure of feeling'. Although Bordwell sees modernism predominantly
as a formal practice, Kovács sees it as an ideological project belonging to a specific historical moment that is
abandoned when cultural relations decisively shift during the mid- to late 1970s—a chronological transition
that corresponds to both Jameson and David Harvey's accounts of the emergence of postmodernism
(although Jameson does imply that aspects of postmodernism first become visible in the 1950s and 1960s)
127
(see Harvey 1989; Jameson 1991).
Kovács notes that, after the collapse of modernism at the end of the 1970s, “it is quite obvious that
some of its stylistic and narrative innovations continued to enrich different cinematic practices” (Kovács
2007: 45). The slow cinema that comes after this period directly embraces a select set of strategies from late
modernist style, and continues to deploy those strategies in a different ideological and economic context. To
this end, Kovács notes that filmmakers such as Béla Tarr and Abbas Kiarostami first accepted the
“minimalist continuity style” of late modernism (pioneered by Wim Wenders' Kings of the Road [1976] and
Akerman's Meetings with Anna [1978]) in the 1980s by adopting a set of stylistic devices in which “empty
space/time is the prevalent subject matter of the images” (Kovács 2007: 156, 158).
In light of this continuum, the aesthetic or narrational tendencies of slow cinema do not appear to
comprise, I would suggest, a postmodern cinema—in the sense that they are not explicitly reliant upon
nostalgia, pastiche of popular convention, or a fragmentation rather than continuity of (the impression of)
duration—but, equally, they do appear to be deeply symptomatic of the cultural shift in media and
temporality under late capitalism since the late 1970s. As in Shaviro's account of recent commercial film's
movement toward post-continuity, the historical emergence of forms of slowness in contemporary cinema
might be most productively placed within the context of the intensifying neoliberal turn over the last three
decades—and, indeed, their emergence coincides with the point at which neoliberalism is initially
implemented as a set of economic reforms in key regions of the capitalist world system.
Before continuing, it is worth noting that although I do not consider slow cinema to be a postmodern
rather than simply post-modern cinema, it is not my aim here to contest the conceptual rupture between
modernism and postmodernism. By invoking Bazin and Deleuze's conceptions of 'modern cinema' (which
are, inevitably, not limited to distinctions between modernist and postmodernist aesthetics), I have thus far
implied that visible aspects of the modern continue to exist in a postmodern era. However, it is important to
clarify here that the cultural shift signified by postmodernity does, in fact, most likely condition the historical
emergence of recent trends in durational cinema.
Fredric Jameson, in his essay “The End of Temporality” (2003), suggests that the “epochal change”
between modernism and postmodernism is defined by their divergent obsessions with time (for modernists)
and space (for postmodernists). Jameson notes that the modernist obsession with time has tended to manifest
128
itself on an “interior” level of consciousness, subjectivity and desire, whereas the postmodern concern with
space comprises “a realm of exteriority” that is centred on relations between people and place, and trends
such as urbanisation and globalisation (Jameson 2003: 697). The spatiality of postmodernism attests to
changes in the network of the capitalist world system since the 1970s that have been brought about by an
intense consolidation of existing power structures—which, as Michael Hardt and Antonio Negri contend in
Empire (2000), have provided “capital with potentials on a scale previously unimaginable.” Hardt and Negri
observe that postmodernist thinking has, perhaps inevitably, flourished “in the various fields of practice and
theory proper to capital, such as marketing, management organisation, and the organisation of production,”
and suggest that it fully incorporates “the logic by which global capital operates” today (Hardt and Negri
2000: 151).
The postmodern era, for Jameson, is defined by late capitalism's “dramatic and alarming shrinkage of
existential time and the reduction to a present that hardly qualifies as such any longer” (Jameson 2003: 708).
This shrinkage of time is effected in a number of ways: in the market dynamics that produce and accelerate
vast speculative capital flows (such as in the form of the derivative, as described above); global changes in
capitalist production and its mobilisation of labour (as de-industrialisation and financialisation in Europe and
North America has been met by a concomitant expansion, and urbanisation, in developing Asian, African and
South American economies); the dissolution of national boundaries through the liberation of trade (that is, of
commodity production and exchange); and the innovation of digital and media technologies (mobile
telephones, the internet) that have not only enhanced but enabled new forms of connectivity.55
These factors of temporal (and spatial) compression amount to what Jameson calls the 'end of
temporality' (as a rebuke to Francis Fukuyama's designation of 'the end of history'): a situation that “is faced
by postmodernity in general and to which its artists and subjects are obliged to respond in a variety of ways”
(Jameson 2003: 708). In terms of the multitude of artistic reflections on contemporary everyday life,
Jameson, interestingly, suggests that the effects of late capitalism on lived experience are best articulated in
terms of temporality rather than spatiality. It is in this context of a 'temporal articulation' of the world that we
might understand the work of contemporary slow filmmakers: that is, in their production of excesses of
55
Jameson also suggests that, in the passage from the modern sovereignty of imperialism to contemporary
globalisation, “what could not be mapped cognitively in the world of modernism now slowly brightens into the very
circuits of the new transnational cybernetic. Instant information transfers suddenly suppress the space that held the
colony apart from the metropolis in the modern period” (Jameson 2003: 701).
129
duration that flaunt late capitalism's shrinkage of time, as well as the immobility and 'wasteful drifts'
characteristic of acinema. The field of slow cinema, then, not only reflects on but is produced by the system
of conditions of the postmodern era: a cultural symptom generated as “a structural effect of the temporality
of our socioeconomic system” (Jameson 2003: 718).
As Hardt and Negri suggest in their designation of “today's real enemy,” the paradigm shift in
capitalist power represented by neoliberalism comprises the world system of the postmodern era, and,
therefore, the driver of the set of conditions from which slow cinema emerges (Hardt and Negri 2000: 137).56
Neoliberalism decisively arises between 1978 and 1980 as a direct consequence of the systemic turbulence
felt by the capitalist world-economy in the early 1970s.57
According to Giovanni Arrighi's mapping of the
historical cycle of major capitalist developments in his seminal study The Long Twentieth Century: Money,
Power and the Origins of Our Times (first published in 1994), the structural crisis faced by the world
economy in the 1970s forced an “authentic cessation of expansion” in global trade and production—one that
had to be met by neoliberalism's shift to financial intermediation and speculation (the consequences of which
can be measured today by the competitive temporality of the derivative) (Arrighi 2010: 240).58
Under
neoliberalism, dominant commercial cinema's intensified continuity style reflects changes in media and the
circulation of capital by accelerating the production of image-sound situations, in turn blurring the
boundaries between the filmic image and the visual language of advertising (as Daney criticised, at an early
phase, in his analysis of L'Amant [1992]).
Post-continuity style, in Shaviro's conviction, signifies a mode of commercial film that might finally
break through the neoliberalisation of image-sound production, by actualising the spaces and flows of capital
in its most formidably mobile form.59
Post-continuity film represents only a marginal faction of
contemporary dominant cinema, however, and appears to be increasingly defined by a shrinkage of time that
56
David Harvey suggests that the neoliberalisation of the capitalist world system can be interpreted “either as a
utopian project to realise a theoretical design for the reorganisation of international capitalism or as a political
project to re-establish conditions for capital accumulation [after the systemic crisis of the 1970s] and to restore the
power of economic elites” (Harvey 2005: 19).
57
This period signifies Deng Xiaoping's first wave of market reforms in 1978, followed by the transatlantic election of
Margaret Thatcher and Ronald Reagan in 1979 and 1980 (see Harvey 2005: 39-63).
58
For a similar contextual summary of neoliberalism's emergence from post-war late capitalism, see Duménil and
Lévy 2011: 10-22.
59
The present systemic crisis that began in 2007 (with the onset of the U.S. subprime mortgage crisis, now in its
second phase with the Eurozone sovereign debt and balance of payments crisis that began in 2009), appears to have
both consolidated and intensified the process of neoliberalisation yet further.
130
Evan Calder Williams describes as an evacuation of “any sense of weight that might mark a difference
between one possibility or another,” and in which narrativity constitutes “a hot glassy wind of irrelevant
unmaking and particle effects [...] that at least knows how to puncture holes in everything, even if the holes
are barely compelling and hardly ragged” (Williams 2011).
Slow cinema signifies an apparent immobilisation of this process on the level of the temporality of
narration and the shot itself, and displays a radical deceleration that is capable of manifesting the aesthetic of
acinema. On the one hand, a significant number of slow films have turned their attention to the everyday
lives and 'invisible' places that late capitalist commercial films tend to ignore (such as in the examples cited
earlier in this chapter: Alonso's La libertad, Liu's Oxhide I and II, and Tsai's Vive l'amour) or opt to codify
into subjects for advertising and kitsch. On the other hand, however, slow cinema finds itself marginalised by
the very socioeconomic conditions of which it is a symptom. The onset of the neoliberal turn in the 1980s
represented, as Adrian Martin comments, a particularly “unusual and difficult phase in the adventure of
modern cinema,” in which modernism's formal experimentation and political militancy (in the work of, for
instance, Miklós Jancsó, Glauber Rocha, Nagisa Oshima, or Jean-Luc Godard and Jean-Pierre Gorin's
Groupe Dziga Vertov) was “successfully muffled by market forces all over the globe.” Martin singles out
only Chris Marker's Sans Soleil (1982) as an indicator of how elements of modern cinema developed in the
immediate post-modernist period, and suggests that the essayistic mode of Marker's film, with its
incorporation of emergent video and communications technologies, at least signalled that “the barriers
between documentary and fiction [and] between objective reportage and imaginative speculation” were
being comprehensively redrawn (Martin 2010).
Thom Andersen, in a recent essay on what he refers to as the “neo-neorealism” of Jean-Pierre and
Luc Dardenne's recent cinema (in particular, Lorna's Silence [2007]), directly connects a manifestation of the
time-image in contemporary realist cinema to the emergence of neoliberalism (Andersen 2007: 26). Echoing
Deleuze's conception of modern cinema, Andersen suggests that contemporary time-image films, after being
muffled by the market forces that Martin mentions above, reflect a loss of belief “that our actions (both
individual and collective) can respond to the demands of our situation [in the world] and change it.”
Andersen speculates that this loss of belief (a deeply modern condition) was brought about by the political
impasse of the Cold War and the global installation of neoliberalism's 'real enemy' “beyond the reach of
131
democratic institutions” after the collapse of authoritarian socialism. Andersen suggests that the form of
Jean-Pierre and Luc Dardenne's films respond to this “triumph of neoliberalism” with “a scrupulous
examination of its workings on a human level,” displaying an insistence that “action is character” by
attempting to restore belief in the possibility of effective human (individual and collective) agency
(Andersen 2007: 27). Many slow films, however, appear to retreat from this mode of realism, insisting
instead on observational detachment (and even a removal of human action), and a return to a cinema of
'seeing' (in Deleuze's term) where vision occupies the place of effective agency and action.
In the 1980s, the nascent tendencies of contemporary slow cinema were represented by art and
experimental filmmakers that survived the folding of modernism by continuing to foreground strategies of
undramaticness, literal duration, and the long take: Chantal Akerman's Toute une nuit (1982), Theo
Angelopoulos' Voyage to Cythera (1984) and Landscape in the Mist (1988), Philippe Garrel's Elle a passé
tant d'heures sous les sunlights (1985), and Straub-Huillet's Trop tôt, trop tard, Klassenverhältnisse, and
Cézanne: Dialogue avec Joachim Gasquet (1989). Films such as Béla Tarr's Kárhozat (1987) indicate the
direction that would be taken by slow cinema in the 1990s (primarily by Tarr's next film and Kelemen's
Frost); and the trope of observational landscape film is represented in this period by Peter Hutton's Budapest
Portrait (Memories of a City) (1986) and Landscape (for Manon) (1987), as well as James Benning's
Landscape Suicide (1986) and Used Innocence (1989).
In later post-modernist cinema, certain hyperbolic aspects of this last gasp of modernist practice are
systematically modified or hollowed out: the tense hyperrealism of Akerman's Jeanne Dielman, 23 quai du
Commerce, 1080 Bruxelles is relieved of its psychological crisis, the sedate ennui of Akerman's Meetings
with Anna is distended yet further, and the dense metaphysical materialism of Tarkovsky's Stalker is
exhausted of its transcendental quality. Recent durational films drain the interplay between form and content
in these works to a point of extreme rarefication—from the systematic construction of causal tension in
Jeanne Dielman, only literalness, referentiality and stasis is extracted, and the defining aspects of
Tarkovsky's modernist style are, as implied by Jonathan Rosenbaum's description of Tarr's cinema, wholly
'despiritualised' when appropriated in the era of contemporary slow cinema.
In this transition to a post-modernist era, Kovács suggests that the key aspect of modernism that is
abandoned is its expression of “nothingness”—that is, “the only verifiable reality behind the surface of the
132
empirical world.” Modernism, as an artistic, literary and cultural movement, is defined by a replacement of
the (classical) sense of knowable empirical reality by a process pointing to its lack, and its ideological project
draws to a close when the paradigm of an expression of 'nothingness' disappears (Kovács 2007: 395). With
this assumption of an absolute closure to the modernist era in mind, I would like to briefly focus on a recent
intervention that challenges Kovács' emphasis on modernism's total demise as a conclusion to this chapter.
Post-modern slow cinema and the persistence of parametricity
In a valuable article on the endurance of aspects of modernist practice after the onset of postmodernism
(entitled “Beyond Europe: On Parametric Transcendence” [2010]), Mark Betz attempts to account for the
persistence of certain stylistic devices by reviving and extending David Bordwell's concept of parametric
narration (see Bordwell 1985: 274-289). Although Bordwell's analysis of parametricity focusses on
modernist works such as Robert Bresson's Pickpocket (1959), Alain Resnais' Last Year in Marienbad (1961)
and Jean-Luc Godard's Vivre sa vie (1962), the concept itself is not historically confined to modernism. Betz
proposes that Bordwell's analysis can be productively expanded from a consideration of a few 'fugitive films'
in order to shape a broader notion of a “parametric cinema” that indicates “a properly historical
phenomenon” (Betz 2010: 38).
As it refers to a narrational system that privileges order and stylistic patterning over meaning and
narrative development, Bordwell's concept of parametricity is particularly useful in the context of slow
cinema. In parametric films, as Bordwell notes, the narration “literally stylises the represented events” so that
“images and sounds stand like translucent filters between the syuzhet [plot] organisation and the spectator”
(Bordwell 1985: 292). A parametric film will always be fully unified by a set of intrinsic norms that are open
to patterned reiteration, and Bordwell identifies two strategies by which the narration will operate: an
'ascetic' option, in which the film limits its range to a narrow range of codified procedures (such as in
Yasujiro Ozu's cinema, to which I shall return in my next chapter), and a 'replete' option, in which each
stylistic 'event' “exploits a wide range of paradigmatic procedures” (Bordwell 1985: 285).
Betz proposes a renewed focus on parametricity in order to counter an insistence that modernist
practice is 'over' in contemporary cinema, as well as Kovács' assumption that modernism should be situated
133
primarily within a “European sphere of influence.” Betz suggests that an insistence on modernism's terminal
demise is “counter-intuitive” in relation to the aesthetics and maturation of the 'festival' and 'slow art' film,
which instead appear to signify “a modernism that seems quite healthy and hale” (Betz 2010: 39). I am in
broad agreement with Betz's efforts to establish a continuum of the modern beyond the era of modernism
here, and particularly with his attempt to free analysis of modernist film from “the strictures of a historical or
geographic stagism” by emphasising the degree to which “historical time is palimpsestic and dispersive in all
cultures [and] how aesthetic forms may be translated across cultures in multiple circuits of exchange and
appropriation” (Betz 2010: 40). This approach permits, as in Deleuze's model, a consideration of modern
cinema that extends far beyond its geographic roots in post-war Italian neorealism and European art cinema.
Betz loosely ascribes parametricity to a number of filmmakers that I am focussing on in this thesis,
including Alonso, Apichatpong, Hou, Jia and Tsai, as well as, curiously, contemporary art filmmakers such as
Wong Kar-wai, Hirokazu Kore-eda and Abderrahmane Sissako (Betz 2010: 41). My only reservation with
Betz's proposition here is that Bordwell is quite particular about the presentational rigours of order,
specificity and repetition in parametric films, and rules out filmmakers whose narrational styles are not
decisively patterned. This would exclude, for instance, the wildly diverse and unpredictable narration in
Wong's work (such as in Fallen Angels [1995], Happy Together [1997] and 2046 [2004]) or Kore-eda's
adoption of fairly standard minimalist art cinema techniques (except, perhaps, in the austere Maborosi [1995]
and his disciplined visual homage to Ozu, Still Walking [2008]). As Colin Burnett has illustrated, parametric
narration is fully in operation in recent durational film in rarefied long-shot long-take works such as Hou's
Flowers of Shanghai (1998) (see Burnett 2004), but expanding Bordwell's distinction beyond the
fundamental basis of its own parameters is perhaps problematic.
In terms of the aesthetics of slow cinema, Tarr's Sátántangó (1994) is an exemplary instance of
parametric film. In Sátántangó, Tarr adopts a system of stylistic 'repleteness' that not only adheres to but
systematically patterns an unconventional set of formal devices. Every scene in Sátántangó is stylised by a
recognisable and broadly predictable set of variables: the frequent use of the sequence shot (usually
bookended by a substantial pre- or post-action lag); decelerated camera movement that tracks laterally,
vertically, horizontally or in an arcing or circular motion; the immobile set-up and fixed frame; variations
between extreme-close-ups and 'cosmic' long shots; and a repetition of planimetric compositions. Within this
134
expansive framework, Tarr retains a number of consistent diegetic aspects that remain unaffected by the
patterning of style: a consistent slow pace of camera movement; the periodic evacuation of activity from the
frame; inexpressivity of performance; and a delayed disclosure of narrativity within shots and sections of the
syuzhet. This, in turn, draws attention to the operation of a durational style that is ostensibly derived from
modernist practice (for instance, what Tarr has referred to as the “real cinema” of Miklós Jancsó's mobile
long-take style), and produces stretches of temps mort that permit a contemplation of the material content of
the image (G. Andrew 2007: 19).
To adequately account for the continuum of the modern in contemporary cinema, then, it is perhaps
also necessary to return to more than the operation of stylistic devices: that is, to the modern problematic of
the uncertainty of vision and a need to restore a dialogue with the world described by Gabriel Josipovici, or
to Deleuze's conviction of cinema's ability to 'think' and restore belief in the world (which I shall return to in
my next chapter). As the function of modern film is to conduct this dialogue between humans and the world
(through either realist or anti-illusionist, or narrative or non-narrative forms), we might say that the
foundational aspects of modern cinema will invariably persist in their aesthetic endeavours until the modern
condition changes.
As Jameson himself has suggested, this conception of a continuum of the modern could override the
category of the postmodern. In his essay “The Antinomies of Postmodernity,” Jameson notes that the
historical break between modernism and postmodernism could logically be repudiated so that any works
associated with the postmodern would be “assimilated back into classical modernism proper.” Hence the
postmodern might become “little more than the form taken by the authentically modern in our own period,
and a mere dialectical intensification of the old modernist impulse toward innovation” (Jameson 1998: 26).
This, I would suggest, is a useful way to consider the broader framework of the emergence of slow cinema
(alongside its position in the hypermediated culture of the neoliberal era), as well as its development of a
particularly rigorous and adaptable acinematic style.
135
CHAPTER FOUR
THE STILL LIFE, TWO TRIBUTES TO OZU, AND THE DIGITAL REGIME
The following chapter consists of four broad sections. The first section concerns an aspect of slow cinema
that derives from Gilles Deleuze's consideration of the 'still life' in Yasujiro Ozu's cinema, whilst the second
and third sections look at two exemplary contemporary films that pay tribute to both the use and tone of this
device in Ozu's work: Hou Hsiao-hsien's Café Lumière (2003) and Abbas Kiarostami's Five: 5 Long Takes
Dedicated to Yasujiro Ozu (2003). The final section, which is inspired by the unique production and aesthetic
of Kiarostami's film, looks at the role of digital technology in contemporary durational cinema.
To introduce these sections, I will first elaborate on Deleuze's model of modern cinema, and its
introduction of concepts that are integral to the operation of contemporary observational cinema: the timeimage
and the 'still life'.
The creation of concepts
Deleuze's conception of modern cinema represents both an addition to, and radical break from, the Bazinian
model of an aesthetic transformation in realist style.60
This radical break, for Deleuze, occurs with a shift in
the production of duration in global film practice after the Second World War, as a response to the traumatic
dislocation in the human relation to the world in the immediate post-war period. In positing this radical
break, the Deleuzian model appears to be contiguous with conventional definitions of modernity that, as
Bruno Latour observes, always “point, in one way or another, to the passage of time.” Latour reminds us that
the adjective 'Modern' always designates “a new regime, an acceleration, a rupture, a revolution in time” that
disrupts the precedent of an “archaic and stable past”—here, of a dominant industrial mode of production
that is interrupted by new forms of post-war art and experimental film practice. As Latour suggests, the
assignation of Modern “is thus doubly asymmetrical: it designates a break in the regular passage of time, and
it designates a combat in which there are victors and vanquished” (Latour 1993: 10). The victor, in both the
60
For an excellent analysis of both the convergence and difference between Bazin and Deleuze's models of modern
cinema, in relation to the evolution of the durational long take, see Beasley-Murray 1997.
136
Bazinian and Deleuzian model, is a new production of duration, and Deleuze's model introduces a set of
concepts that work to extend the formal attributes of Bazinian realism, as well as reflecting on modern
cinema's radical means of 'imaging' time.
Deleuze's model of modern cinema does not derive from a classification of formal or narrational
devices, as in Bazin's theory, or from traditional distinctions of 'classical', 'art' and 'avant-garde' film.
Deleuze's approach to cinema is philosophical rather than critical or theoretical, and is based upon the
assumption that cinema thinks with images, and thus creates concepts. These concepts, as Deleuze suggests
at the very end of his second volume on the medium, “are not given in cinema, and yet they are concepts, not
theories about cinema” (Deleuze 2005b: 269). As Paola Marrati has observed, for Deleuze “philosophy and
cinema are both creative practices” in which “philosophy creates concepts, [and] cinema creates images.”
Hence when cinema encounters philosophy, it does so with a conceptual practice unique to the medium, in
which “creating new paths of thought is the shared task and ambition of philosophers and filmmakers” alike
(Marrati 2001: 227).
I am invoking Deleuze's philosophical model here for two key purposes. The first is to address slow
cinema's direct relation to time and stillness, and the way it produces duration and a sense of immobility
through conceptual devices such as the 'still life'. The second relates to contemporary slow cinema's acute
demonstration of what Deleuze identifies as the medium's primary capacity: to think, by filming, our relation
to the world. In Deleuze's philosophy, the Modern condition is, at root, a lack of faith: one in which we
moderns “need to believe in the world,” for “our problem is not the absence of a God but instead our absence
from this world” (Marrati 2001: 227). Marrati suggests that this lack stems from a lack of belief in the
possibility of creating new forms of life: a belief that is “a matter neither of knowledge nor of representation
but rather of a conversion of thought” (Marrati 2001: 228). The power, and task, of cinema is to embody this
conversion as a matter of faith, and I would suggest that in the field of slow cinema this is sought in
strategies that are primarily realist. The dominant tendencies of slow cinema appear to merge Bazin's vision
of cinema as a tool that can restore our ability to adequately observe the world (in all its ambiguity) with
Deleuze's conception of its creation of new paths of thought. In placing a unique emphasis on the time and
objects of life and nature, it could be argued that slow cinema manifests a potential for restoring, to some
degree, our belief in the world.
137
One of the most important aspects of Deleuze's philosophy in relation to the evolution of slow film is
his consideration of everydayness in the work of Yasujiro Ozu. In Ozu's films, the formulation of an
“everyday banality taken as family life in the Japanese house” serves as a 'container' for new modes of
undramatic narrativity (everyday image-sound situations) and the time-image, in a device that Deleuze refers
to as the 'still life'. Although Ozu's cinema spans the period between the last century's two irruptions of
cinematographic modernity—from the 1920s to the 1960s—it is rarely classified, unlike the other primary
subjects of Deleuze's volumes on film, as strictly modernist. Deleuze suggests that Ozu's radical innovations
were not imitated in the immediate post-war era by European modernists, but, in a tradition that is continued
by slow cinema, filmmakers “came back to him later via their own methods” (Deleuze 2005b: 13). In light of
the centrality of Ozu's work to modern cinema, it seems appropriate to consider Deleuze's analysis of Ozu
here, as well as two contemporary films that derive from his cinema: Hou's Café Lumière (2003) and
Kiarostami's Five: 5 Long Takes Dedicated to Yasujiro Ozu (2003).
Kiarostami's tribute to Ozu points to another key aspect of slow film that is expressed in Deleuze's
model: the development of a purely contemplative mode of image-sound situation. Deleuze's notion that
cinema is able to 'think' the world illustrates this contemplative tendency more satisfactorily than Bazin's
realist model, and also provides a conceptual base from which to approach Kiarostami's minimalist attention
to objects and their duration. In this respect, my approach to Deleuze's philosophy of cinema is intended to
provide a framework to address a few core aspects of style and content shared by a group of particularly
rigorous art and experimental films, rather than to enter into a philosophical reflection on cinema's creation
of concepts.
The Deleuzian model of modern cinema
Deleuze divides film history into two broad regimes of the movement-image and the time-image, which
differ in their oppositional thinking and 'imaging' of time. In this model, the montage-based cinema of the
pre-war movement-image era (including the “great organic unity” of the American classical system, Soviet
dialectical montage, and German Expressionism) is drawn from “an indirect image of time” in which relative
or absolute movement (Bergsonian durée) is selectively divided into objects, then reunited to form a whole
138
that is comprehensible (Deleuze 2005a: 31). Montage is a critical intermediary in this process, and
movement-images represent mobile sections ('cuts' or 'blocks') of duration that produce the movement
(durée) from which image-sound situations and narrativity flow. Hence the movement-image regime is an
assemblage of coherent chronological relations where “perception is organised in obstacles and distances to
be crossed, while action invents the means to cross and surmount them”: that is, the ideal basis of classicism
(Deleuze 2005b: 38). Central conflict and continuity style derive from this ordered basis, drawing on a
relation that Deleuze terms the 'sensory-motor schema': the way we coordinate our sensory and motor
faculties in order to meet our needs, purposes or desires, and shape our “commonsense” world. Ronald
Bogue notes that the pragmatism of orderly, linear narrative cinema represents “the practical application of
our perceptions and actions to meet those ends”—a process that is entirely dependent on the effective, and
successful, interconnection of our sensory and motor faculties (Bogue 2003: 66).
The sensory-motor schema of movement-image cinema is irrevocably loosened in the post-war timeimage
regime, which brings about a relation in which “we no longer have an indirect image of time which
derives from movement, but a direct time-image from which movement derives” (Deleuze 2005b: 125). The
direct time-image cuts loose from the assemblage operation of movement-image cinema, and reaches a point
where movement mutates to signify the perspective of time (Deleuze 2005b: 21). As only weak sensorymotor
connections survive in the time-image regime, 'false' continuities and impulsive, aberrant movements
(potentially analogous to those in Lyotard's conception of acinema) are produced, and bring about 'pure
optical and sound situations' that reveal empty spaces “disconnected from the commonsense coordinates of
their standard usages and practices” (Bogue 2003: 109). This process takes a variety of forms during the
evolution of modern cinema, ranging from the 'organic' regime that produces the “instances of pure
contemplation” in Ozu's cinema, to the 'crystalline' regime of high modernist hallucinatory topography in
Alain Resnais' Last Year in Marienbad (1961) or Alain Robbe-Grillet's L'Eden et après (1970) (Deleuze
2005b: 15). It should be noted here that the key dynamic of the montage cinema of Resnais, Robbe-Grillet
and Godard arises, as Denis Lévy notes, from the fact that it is “no longer discursive, or rhetorical; its
coherency is no longer composed of sequences, but of conglomerates around a void.” These interstitial voids,
introduced by causal and temporal discontinuity, are “opposed to the plenitude of Realist [classical] editing,
which fills voids” (Lévy 2011). This 'irrationality' of the montagist operation therefore demonstrates that “it
139
is necessary to make a division or make emptiness in order to find the whole again” (Deleuze 2005b: 21).
Deleuze does note that, in the post-war period that leads to the emergence of slow cinema, only a
minority of time-image films attempt to wholly decompose montage in order to manifest time in the form of
the long take or sequence shot. In most cases, active montage remains the principal cinematographic act, and
has enabled, in the false continuities of modernist practice, the emergence of “all possible movements” from
cinema's emancipated sheets of time (Deleuze 2005b: 126).
Key aspects of slow cinema derive, in part, from Deleuze's notion of the post-war “cinema of the
seer and no longer of the agent,” in which characters in fiction films become spectators of new realities
where they are able to see and hear but not respond to their diegetic worlds. These seers record instead of
reacting, as if “prey to a vision, pursed by it or pursuing it, rather than engaged in an action” (Deleuze 2005b:
3). Roberto Rossellini's Viaggio in Italia (1953) and Michelangelo Antonioni's 'great tetralogy' (from
L'Avventura [1959] to Il deserto rosso [1964]) represent the modernist root of this tendency, with the latter
opening, as Jacques Rivette famously observed in his “Letter on Rossellini” (originally published in Cahiers
du Cinéma in 1955), a breach “that all cinema, on pain of death, must pass through” (Rivette 1985: 192). In
this way, modern cinema replaces the faltering sensory-motor situations of the movement-image regime with
sets of pure optical and sound situations, creating a cinema in which spectators (both in and of the film) enter
into what Rivette refers to as an “impassive web of duration” where the only action is vision (Rivette 1985:
198).
The most prominent contemporary analogues of this cinema of the seer include Costa's Colossal
Youth (2006), Jia's Still Life (2006), Alonso's Los muertos (2004) and Liverpool (2008)—all films that render
a visual (and sonic) impression of the modern disconnection between man and an uncertain, intolerable
world. The protagonists of these films (Ventura in Colossal Youth, Sanming and Shen Hong in Still Life,
Vargas in Los muertos and Farrel in Liverpool) appear to not be themselves in the situations in which they
find themselves, and act as if barely able to function in an everyday to which they can no longer respond.
They move through pure optical and sound situations where, as Deleuze describes, “the intolerable is no
longer a serious injustice, but the permanent state of a daily banality” (Deleuze 2005b: 164). In these
situations, thought has become unthinkable, and the link between man and world, people and objects, has
broken down. All the characters can do is look: their bodies become weary, which delays and drags out their
140
movements, and strings of image-sound situations depict activities of staring and waiting, and states of
fatigue and despair. Here, as in the images of looking and lingering in Antonioni's work, slow cinema shows
“not the drama of communication, but the immense tiredness of the body” (Deleuze 2005b: 183).
What gives rise to this type of modern cinema, then, is the modern fact that “we [can] no longer
believe in this world” (Deleuze 2005b: 166). The solution to this problematic, for Deleuze, is the restoration
of a link between man and world as an object of faith, as “only belief in the world can reconnect man to what
he sees and hears. Hence the cinema must film not the world, but belief in this world, our only link.” The
task of modern cinema might then be summarised as: the discovery and restoration of belief in the world
“before or beyond words” (Deleuze 2005b: 166; my emphasis). The primary potential that cinema has for
renewing our belief in the world lies in the time-image, and formal strategies employed by modern
filmmakers can be said to actualise the capacity to film this link. Perhaps tellingly, some modern filmmakers
have spoken of their cinema in ways that express this potential quite clearly—for instance, Andrei
Tarkovsky, in reference to Stalker, reflects that: “In all my films it seemed to me important to try to establish
the links which connect people (oilier than those of the flesh), those links which connect me with humanity,
and all of us with everything that surrounds us” (Tarkovsky 1987: 192). The method of establishing and
thinking the nature of these links varies throughout modern cinema, but one strategy in particular endures in
recent durational film.
The 'fullness' of the time-image, and the still life
An important component of slow cinema derives from what Deleuze refers to as a particular 'fullness' of the
filmic image in time. This type of time-image occurs, firstly, in the work of Yasujiro Ozu (both before and
during the post-war modern era), and, in an alternative mode, in the landscape films of Jean-Marie Straub
and Danièle Huillet. Although Ozu's cinema will be my primary focus here, it will be worth considering
Straub-Huillet's intervention first, as their work has served as an important foundation for the realist
operation of the long take (and observation of landscape) throughout the last four decades of durational
cinema.
Ozu and Straub-Huillet's work is connected by the way in which they submit the filmic image to
141
what Deleuze describes as a “stratigraphic reading” (Deleuze 2005b: 237). In this reading, the imagination
and memory needed to read a seemingly 'empty' image (one that is devoid of human or causal content)
produces a new analytic in which our act of perception, over the course of a duration, reverses the status of
the image and “constantly converts the empty into full” (Deleuze 2005b: 235). In Ozu's cinema, this
reversion occurs in 'still lifes' that depart from narrative relations in order to produce “pure and direct images
of time” that fully transform empty and disconnected spaces into cumulative time-images (Deleuze 2005b:
17). In Straub-Huillet's cinema, the unbroken static and panning shots across landscape that are emblematic
of their films (from Les Yeux ne veulent pas en tout temps se fermer, ou Peut-être qu’un jour Rome se
permettra de choisir a son tour (Othon) [1969] to Straub's O somma luce [2010]) work against 'emptiness' to
reveal, in an apposite temporal field, “the strata of the visual image” (Deleuze 2005b: 237).
In Straub-Huillet's work, Deleuze proposes that the visual image becomes “archaeological,
stratigraphic, tectonic” through a presentation of “the deserted layers of our time”—that is, in the filming of
places chosen for their 'coalescence' of history and the present (Deleuze 2005b: 234). The sequence shots of
Straub-Huillet's landscape films frequently decouple the image from both narrativity and human presence,
and display documentary records of the earth (often sites of political struggle) accompanied by sound
recordings of a literary text or historical document. This structural contingency is employed throughout, for
instance, Trop tôt, trop tard (1981), where a series of extended takes (in one instance lasting the entire length
of a 35mm reel) display the depopulated, rural locations in France and Egypt cited by three political texts
recited on the soundtrack.61
Equally, in Dalla nube alla resistenza (1979), Straub-Huillet document a number
of bucolic sites were blood was spilt on the soil in anti-imperialist struggle, such as a cornfield in Piedmont,
Italy where anti-Fascist resistance fighters were massacred during the Second World War.62
Huillet has spoken of this layering of separate elements as a critical facet of their films, and indeed
uses the term 'archaeology' to describe the density of materiality and time recorded by the images of
61
There are three texts recited on the soundtrack in Trop tôt, trop tard: a letter from Friedrich Engels to Karl Kautsky
(sent on 20th February 1889), extracts from the Cahiers de Doléances (written by village mayors on the eve of the
French Revolution in 1789), and extracts from Mahmoud Hussein's book Class Conflict in Egypt, 1945-1970 (first
published in 1973).
62
Another primary instance of this practice in Straub-Huillet's work is Toute révolution est un coup de dés (1977), in
which nine speakers (including Huillet herself) read Stéphane Mallarmé's poem Un coup de des jamais n'abolira le
hasard (1897) by the wall by which 147 defenders of the Paris Commune were executed in Père Lachaise cemetery
on 28th
May, 1871. The film consists of three similarly 'concrete' elements: the text, the speech act, and a visual
impression of the material traces of history.
142
landscape in their work. In a 1976 interview, Huillet follows Straub's suggestion that “I think what interests
us is to show layers” by stating that their aesthetic intent is always “not to eradicate traces, but to build on
them” (qtd. in Byg 1995: 66). Hence, Straub-Huillet's sequence shots possess a descriptive function that aims
to expose the geology of the image: to show, spatially and temporally, that “the earth stands for what is
buried in it.” In this way, Deleuze suggests, Straub-Huillet's images of landscape are distinguished by the
principal fact that “an empty space, without characters (or in which the characters themselves show the void)
has a fullness in which there is nothing missing” (Deleuze 2005b: 235; my emphasis). Such a sustained
concentration on place reveals layers of duration and history, and an attendant politics, that can only arise
outside of the 'coherency' of sensory-motor relations emblematic of the movement-image regime. StraubHuillet's
unpeopled images, anchored by their documentary transcription, have often developed powerful
political meanings, and Mark Fisher notes that the stratigraphic capacity of the time-image in their work is
representative of an expansive 'social stratification' that forms “an expression of the way in which both
human populations and the earth are shaped by vast impersonal processes.” In revealing a fullness of the
material and temporal layers of the filmic image, Straub-Huillet's work shows that “the unpeopled is
therefore not the same as the empty” (Fisher 2010: 50; my emphasis).
In Yasujiro Ozu's work, it is the attention to the everyday that irreparably weakens the sensory-motor
connections of the movement-image regime. This everydayness is drawn from a basis of understated
narrativity that, as Ozu himself wrote of Late Autumn (1960), seeks to portray character and incident “by
eliminating all the dramatic devices [...] to make people feel like what life is like without delineating all the
dramatic ups and downs” (qtd. in Schrader 1972: 19). As in Straub-Huillet's films, Ozu dedicates time to
spaces without characters or drama, frequently cutting away to interiors evacuated of human presence, and
equally deserted, wholly calm exteriors (which, in Straub-Huillet, do at least contain a quickening pulse of
the wind in the trees, or bounteous patterns of light shifting on the ground). In these cutaways, humans and
their fictions are replaced by configurations of household objects, sections of built environments and natural
landscapes (forested mountains, shorelines, the sea, the sky) that signify the 'real diversions' conventionally
prohibited by central conflict theory. These 'still lifes' are best summarised as optical and sound images that
no longer extend into action, a distinction that I shall retain here. It is also worth noting that Noël Burch, in
his canonical study of Japanese cinema To the Distant Observer (1979), directly interprets the concept of the
143
Fig. 32: There Was a Father (1942)
still life as a “pillow-shot,” a term that proposes a loose analogue with the 'pillow-word' of classical Japanese
poetry (in short: a conventional epithet or attribute for a word that occupies a short five-syllable line in order
to modify a word in the next line) (Burch 1979: 160).63
In Ozu's body of work, the still life signifies a fixed, largely inanimate tableaux that pauses or defers
narrative activity to foreground instead a variety of places, objects and sights: unpeopled rooms or corridors;
light sources (exposed bulbs, table lamps, street lights and neon advertising signs); liquid receptacles (teakettles,
multicoloured glasses, beer bottles, shallow bowls—see [Fig. 32] for a typical example from There
Was a Father [1942]); laundry hung from lines; and icons of industry or communication (modern office
blocks, smokestacks, electricity pylons, radio towers). As Deleuze notes, Ozu's attention to this range of foci
for the undramatic image is purely observational, and free from “an apparent value which is relative (in
63
Burch notes that the meaning of a 'pillow-word' might be known or not known, and is often ambiguous. He is also
mindful to add that the adaptive designation 'pillow-shot' was never an “entirely adequate” term (Burch 1979: 160).
144
relation to a story) or consequential (once the action is done with)” (Deleuze 2005b: 15).
In terms of narration, the parametric system employed by Ozu to contain these images is, as David
Bordwell has observed, “highly self-conscious,” and favours ellipses and decentred narrative relations over
classical linear clarity (Bordwell 1988: 70). Ozu's narrational system frequently elides events and 'flaunts'
gaps in its storytelling, letting us notice and puzzle over disruptions when they occur. Cutaways to still lifes
serve to reflect and visualise these intentional gaps in storytelling, but also operate on a more radical level
that is, I would suggest, adopted later by slow cinema. The still life is particularly prevalent in Ozu's late
silent films (most notably, from Tokyo Chorus [1931] to A Story of Floating Weeds [1934]), yet, during his
mature work of the 1950s and early 1960s, Ozu's placement of the images begins to shift their relation to
narrativity. In the late sound films, the device becomes, as Noël Burch notes, “less specific, less about a
tension between the suspension of human presence (of the diegesis) and its potential return,” and instead
marks a point of disconnection with places or objects explicitly marked out by the narration (a tendency that
commonly structures the placement and order of pillow-shots in earlier films such as An Inn in Tokyo [1935])
(Burch 1979: 161). As Burch suggests, during Ozu's late films, the still life gains a greater autonomy as a
pure instance of contemplation, and achieves what Burch refers to as “a satori-like suspension of meaning”
(Burch 1979: 172).
In time-image cinema, the dissolution of sensory-motor relations and suspension of diegetic flow
enables the opening up of a field external to the diegesis, in turn creating new contemplative possibilities for
the image. Key aspects of post-war slow films begin to emerge from this basis, for instance in the production
of temps mort that allows time for reflection, or a tendency to linger on objects or natural landscapes during
long takes to reverse the status of the image from 'empty' to 'full'. Deleuze suggests that it is the radical
stillness and immobility of objects in Ozu's compositions that transforms the temporal dimension of the
image. Writing of a sequence in Late Spring (1949) that depicts Noriko and her father Somiya on holiday in
Kyoto, Deleuze describes the construction of an image-sound situation that takes place after both characters
retire to bed. In a series of shots, Ozu cuts, twice, from Noriko's face and back again to a vase in a different
part of the room, positioned centrally in the frame below a silhouette of bamboo branches buffeted by the
wind, and illuminated by soft light leaking through the shoji [Fig. 33]. Deleuze suggests that, in the montagecut
between Noriko, the vase and back again (after Somiya has fallen asleep), Noriko's emotional state
145
Fig. 33: Late Spring (1949)
visibly alters, and the still life of the vase is “interposed” to make visible a “becoming” in the passage of time
in and between the shots. This little stretch of “time in its pure state” emerges from the dialectic between the
duration of the object in the frame (“the representation of that which endures”), and the “succession of
changing states” between shots, where the concentration on the still object within the shot serves to
“represent the unchanging form of that which moves, so long as it is at rest, motionless” (Deleuze 2005b:
16). Enabled by the temperate relation of the montage-cut, images of immobilised objects such as the vase in
Late Spring embody blocks of duration that “bring the emancipated senses into direct relation with time and
thought” (Deleuze 2005b: 17).
The establishment of this direct connection with thought, enabled by a stillness and duration that
manifests a space to think, is a critical development in the emergence of slow cinema. It is worth
remembering that, in Deleuze's conception, modern cinema is primarily based on a “system of exchange
146
between the imaginary and the real” (Deleuze 2005b: 9), where the real and the imaginary become
indivisible. This system of exchange between the mental (thought) and the physical (the world) is a
prerequisite for a contemplative mode in representational art. In Ozu's singular attention to objects and their
duration, cinema, perhaps for the first time, is given over entirely to a contemplative act, and thus to a
renewed ability to think the world. In this way, Ozu's work attempts, on a foundational level, to reconnect
man to what he sees and hears, and his effort is upheld by more contemporary filmmakers.
As I will explore in the following sections, two exemplary films are particularly significant in this
context: Hou Hsiao-hsien's Café Lumière (2003) and Abbas Kiarostami's Five: 5 Long Takes Dedicated to
Yasujiro Ozu (2003). Both were produced to mark the centenary of Ozu's birth in 1903, and represent two
distinct aspects of recent slow cinema. Hou's narrative film was originally commissioned as one segment of a
three-part portmanteau film (extended to feature-length when the other two contributors, Aki Kaurismäki and
Wim Wenders, dropped out of the project), and is the first film Hou shot outside of his native Taiwan. As a
direct homage to Ozu's work, Café Lumière was filmed entirely on location in central and suburban Tokyo
with a local crew, accompanied by a core unit of Hou's regular collaborators (writer Zhu Tianwen, director of
photography Mark Lee Ping-Bing, editor Liao Ching-song), and a few professional and non-professional
actors.
Kiarostami's film, on the other hand, is a non-narrative observational piece, shot solely by
Kiarostami in a few locations over the course of several months, near the Caspian Sea in Northern Iran. Five
was filmed, like Kiarostami's previous experimental work Ten (2002), on rudimentary digital video, which
enables both a more solitary, unhurried filming practice and an option to vastly elongate the duration of
individual shots. As Kiarostami comments in Ten's companion film 10 on Ten (2004), Five shares the former
work's singular emphasis on the structural quality of the film (which, in Ten, comprises a restrictive frame of
a dashboard-mounted camera and an iterated measure of shots) at the behest of any ostensible subject or
story. From these divergent bases, Hou and Kiarostami's films take on, and modify, two primary elements
from the modern cinema of Ozu described above: respectively, an undramatic selection of images and
narrative events in which “everything is everyday!” and a radical renewal of the device of the 'still life',
which is given over to “instances of pure contemplation” over periods of substantial duration (Deleuze
2005b: 15).
147
Ozu and Café Lumière
Hou and Kiarostami both adopt minimal stylistic systems that modify principal aspects of Ozu's cinema and
the concept of the still life. Café Lumière, by modifying Ozu's cinema on the level of form and narrativity,
belongs to a particularly attenuated strain of contemporary art cinema, whilst the relation of Kiarostami's
experimental film to Ozu's work is primarily conceptual, and belongs to the taxonomy of durational
landscape films outlined by Francisco Algarin Navarro in my first chapter. Both of these modes dispose of
the montagist operation that encloses the still life in Deleuze's analysis of Late Spring—a displacement that
is contiguous with the stylistic transition from late modernist practice to contemporary slow cinema, and its
elimination of effects derived entirely from montage. Whereas Kiarostami's film comprises a set of five
isolated still lifes, Café Lumière discloses fictional narrativity seemingly in addition to the states of stillness
signified by the pillow shot, and creates an impression that its events (and non-events) emerge organically
from 'empty' frames rather than sequences designed solely for the efficient disclosure of narrative activity.
This is a contradictory effect to that established by Ozu's cinema, in which the still life tends to act as an
instance of contemplation by detaching itself from other expressive elements within a highly stylised
parametric system.
The typical parametric system in Ozu's work, as specified by David Bordwell's definition, is quite
determinedly “ascetic,” and limited to a narrow range of codified procedures unified “by a set of identifiable
intrinsic norms” (Bordwell 1985: 285). Within these boundaries, as Andrew Klevan observes, any “shots
which do not ostensibly further the plot are, in fact, crucial to the film's narrative,” and mostly serve to offset,
and reflect on, human incident rather than disrupting a perceptible narrative pattern (Klevan 2000: 144). This
suits the highly ritualised nature of everyday life portrayed by Ozu's films, and is loosened in the slackened
fiction, and less rigid narrational form, of Café Lumière.
It is worth noting that this interpretation of Ozu's narration is open to another that is perhaps better
suited to Hou's tribute: Paul Schrader's suggestion that the silences and voids of Ozu's cutaway pillow-shots
form the narration's primary operative element, rather than narrativity or human incident. In his influential
text Transcendental Style in Film, Schrader compares Ozu to a “traditional Zen artist,” and suggests that the
codas and still lifes in his films embody the first koan of Zen: mu, the concept of negation, emptiness and
148
void. Schrader suggests that whereas “in Western art one would naturally assume that the codas are inserted
to give weight to the paragraphs, for Ozu, as for Zen, it is precisely the opposite: the dialogue gives meaning
to the silence, the action to the still life” (Schrader 1972: 29). In an echo of Deleuze's distinction of the
fullness of the time-image, Schrader notes that silence and stillness in Zen art signify presence rather than
absence, and that as “mu is the character used to refer to the spaces between the branches of a flower
arrangement,” 'emptiness' becomes the operative part of the narration (Schrader 1972: 27).
It is also worth noting here that, although I am drawing an active comparison between aspects of
Hou and Ozu's work, I do not wish to fall into the trap of the 'colonial encounter' that Gary Needham has
identified, and criticised, in relation to Western critical discourse on Hou's work. In an incisive article
published after the release of Café Lumière, Needham outlines how European and North American critics,
when straining to locate nominal similarities between Hou and Ozu's work, attempt to smooth out any
questions of “cultural, ethnic or national specificity and identity” that inevitably arise. This approach, in
Needham's estimation, tends to result in crass definitions of a film style “that can be seen to represent a
fictive and localised 'oriental cinema',” without much consideration of historical reality or the nuances of
Hou's evolving style (Needham 2006: 374-375). The willingness of Western critics to seek out 'oriental'
equivalences based on visual signifiers (the stillness of framing, or a similar appearance or presentation of
domestic space) of course has its risks, and Needham outlines one in detail: the identification of a visual link
between locations in Ozu's work and the Japanese (or, perhaps more appropriately, 'Japanised') exterior and
interior architecture, and the presence of tatami mats and shojis, in Hou's films of the Taiwanese New
Cinema era. The presence of 'Japanised' architecture and domestic furniture functions as a reminder of the
cultural and architectural changes that were enforced upon Taiwan under occupation by Japan between 1895
and 1945, but have been mistaken by some Western critics (such as, in Needham's summary, Alan Stanbrook
and Godfrey Cheshire) as indicating a continuum of some sort of unified 'oriental aesthetic' (Needham 2006:
371-372).
Needham, in opposition to such problematic critical tendencies, repeatedly makes the point that
Hou's stylistic intent is quite radically different to Ozu's, not least in his total transformation of montage
relations in Ozu's work. Whereas Ozu cuts into the spatial and temporal order of scenes within a rigorous
parametric system, Needham notes that Hou “prefers to preserve the natural order of space through editing
149
restraint, long takes, and strategies of camera placement that all suggest a detached observation in as real a
time as possible for the event being filmed” (Needham 2006: 379). It is here, in Hou's employment of
sequence shots to preserve the duration of the undramatic field in time, that the broader tendency toward an
attenuated Bazinian realism in contemporary film regains prominence. Hou's style, in this way, obliterates a
linear relationship with its parametric precursor: as Needham observes, “Hou is concerned with what is
taking place within the shot and not with the relations between the shots” (Needham 2006: 380). My
proposition here is that Ozu and Hou's cinema are linked only in terms of their conceptual relations within
modern cinema (in terms of stillness, everydayness and temporality), and by Hou's explicit tribute to Ozu's
work in Café Lumière. In this context, Needham's attempt to stress the other considerable differences in form
and cultural specificity between both filmmakers' work is particularly instructive: “If anything, Café Lumière
reveals to us through Hou's vision of Tokyo, lingering shots of tempura restaurants and character silences, a
reversal of what Ozu might have looked like if Hou had influenced him” (Needham 2006: 382).
Alongside its transformation of montage relations in Ozu's cinema, Café Lumière's primary departure
from Ozu's style concerns a dynamic that Andrew Klevan locates in Late Spring: the fundamental
maintenance of “the rigour of a still frame in order to register the slightest movement from its actors”
(Klevan 2000: 151). This immobility of the frame is stringently maintained throughout Ozu's work (with
only a few exceptions of camera movement, most notably in earlier films such as Dragnet Girl [1930], and,
in single instances, later works such as Early Summer [1951]). Hou, however, freely alternates between static
camera set-ups and mobile long takes: in Café Lumière, the camera will retain either a still position or pan,
languorously, both toward and away from characters without clear direction, often in a single shot. Hou's
style is not concerned with the inflexible application of the still frame, but rather of a method of observation
that continues to register the slightest motion within the visual field (whether the camera is lightly mobile or
not).
Hou sets up this stylistic tendency in the second shot of Café Lumière. After a fixed opening shot in
which a tram passes through the frame from left to right, a gradual fade reveals Yoko (the film's primary
character: a young writer played by pop star Yo Hitoto in her first and, to date, only film performance)
hanging laundry out to dry on her apartment balcony [Fig. 34]. As she calmly performs this everyday
activity, the wind lightly blows into the apartment through the open window, ruffling her hair and clothes.
150
Fig. 34: Café Lumière (2003)
Initially the camera is still, but when Yoko moves across the interior of the room to answer a phone call it
responds to her movement by fully panning to the left, with a slight delay, and back again on her return.
Yoko then moves offscreen (through the kitchen) to open the door to her neighbour (from whom she will
later borrow a bottle of sake in a direct reference to a scene in Ozu's Tokyo Story [1953], in which Yoko
performs the same task as Noriko in Ozu's film to the considerable embarrassment of her mother). While
Yoko is offscreen the frame remains fixed like a still life, inviting us to turn our attention to the array of
visible objects: a rotating fan, the shape of the curtain, and the motion of the laundry blowing in the breeze.
After a minute Yoko re-enters and resumes hanging up her clothes, whilst continuing to talk on the phone,
and the shot ends after three and a half minutes. The rest of Café Lumière then employs stylistic variants on
this observational method, most notably in the two extended sequences in Hajime's bookstore, and the casual
reframings during interior scenes at Yoko's father and stepmother's house.
Perhaps the most significant dynamic within the second shot of Café Lumière is the integration of the
device of the still life within an extended analysis of the undramatic field in time. In this sequence shot, the
still life forms part of a broader temporal continuum of human activity within a domestic space, rather than
being sectioned off by a montage-cut in order to create the effect of a succession of changing states. There is
151
no attempt by Hou to render an emotional effect by introducing montage relations, and the parameters of the
still life are somewhat softened, becoming expressive of duration in the manner of what Deleuze referred to
as “the representation of that which endures” rather than the more dynamic state induced by a montage-based
transition (Deleuze 2005b: 16).
Still lifes and objects
The use of the long take to display undramatic activity in the second shot of Café Lumière is emblematic of
the film's overall style, and I will return to its system of undramatic narrativity, and Hou's particular
employment of the long take, shortly. Hou's integration of the dynamics of the still life into a different
narrational system is indicative of a broader tendency in contemporary durational film, which radically
departs from the spectatorial practice encouraged by classical and intensified continuity cinema. David
Bordwell notes that, as archetypal spectators of a classical or continuity film, when we are “given a narrative
mandate, we’re on the lookout for pictorial factors that affect the dramatic situation.” But when narrativity is
purposely slowed by a filmmaker who is disinterested in a classical organisation of the image, we might then
enter into “a visual search that isn’t wholly driven by plot considerations” that, in turn, heightens the
visibility of objects not otherwise activated by narrative. Bordwell suggests that Ozu has been uniquely
influential for contemporary filmmakers in this respect, by drawing attention to objects with careful
inflections of light, colour and composition that aid the spectator's visual search external to narrativity. In the
still life, Ozu's careful focus on objects and their composition means that “even though he cuts rather fast,
we’re given time to see everything” (Bordwell 2011).
Bordwell remarks that a similar technique has been adopted by Béla Tarr and other contemporary
slow filmmakers, but suggests that in their work such visual guides are often “more a matter of texture and
tactile qualities” (which, in Tarr's films, is provided by the tactile properties of “mud, oily puddles, dusty
corners, and tearing winds”) (Bordwell 2011). Throughout Café Lumière, there is a distinct echo of Ozu's
configurations of household objects in domestic scenes, but, tellingly, in Hou's film such objects are
disclosed casually, in a manner far less mindful of their aesthetic effects. In Ozu's work, we are frequently
reminded that the image's task is to bring out, as Deleuze suggests, the thing in itself, literally, in its excess of
152
beauty or quietude. But in the light sources and domestic clutter of Café Lumière—just as in Hou's
Millennium Mambo (2001) and Flight of the Red Balloon (2007)—domestic objects appear in ornamental
states that are seemingly accidental, routine and banal. In Café Lumière, the objects that catch our eye—the
damp clothes that Yoko hangs from the balcony of her apartment; the diagonal line of half-empty glasses and
a flask in Yoko's parents' home; a strikingly neat line of bottles at a restaurant counter; or the patchwork
blanket slung up by Yoko to dim the afternoon sun—appear to simply be present, lent their concreteness by
the duration in which they remain on screen, elusive of narrative import.
As Bordwell implies, however, other contemporary filmmakers are more active in adopting Ozu's
attention to objects and their textural qualities, and it is worth considering a couple of examples here. Tarr's
Sátántangó (1994) offers an abundance of objects and tactile patterns not activated by narrativity, in still
images that last minutes at a time (usually featuring sporadic human activity), or slow tracking shots that
accumulate detail by recording the surface of material phenomena (reminiscent, above all, of analogous
mobile long takes in Tarkovsky's Stalker [1979]). Tarr's parametric narration is designed to enable the display
of things and objects that appear as a surplus to narrativity, impelling the spectator to contemplate the density
of a world that is, to borrow a concept from Victor Shklovsky, systematically made strange.
In Sátántangó, things and objects are disclosed in a visual itinerary that heeds one of Shklovsky's
determinants of the purpose of art: “to recover the sensation of life” and “impart the sensation of things as
they are perceived and not as they are known” (Shklovsky 1965: 12). Tarr lingers, for exhaustive durations,
on images of the sodden and weathered brickwork of external walls; patches of peeling plaster in damp
interiors; damp, sagging net curtains; wilted plants in dank rooms; bulbous gatherings of condensation on
windows; parched exercise books and overflowing ashtrays; stacks of just-washed pint glasses in a bar; piles
of sodden leaves in woods; and clods of earth on well-worn boots. The visual search for sensation that is
activated by Tarr's attention to objects here—that is, for the experience of the artfulness of the object—
overwhelms any other diegetic significance that the objects themselves might possess.
In Tarr's cinema, image-sound situations are designed to prolong the perception of the fullness of
things in themselves. Shklovsky's note on art's impediment of perception in his essay “Art as Technique”
(originally published in 1917) is useful here, as he suggests that “the greatest possible effect is produced
through the slowness of the perception,” and that art should encourage a “lingering” spectatorial practice in
153
Fig. 35: Still Life (2006)
which any representational object will be “perceived not in its extension in space, but, so to speak, in its
continuity” (Shklovsky 1965: 22).
Jia Zhang-ke's Still Life (2006), by comparison, is structured by a figurative series of objects, divided
into four thematic parts that appear as intertitles: 'cigarettes', 'liquor', 'tea', and 'toffee'. Still Life's title, as Jia
has noted, is a reference to its connotation in painting rather than to Ozu, but a number of the film's cutaways
do operate in a similar manner to the device in Ozu's work. This strategy is particularly evident in, for
instance, the images of a shelf of abandoned belongings in a deserted factory locker, or the curious
composition of a pocket watch, alarm clock and set of wristwatches strung up across an interior wall in a
house belonging to one of the characters (Wang Dongming) [Fig. 35]. In an interview published in Cahiers
du Cinéma in 2007, Jia suggested that these “banal” objects expressly embody the film's thematic content,
foregrounding the visible traces of “a life that is no longer there, or rather of a way of life about which there
remains only the most basic evidence.” The attention to simple objects such as cigarettes, wine and food acts
as an intermediary for attempts to reconnect with the world, before and beyond words: as Jia observes, “it is
through such things that people communicate no matter what, that they sense being together even without
talking” (qtd. in D. Andrew 2010: 56).
154
Still Life, like the first half of its documentary counterpart Dong (2006), was filmed in a place of
intense dislocation: the town of Fengjie, during its dismantlement and relocation as part of the construction
of the Three Gorges hydroelectric dam project.64
When Jia filmed Still Life and Dong in 2005, the twothousand-year-old
town was at an advanced stage of demolition, in preparation for its future submergence by
the rising water level of the Yangtze River. The subsequent completion of the dam barrier has resulted in the
flooding of thousands of towns and settlements, and official figures released by the CPC register the
displacement of 1.13 million people from affected areas.65
Jia has described the rapidity of the destruction of
Fengjie during this process as “simply unimaginable,” and constructs Still Life's narrative diptych to render a
sense of the collective experience of such an acute disruption of population and place (Lim 2008). The film's
narrative focuses on the search of two individuals, Sanming and Shen Hong, for their partners, both of whom
have disappeared during the town's relocation. In a manner typical of time-image cinema, Sanming and Shen
Hong's search is enacted principally through the activity of looking, and is structured by a series of imagesound
situations in which they see the changes in their immediate environment, yet are unable to respond to
the new world that is signified by those changes.
Perhaps more so than any other contemporary film, the rubble-strewn landscapes of Still Life recall,
in purely visual terms, the wreckage of post-war Italian neorealism—in particular, the streets of Naples in
Rossellini's Paisà (1946), and Berlin in Germania anno zero (1948)—and point to the very roots of the timeimage
regime. Jia films the ruined environments of Fengjie in a series of extreme long shots, and draws on
Straub-Huillet's visual strategy of the slow, expansive pan across landscape (here, that is, of the Yangtze
valley and the stark gradient of the townscape), tracing the arc of its vanishing past and devastated future. As
in Straub-Huillet's work, these images develop a stratigraphic quality, in which the “silent piling-up” of
layers of time display the material evidence of dislocation (the landscape soon to be submerged), and the
struggle of the people to survive it (Deleuze 2005b: 245). In this way, Jia formulates a time-image cinema
that is able to depict the actual 'event' of the film: the manifestation of a shifting state of time, in a place
64
The construction of the dam barrier was completed just after filming in 2006, and the structure has been fully
operational since May 2012.
65
See, for instance, Yardley 2007. Jia's diptych of Still Life and Dong is perhaps best viewed as a counterpart to a
number of independent documentaries that address how the dam's construction has affected the lives of citizens
caught in its path (due to the flooding of the valley and disintegration of the surrounding landscape): see, in
particular, Yan Yu and Li Yifan's Yan Mo (2005), Feng Yan's Bing Ai (2007) and Yan Yu's Yan Mo II—Gong Tan
(2008).
155
whose future is already its past.
Hou's style and the undramatic long take
Hou's adoption of a distinctive long-shot long-take style in Café Lumière and his preceding body of work is
shared by other Taiwanese filmmakers of the pre- and post-New Cinema era, most notably Edward Yang and
Tsai Ming-liang. Of these three filmmakers, Emilie Yueh-yu Yeh and Darrell Davis have observed that Hou's
style appears to be the most “Bazinian,” tending toward a “haptic view of perception” that attempts to
“sustain relations between screen and spectator that are perceptually closer to the phenomenological real”
(Yeh and Davis 2005: 100). This approach produces both quite conventional realist effects and a more radical
aesthetic of delay across the broad base of slow cinema, which Yeh and Davis suggest is due to the coming
forward of “positive and negative aspects” of time in the image. The long-shot long-take aesthetic, in their
summation, promotes either “an impression of discovery and contingency (a sense of 'could have been
otherwise'), while an uncomfortable sense of pure waiting may translate into meditative states of
detachment” (Yeh and Davis 2005: 101).
Adrian Martin suggests that Hou's long-take style is defined by a strategy of selecting “a single,
central position from which to take in an event—sometimes a static camera position, sometimes more of a
zone within which the camera will track, pan and reframe” (Martin 2008: 266). This 'zone' to which Martin
refers is described by Yeh and Davis as providing a “convex” observational field, and is carefully
predetermined by Hou to enable the sustained observation of aleatory human activity (Yeh and Davis 2005:
101). When setting up the spatial conditions for (in)activity to unfold, Hou loosens restrictions on the filming
of scripted events that would, in classical and continuity cinema, possess narrative primacy, and at times
simply lets things take place in front of the camera (which either records from a fixed position, or hesitantly
responds, or overlays, with movement). Hou's formal mediation does, of course, stylise the representation of
what is taking place, but it also signifies an aesthetic attempt to surrender to reality in both spatial and
temporal terms—a strategy that once inspired Fredric Jameson to comment that realism, if defined as an
absence of subjective style, “can only be attained by the deep shot and the surrender of eventfulness to its
contents, rather than to the machine that records them.” With this acceptance of what has become a core
156
precept of realism in slow cinema, Jameson suggests that, “perhaps, in this sense, Bazin was right after all”
(Jameson 1997: 83).
As Genevieve Yue discloses, after a conversation with Ichirô Yamamoto (one of Café Lumière’s four
producers), during the filming of Café Lumière Hou was apparently content to simply observe the behaviour
of his actors rather than actively direct them, and “only ever called 'cut' at the end of a reel of film, or after
eleven minutes' worth of footage, where a lot could and did happen.” Yue reveals that Café Lumière was
filmed without any rehearsals, and Hou instead relied on “an openness to the outside world and the generous
space given to the actors to move about and inhabit their roles” (Yue 2008). Like similarly patient,
'unproductive' modern and contemporary filmmakers such as Straub-Huillet, Pedro Costa and Albert Serra,
this strategy left Hou with an extensive amount of footage—180,000 feet of 35mm film—from which to
extract the series of long, uninterrupted takes (with an ASL of 66.7 seconds) that constitutes the final version
of the film.
The method by which Hou constructs image-sound situations is symptomatic of this Bazinian
precept: the staging and recording of a spatial and temporal continuum whenever possible (or desirable).
This unwritten rule is clearly expressed by Theo Angelopoulos, who, when interviewed in 1974 whilst
editing The Travelling Players (1975), claimed that
the basic principle governing all the film is the sequence shot, whether the camera is moving (which
it is most of the time) or immobile. This way, the scenes gain much in depth and detail, with the
editing being done inside the camera. We never shot two scenes, if we had the option of doing it in
one (Angelopoulos 2001: 21).66
As Hou's staging of the aforementioned second shot in Café Lumière illustrates, spaces and (non-)events that
continuity style would have to record separately for the purpose of narrative efficiency are here permitted to
flow into one another, or to run their course without interruption.
Although I would argue that the fundamental principle holds for both filmmakers, it is worth noting
66
This sentiment is echoed by an epigrammatic comment that Jean-Luc Godard made regarding Jean Renoir's Eléna et
les hommes (1956): “He achieves in one shot what others do in ten; and where they make do with one, Renoir can
do without. Never has a film been so free” (Godard 1968: 64).
157
that Hou's technique is markedly more understated than the sequence shots in Angelopoulos' work, and that
the strategies for editing 'in camera' in their films serve very different purposes. In Angelopoulos' cinema, the
intersection of spaces and events within the same shot primarily serves to visualise a sense of historical flux,
and, more specifically, the breakdown and re-establishment of political order in Greek society. Angelopoulos'
late modernist style is consistent throughout his mature body of work (which spans Days of '36 [1972] to The
Dust Of Time [2008]), and includes such strategies as the negotiation of spatial transitions by switching from
the narrative path of one individual character to another in a single tracking movement, or halting in a space
after a narrative situation has dissipated until a subsequent event is staged to take its place. At times,
Angelopoulos extends these strategies to cover, quite audaciously, sweeping historical shifts within the
spatial field of a single 'implausible' shot. For instance, in one sequence shot in The Travelling Players, after
the action of a fascist group dispersing into the streets on New Year's Eve in 1946 has ended, Angelopoulos'
camera holds back and, after a brief period of temps mort, begins to track toward an approaching election
rally for Alexander Papagos (then soon to be elected Prime Minister), revealing that we have been transposed
to 1952 seemingly without missing a beat.67
Both Hou and Angelopoulos' use of the long take, regardless of ambition, recalls Brian Henderson's
notion of “intra-sequence cutting” in his critique of Bazin's theory of film style. Henderson argues that, in the
sequence shot, “rhythm is achieved not by the lengths of shots themselves (even where multiple), but rather
within each shot, through movement—or lack of it—by camera, or both.” An intra-sequence cut enables a
rhythmic or spatial shift without severing temporal relations through montage, and serves to break the
rhythm of the sequence in order to “reconnect it on a new basis” (Henderson 1980: 54). This sense of
rhythmic design, manifested in the interstice between immobility and movement in Hou and Angelopoulos'
films, is also shared by a few other modern and contemporary filmmakers. Béla Tarr is a master of
consolidating mobility and stillness, and frequently pairs lengthy tracking movements with static set-ups that
render actions in interchangeable devices (as I suggested at the end of the last chapter, in relation to the
parametricity of Tarr's cinema). In Sátántangó, this often happens in two stages: for instance, after following
a character's approach from a distant point toward the camera (either in a static long shot that observes from
an extreme distance, or a mobile long take that trails close to their body), the camera will reverse its position
67
For a fuller description of this sequence shot and similar strategies in Angelopoulos' work, see Mitchell 1980.
158
Fig. 36: Café Lumière (2003)
for the character's inevitable retreat into the distance by merely selecting the opposite observational stance (a
fixed frame or tracking shot) to that first adopted. Tarr deploys a series of variations on this strategy
throughout Sátántangó, and exploits their adaptability to create a parametric system that is as slyly playful as
it is formally complex.
David Bordwell, in the most comprehensive study of Hou's staging and style yet published (see
Bordwell 2005: 186-237), suggests that the patience and stillness of the long takes in Hou's work are
designed to let new locations “register initially as a space, not a container or background for well-defined
narrative action.” In what might be described as a pre-action lag, the spectator of Hou's films is consistently
encouraged to “take in the vastness of a landscape or the details of an interior without yet knowing how it
links to a larger story rhythm” (Bordwell 2005: 217). A typical example of this occurs in Café Lumière in one
of the scenes in Yoko's father and stepmother's house, in which Yoko wakes after dark after having slept
through the afternoon. The scene comprises two shots, and opens in darkness before Yoko enters the kitchen
and turns on a light. The camera is positioned in the living room, which at first is not lit, meaning that the
space of the kitchen is only visible through the vertical letterbox framing of the narrow doorway [Fig. 36].
Yoko quickly disappears from the frame after entering the kitchen, leaving the image as a still life empty of
159
Fig. 37: Café Lumière (2003)
human presence for almost half a minute. At this point, our visual search of the frame scans the contours of
the objects within view, quickly adjusts to the camera's immobility, and notes a visual similarity between the
pan on the hob and one of Ozu's signature tea kettles.
After Yoko's stepmother enters the room, Yoko pours herself a drink and moves into the foreground
of the living room. Hou racks focus briefly as Yoko sits at the table in front of the camera, and until her
stepmother leaves the kitchen she remains visible in the second plane of a deep focus composition. After a
minute and a half, an uncharacteristic jump cut (the only one in the entire film) elides the act of Yoko's
stepmother sitting down to join Yoko. As they sit and talk, Yoko, with her back to the camera, calmly tells
her stepmother that she is pregnant [Fig. 37]. Her stepmother responds by shifting position almost
imperceptibly (folding her arms and resting them on the very edge of the table), and the second shot ends a
minute and a half later. The sequential dynamic here is best described as one of narrational delay: as
Bordwell notes, it is only quite late in the scene, when Yoko addresses a need (to inform her stepmother of
her pregnancy), that we understand what is transpiring in terms of narrativity. In the meantime, as a result of
Hou's willingness to spare the time necessary to let the space of the scene register before accomplishing its
more conventional function, we are simply “obliged to study the shot in itself” (Bordwell 2005: 218).
160
It is also worth noting here, in relation to Hou's method of filming Café Lumière, a more pragmatic
reason for his adoption of the long take. In the early period of the Taiwanese New Cinema era, Hou opted to
work with non-professional actors (after his brief foray into commercial production ended in 1983), and soon
found that the practical constraints of that ostensibly realist decision began to guide his style. In a 1989
interview cited by Bérénice Reynaud in her monograph on A City of Sadness, Hou stated that “it is difficult
for non-professionals to respond quickly in front of the camera, so the long takes provide them with a chance
to develop their emotions slowly.” To accommodate this slowness of performance, Hou also designed “some
daily life actions—such as washing clothes or eating—for [the actors] to perform, which gives them a chance
to familiarise themselves with the situation” (Reynaud 2002: 28). In working with non-professional
performers in this way, Hou appears to heed a Bazinian dictum: that the actor in realist film should “be
before expressing himself,” and that “the ignorance of theatrical technique [is] less a positively required
condition than a guarantee against the expression of traditional acting” (Bazin 2005b: 65). Both Hou's longshot
long-take aesthetic and attention to everyday activity are rooted in this pragmatic response to actualities
of production, and both are progressively refined as his style develops.
Hou's approach is shared by other modern filmmakers who have dealt with similar practical
contingencies. For instance, Jon Jost, when filming Slow Moves (1975)—a film shot on 16mm in five days
for approximately $8000—purposely used extended takes in order to allow his actors to improvise and
collaborate without worrying about the efficiency of their performance. As Jost recalls, “I wanted to get
towards working with actors in ways that leave them a lot of free space in order not to be stuck to a script
and to be able to have time to develop their acting trajectory” (Lehman and Triplett 1980: 27-28). This
approach was aided by an elimination of dramatic narrativity, and both the filming and editing of Slow
Moves became, as Jost comments, “an experiment in getting rid of the stuff we know makes a narrative work
—the famous, beautiful star, a tight plot, adventure, the order of it, its emphasis—and seeing if there’s
something left that holds the audience anyway” (Lehman and Triplett 1980: 32).68
More recently, Tsai Ming-liang has spoken of the manner in which, in the early period of his working
relationship with Lee Kang-sheng (which now spans eleven films, from Boys [1991] to Visage [2009]), Lee's
68
Jost used the same method, equally successfully, for Last Chants for a Slow Dance (1977), an important modern
film that consists of only seventy-five shots (with an ASL of 67.7 seconds), and a narrative conceit that plays out
like a patriarchal sequel to Akerman's Jeanne Dielman, 23 quai du Commerce, 1080 Bruxelles.
161
'slowness' of performance initially appeared to be a major obstacle during filming. Tsai comments that
“whenever I asked him to do something, I would always have to wait for it to sink in, so there was always a
momentary lag. It was really strange, so I kept rushing him and telling him to hurry up.” However, on
learning to accept Lee's natural movements, Tsai began to adjust the rhythm of both production and form in
his films, which led to an attenuation of his long-shot long-take style: “His actions may be slow, but why do I
always have to get him to hurry up? Why can't I slow down to wait for him?” (M. Berry 2005: 378-379).
After Vive l'amour, Tsai also began to narrativise Lee's physical slowness by designing conditions
that impede and delay his performance of activities. The River (1997), The Hole (1998), Goodbye, Dragon
Inn (2003) and I Don't Want to Sleep Alone (2006) all contain central tropes or scenarios that depict the
obstruction of simple tasks, or the performance of arduous (or routine) activities that take an interminable
length of time to complete. In The River, Lee contracts a mysterious virus from filthy water (whilst shooting
a film with Ann Hui) that leaves him crippled and unable to turn his head for the rest of the duration of the
film. This disablement gives rise to one of the film's best visual absurdities: a touching moment when Lee's
father perches behind him on his motorbike, cradling his head to keep it upright (the only way that Lee is
able to move at speed). Likewise, in The Hole, a constant torrential downpour and freakish interior flooding
makes it difficult for Lee and Yang Kuei-mei to navigate walkways and claustrophobic apartments, and the
opening sequence of I Don't Want to Sleep Alone depicts the slow and painstaking activity of migrant
workers dragging a flea-ridden mattress down the streets of Kuala Lumpur. Tsai also adds individual
sequence shots to his films that are configured to display unusual or decelerated movements: for instance, in
one poignant shot in The Wayward Cloud (2005), Lee and Chen Shiang-chyi waltz across a pedestrian bridge
in Kaohsiung, with Chen standing on Lee's feet to slow their progress. Tsai films the activity in an extreme
long shot that displays the entirety of their journey in both spatial and temporal terms, and, of course, only
cuts when they have managed to traverse the entire bridge.
To add another example, the image-sound situations of Goodbye, Dragon Inn, in particular, are
almost entirely comprised of delayed movement due to the centrality of Chen Shiang-chyi's character: a
ticket seller who walks with a heavy limp. Goodbye, Dragon Inn takes place in the rundown Fu-Ho Theatre
in Taipei during its final screening (of King Hu's Dragon Gate Inn [1967]) before closure, and depicts the
traversal of space by a few characters who spend the venue's last night simply seeing out time: eating,
162
Fig. 38: Goodbye, Dragon Inn (2003)
shifting seats, slouching and cracking nuts, cruising, (not really) urinating, and occasionally watching the
projection of Hu's film. In the five sections of the narration that are given over to her movements, Chen
struggles around the large, seemingly labyrinthine building, visiting the projection booth and completing
tasks such as positioning buckets under a leaking roof. The longest of these sections lasts over six minutes,
and they culminate with a static long shot of the empty theatre (after the projection has ended) that fixes on
Chen's awkward mobility for five and a half minutes.
In this static shot, Chen, after the house lights come on, enters from the right of the frame and climbs
halfway up one of the theatre's two central aisles, using a broom to aid her balance [Fig. 38]. Each of the
seventeen steps takes her two to three seconds to ascend, and her journey up, across a row and down again
takes three minutes (including a pause to sweep up some detritus). After Chen reaches ground level again,
she exits to the left of the frame, and it takes about forty-five seconds for the echoing sound of her footsteps
to dissipate. The final minute and a half of the shot is taken up by a significant post-action lag: a largely
silent still life of the moribund cavernous space of the theatre. Tsai's insistence on presenting Chen's torpid
movements in these rarefied sequence shots (described, as Yeh and Davis note, by many critics as
“intolerable”) serves to further hyperbolise the film's dearth of narrative activity—which, as Tsai has
163
accurately stated, could “be told in two sentences” (see Yeh and Davis 2005: 237; Reichert and Syngle
2004). As a result, Goodbye, Dragon Inn has an extended shot length of 54.5 seconds, and the durational
realism of Chen's physical slowness forms its central narrational conceit.69
Narrativity and de-dramatisation in Café Lumière
Hou's tribute to Ozu in Café Lumière turns to contemporary Tokyo for one principal reason alongside
location: Ozu's modern cinema was always of the present. Whereas Hou's most sophisticated long-shot longtake
films cover historical epochs that open up a lost or censored past—namely, A City of Sadness (1989),
The Puppetmaster (1993) and Flowers of Shanghai (1998)—his more recent films, after the rarefied genre
exercise Goodbye South, Goodbye (1996), have marked a rediscovery of the contemporary that is otherwise
most prominent in his early work. In the New Cinema era, Hou's feature-length films, from The Boys from
Fengkuei (1983) to Dust in the Wind (1986), focussed on largely contemporary, provincial subjects: rural
boredom, geographic displacement from the country to the city, agricultural and low-paid manual labour,
familial and group solidarities, and a desperate resort to crime. From this period, the gentle, undramatic A
Summer at Grandpa's (1984) is most reminiscent of Ozu's cinema (in particular, the lively I Was Born, But...
[1932]) in its portrayal of children as both individuals and in groups, but the equivalence is purely
coincidental—as Needham reminds us, Hou had not seen a film by Ozu until 1989 (Needham 2006: 370).70
In Millennium Mambo (2001) and the third section of Three Times (2005) (titled A Time for Youth),
Hou turns to the alienation, disaffection and muted hedonism of life for a young and affluent class in
contemporary Taipei. The form of these films is slick and speculative, centring on a mobile camera that floats
across aesthetic surfaces, stylising the narrative's sense of perpetual drift. When preparing Millennium
69
For comparison, the longest ASL in Tsai's films (thus far) is 70.9 seconds in I Don't Want to Sleep Alone. The
measure of shots in Goodbye, Dragon Inn is comparable to The River, The Hole and What Time is it There? (2001),
which have ASLs of 55.3, 50.4 and 65.1 seconds, but it is worth noting that Goodbye, Dragon Inn contains a
sequence of rapid montage (when Chen Shiang-chyi is watching a combat sequence in Hu's film) that substantially
distorts the final metric. Vive l'amour and The Wayward Cloud clock in with shorter durations of 36.2 and 34.5
seconds respectively.
70
Later, in Harold Manning's documentary Métro Lumière: Hou Hsiao-hsien à la rencontre de Yasujirô Ozu (2004),
Hou does profess a fondness for the “sense of movement and taste for everyday life” in I Was Born, But.... However,
in its reliance on the observational long shot and a distinct slackening of causal relations, A Summer at Grandpa's
might be more productively placed in the context of Hiroshi Shimizu's films about children, in particular Children in
the Wind (1937) and Four Seasons of Childhood (1939), with which it also shares a couple of close narrative as well
as visual similarities.
164
Mambo, Hou immersed himself in contemporary culture—as he recalls, hanging out “with these kids for two
years, going to all these clubs and bars, trying Ecstasy myself”—but the film deliberately mutes, in a manner
characteristic of contemporary slow cinema, the rhythm of its subjects' existence (Ellickson 2002: 18). Hou's
camera retains an observational distance and foregrounds a floating, hazy affect that inadvertently reflects on
the compression of temporality in contemporary life—as Hou recalls, “the casting director of Millennium
Mambo, who is about the same age as the actors, told me that the pace of their lifestyle is so quick that there
really is no way to catch them” (M. Berry 2005: 263).
The everydayness of Café Lumière is more subdued, however, and depicts a short period in Yoko's
life in suburban and central Tokyo. The urban space depicted in the film, as Shigehiko Hasumi notes,
“includes none of the bustle of government and business in the downtown areas, none of the city's
skyscrapers, and none of the neon signs of the entertainment districts” (Hasumi 2008: 193). The narrative's
focus, as in Ozu's cinema, is structured by the rituals and relations of family and friendship, rather than the
disaffected listlessness of Millennium Mambo or A Time for Youth. In a manner similar to Hirokazu Koreeda's
less candid recent tribute to Ozu, Still Walking (2008), Café Lumière is organised around a few events
common in Ozu's work: an extended visit to the family home, a funeral, and a visit to a cemetery. More
uncharacteristically, its core theme centres on a revelation of pregnancy, an incidence that only forms a
central strand of Ozu's plots in Tokyo Twilight (1957). Hou's narrative otherwise comprises a slight inventory
of things that happen (in line with Deleuze's proclamation that “everything is everyday!”) over a linear yet
unspecified time scale: Yoko visits her companion Hajime at his bookstore; travels to stay at the family
home; carries out research for her book on a Taiwanese composer (Jiang Wen-Ye); hosts a visit from her
parents (whilst her father attends a colleague's funeral); and rides trains with Hajime, who records and
catalogues their sounds. Between these 'pivotal' events lies another layer of the quotidian: drinking coffee,
brief conversations with friends, oversleeping in the morning and night, and travelling from place to place on
public transport. The film spends an extended amount of time both on and observing trains, recording their
sonic rhythms and tracking their movement in measured pans across the cityscape (pans that, in their
slowness and precision, recall Straub-Huillet's stratigraphic images of the curvature of landscape).
Hou's extended shots of trains (and trams) provide both a literal sense of the continuity of everyday
life, as well as a direct homage to Ozu's work. In Ozu's films, the most dynamic images are captured by a
165
camera jutting out of a carriage window, as in Late Spring or The Flavour of Green Tea Over Rice (1952),
but in Café Lumière the impressions of movement mediated by Hou are deliberately more subdued. Hou
films the motion of trains in a variety of ways: through a density of bodies on platforms, in abstractions of
cityscape glimpsed through windows, and as containers of shifting patterns of natural light. In a similar
manner to the extended opening sequence of Clare Denis' recent loose adaptation of Late Spring, 35 Shots of
Rum (2008), the characters' silent experience of time spent in transitory motion provides the pure optical and
sound situations that serve to delay orderly narrative action.71
Shigehiko Hasumi notes that the dominance of
these sensuous image-sound situations “fill the work with a strength that forms its own new context,” and
only after their “visual and sensual strength softens is the story gently spun out” (Hasumi 2008: 186).
Adrian Martin suggests that the stories of Hou's films, like those of Tsai and Liu outlined in my
previous chapter, are predominantly structured by 'activities' rather than 'action': that is, the everyday tasks of
eating (often collectively), drinking (often heavily, or through the night), lazing about and wasting time,
washing and drying clothes, singing, brushing hair and applying make-up, and so on. Martin comments that,
in terms of narrativity, it is possible to “disassemble and reassemble several of [Hou's] films not along the
lines of plot, theme or style, but in terms of the system of some activity.” Martin calls these activities “bits of
business,” which normally serve to “fill out” the realism of scenes in classical dramatic structure (reliant on
central conflict theory), as well as supplying “more charged gestures, expressive of character or psychology”
in melodramatic cinema.72
This distinction leads Martin to suggest that “Hou's bits of business become events
and spectacles in themselves,” a rhetorical implication that I would prefer to downplay here (Martin 2008:
266). It is true that the effectiveness of bits of business in dramatic narrativity, as Christian Keathley has
pointed out, entails the coordination of “physical activities with dramatic expression,” but Hou's equivalents
are, instead, designed to be systematically undramatic and unproductive—bits of business that are designed
to fail, perhaps (Keathley 2011: 106).
71
The train sequences in Café Lumière are, in their length and literalness, also close to a visual precedent set by the
fifth shot of James Benning's 11 x 14 (1977): an eleven-minute take from the front of an 'L' train that depicts the
journey from Evanston to downtown Chicago in full. An image in Hou's film bears a striking resemblance to this
shot, which uses all 400ft of a 16mm reel. The shot's extended duration provides a visual scope that renders an entire
urban topography (moving through de-industrialised zones in impoverished black neighbourhoods to inner-city
concentrations of skyscrapers and office blocks), and, as Benning has commented of its length, “as the minutes go
by, you relate to the image differently: at first it seems to be part of the narrative; then it becomes formalistic. That
shot is a key to the whole film” (MacDonald 1992: 233).
72
For an excellent overview of the typical expressiveness of 'bits of business' in relation to a sequence shot in Otto
Preminger's work, see Keathley 2011.
166
In an anecdote about “writing for the pulps” during the early period of his career, Raymond Chandler
illustrates the key characteristic of this notion of bits of business particularly well. In a letter addressed to
Frederick Lewis Allen (written on 7th
May, 1948) Chandler recalls that a purely notative line in a short story
(referring to a shadow falling across a man's face “like the touch of cool water”) was excised by a magazine
editor before publication, on the assumption that its inclusion fatefully delayed the progression of action. In
his future work, Chandler states that he set out to prove this assumption wrong: “My theory was that the
readers just thought they cared about nothing but the action; that really, although they didn’t know it, the
thing they cared about, and that I cared about, was the creation of emotion through dialogue and
description.” Hence the most memorable, and haunting, bits of business prove to be “not, for example, that a
man got killed, but that in the moment of his death he was trying to pick a paper clip up off the polished
surface of a desk and it kept slipping away from him.” Two details are paramount here: both the material
object (“that damn little paper clip”), and the character's powerless gesture: the way “it kept slipping away
from his fingers” (Chandler 1997: 219). This level of attention to detail and gesture is a prerequisite of
realism (in whatever form it might be constructed or mobilised), and Hou's style is organised to observe
(rather than denote) these minor gestures with a rigour that forecloses the creation of emotion—either in
physical motions that stem from the diegesis, or in the registering of a succession of changing states within a
shot.
The critical difference between Hou's narration in Café Lumière and the precedent set by (for
instance) Chandler's literature is that it refuses to extract dramatic or emotional expression from bits of
business. Activities are performed and presented flatly, in a manner seemingly free from emotional or
psychological exposition. In Hou's system, 'key' bits of business include the family's quiet lunch at a
restaurant counter (filmed, obstructively, from behind), Yoko's father distractedly sipping beer in front of the
television, the waiter's delivery of coffee to Hajime's bookstore (played, incidentally, by one of the great
scholars of Ozu's cinema, Shigehiko Hasumi), and Hajime's holding of the microphone to the noisy air on a
station platform. The casual inexpressiveness and long-shot long-take presentation of these activities is very
different to Ozu's work, in which bits of business are often structured to achieve quite unanticipated dramatic
effects. Andrew Klevan has noted that, in Late Spring, Ozu's careful repetition of shots and activities enables
the narration “to build up its drama from movements which are not urgent or energetic: the slow drop of a
167
head, a shuffling with a small object, or simply one character's changed position in relation to another
character” (Klevan 2000: 149-150). Hou's work, however, tends to portray the latter effect whilst draining
away the former: movements are revealed by still or languidly mobile framings in which little but an
unhurried performance of the quotidian prevails.
Martin's notion of the 'filling out' of realism by bits of business is echoed by Roland Barthes'
conception, outlined in his 1968 essay “The Reality Effect”, of the 'insignificant notation' or 'useless detail'
that validates structures of realist narrativity.73
Barthes suggests that the integration of instances of 'concrete
reality' in realist narratives (that is: the description of material details, cursory movements, trivial objects or
transitory attitudes) complements the insistent signification of other elements within the narrative fabric.
When useless details are divulged by the writer or artist, they are designed to be structurally superfluous, and
only contribute to the narrative's broader system of signification by virtue of their 'insignificance'—because,
as Barthes suggests, if everything contained by realist narration signified, any realistic effect would be
undermined by a surplus of referentiality and meaning (Barthes 1982: 15). Insignificant notations form a
central component of realism by pointing to a lack of signification (except of the 'category of the real'), and
create a dynamic in which only “the very absence of the signified, to the advantage of the referent, standing
alone, becomes the true signifier of realism” (Barthes 1982: 16).
In its concreteness and assumed analogical function, the insignificant notation is somewhat similar to
the indexicality of the cinephiliac moment, but its purpose is rather to fortify realist structure rather than
wholly break free from it. In this sense, its function is closer to Barthes' own concept of the 'third meaning' (a
structurally superfluous level of meaning that is no longer symbolic or obvious but obtuse), and Kristin
Thompson's formalist invocation of material 'excess' (see Barthes 1971; Thompson 1986). I would like to
suggest here that, alongside these allusions, the deprivation of signification in Barthes' conception illustrates
the register of realist attention in Hou's work particularly well, and works against the conventional
expressiveness of bits of business in his films—a tendency that will also apply to other filmmakers who work
in an undramatic narrative mode.
Barthes considers two options for recording what he calls insignificant notations: firstly, by the
means of securing the signification of the real by pointing to its lack, as well as, secondly, the potential for
73
Barthes defines 'realism' here as a reference to “any discourse which accepts statements whose only justification is
their referent” (Barthes 1982: 16).
168
recording a literal plenitude of detail. Barthes suggests, with regard to the latter, that the process of notation
is potentially limitless, and “when discourse is no longer guided and limited by the structural imperatives of
the story (functions and signals), there is nothing to tell the writer why he should stop descriptive details at
one point rather than another.” The writer, if permitted, can sustain a visible focus on the real in a manner
that is “inexhaustible by discourse [as] there would always be some corner, some detail, some nuance of
location or colour to add” (Barthes 1982: 14).
This hyperbolically literal approach is reminiscent of the exhaustive surplus of materiality in modern
films that study the fullness of objects and environments, such as the hyperrealism of Chantal Akerman's
modernist cinema, or the compilation of details (edges of furniture, window panes, shafts of light) and
'nuances of location' in Peter Hutton's Florence (1975) and his New York Portrait trilogy (1977-1991). In
these films the extra-narrative rendering of a plenitude of concrete realities might be thought to provide, as
Thompson writes of the notion of 'cinematic excess', “a perceptual play by inviting the spectator to linger
over devices longer than their structural function would seem to warrant” (Thompson 1986: 134).
Hou's recent narrative cinema might be thought to issue a third option: the formation of a narrational
system that is given over almost entirely to the insignificant notation, detached from the expressive
signification of primary elements. In Café Lumière, the film's narrative base appears to have been stripped
out, leaving only the 'useless' details in its wake, glued together by a few inexpressive bits of business.
Narrativity in slow cinema often appears to function in this way, and it could be argued that the device of the
still life itself, in working against signification in Ozu's later work and in Hou and Kiarostami's cinema,
comprises an exemplary 'reality effect' by superfluously notating details of objects or landscapes external to
the conventional function of narration. Equally, Kiarostami's Five, as I will suggest in the next section, could
be considered to be a wholly notative work: a collection of minor concrete realities that have simply been
placed in an appropriately minimalist stylistic container.
This loose system of insignificant notations formulates a distinctively undramatic register of story,
subject and tone. Writing of Goodbye South, Goodbye (1996), Kent Jones describes Hou's approach to
narrativity by asking: “Is there another film since Warhol with a better sense of just hanging out? It's about
how time feels as it's passing, about the feeling of simply existing, moving through life as most people do, no
big deal, caught in a state of being itchy, nervously under the gun” (Jones 1999: 46). Goodbye South,
169
Goodbye, like segments of Chandler's finest work, comprises a tonal experiment in stripping away the action
and anticipation of crime drama from its generic mode, to leave only in-between activities: gambling,
bickering, hiding out, and the act of simply drifting from A to B in cars and motorbikes that snake up
mountain roads or glide at a glacial pace along wide city streets. This “feeling of simply existing” is
replicated in Millennium Mambo and Café Lumière, and is partly achieved by the narration's indifference to
imparting an understanding of how much, or how little, time passes as each films' characters move through
life. Hou, like Ozu, tends to introduce prolonged ellipses between each system of activity, disclosing little of
prior events, and rarely setting up appointments or deadlines in anticipation of future ones. Conflict is almost
totally absent, and causal relations between one scene and the next often disappear along with drama.
As Adrian Martin notes, in the midst of this narrational uncertainty, we often find ourselves “in the
position of thinking we have perhaps missed out on some vital plot information—when that information has,
in fact, been deliberately withheld, often for remarkably extended periods of running time” (Martin 2008:
261). In Café Lumière, for instance, the central plot strand concerning Yoko's pregnancy is revealed at a pace
that borders on inertness: Yoko informs her stepmother that she is pregnant after twenty-five minutes have
passed, and it takes another hour (after a bout of sickness that causes her to rest on a station platform) before
she reveals that she has been pregnant for three months already. These, in turn, are the only concrete pieces
of information about her emotional life that the narration discloses.
This method of de-dramatisation is evident, at an early stage, in Hou's work from the New Cinema
era. Hou's films of the 1980s are primarily autobiographical (drawn from a variety of texts and collaborative
sources), and less attenuated than his more recent work: in Dust in the Wind (1986), for instance, the
undramatic narration and device of the still life appears to function in a manner that is closer to Ozu's
cinema. I will focus briefly on Dust in the Wind here, as it acts as both a summation of Hou's work of the era
and an indicator of core emergent tendencies in the nascent field of slow cinema during the 1980s. Drawing
on an autobiographical script by writer Wu Nianzhen (one of Hou's regular collaborators, who later directed
a key film of the 1990s: A Borrowed Life [1994]), Hou’s de-dramatisation of Wu's original story disappointed
its author. As Yeh and Davis note, Wu had wanted the project to conform to standard prescriptions of central
conflict cinema, but Hou resolved to suppress the register of emotional turmoil (the agony and grief of
poverty and struggle) in Wu's narrative, downplaying the primacy of the individual and Wu's classical
170
Fig. 39: Dust in the Wind (1986)
depiction of suffering. According to Zhu Tianwen, who documented the entire production process, Hou's
decisive adjustments “were made not out of disloyalty to Wu’s story but rather from an antipathy to generic
and emotional contrivance” (Yeh and Davis 2005: 165).
To displace melodramatic expressiveness in a manner emblematic of time-image cinema, Hou
reduces the characters of Dust in the Wind to seers who respond to their diegetic world by silently or
detachedly gazing toward surfaces and distances. Hou and Mark Lee Ping-Bing (Hou's regular director of
photography) complement such acts of visual contemplation by inserting a number of pillow-shots between
ordered sequences depicting human activity: still images of depopulated interiors, clouds drifting above
electricity wires, waves noisily crashing a shore, and the mountainous landscape surrounding Jioufen. In a
recent interview with Wen Tien-Hsiang and Kelly Yang, Lee recalls Hou's intention for making these images
central to the film's narration: “The reason why I took so many long shots of mountains and things was
because they are natural products of the story and the script. I think that the essence of Dust in the Wind lies
in the passing of time, changing of the landscape, and their variations with human relationships” (Lee 2009:
259). Lee also recalls that he and Hou had “wanted to convey some states of mind, some emotions, or the
atmosphere of a situation purely with impressions of the camera,” and, in order to do so, “the passing of time
171
Fig. 40: Dust in the Wind (1986)
had to be felt in the images” (Lee 2009: 259-260). The cutaways to landscape here present little bits of time
in its pure state, conveying the narration's suggestion of atmosphere and feeling through direct time-images
of becoming, change and transition (Deleuze 2005b: 16).
The final sequence of images in Dust in the Wind illustrates Hou and Lee's method particularly well.
Hou permits the film's only raw display of emotion at the narrative's close, with a fixed shot of Wan reacting
to the news that his love Huen has married (in his absence due to national service) by weeping and striking
the bed on which he lies. This intense emotional release is followed by a slow pan across a dense line of trees
(accompanied by yellow light seeping through the clouds), and a series of pillow-shots: four still images of
an indicating railway signal, an empty train station, shadows passing over a hillside, and children playing in
a street in Jioufen. Wan then returns home to find his mother asleep, and in two shots moves through the
house to greet his grandfather outside, who is tending the family's crops. Wan crouches outside and listens to
his grandfather's reflections on cultivating sweet potatoes and ginseng, the cyclical effects of seasonal
changes, and his doubts concerning the year's harvest. After two minutes pass, Hou cuts to a medium shot in
which Wan lights a cigarette for his grandfather, and both men pause to scan their environment, wordlessly
looking up to the forested hills that surround them [Fig. 39]. After another minute, Hou cuts to an extreme
172
Fig. 41: Life and Nothing More... (1992)
long shot (outside the village) of light drifting across those hills, in which the blue sea in the centre of the
frame divides the grey of the townscape and a vast cumulus above. Patches of sunlight crawl across the
peaks unexpectedly quickly, providing visual markers of the passage of time in the shot [Fig. 40].
This final image is reminiscent of Jonathan Rosenbaum's definition of the 'cosmic' long shot
employed by Kiarostami at the end of Life and Nothing More... (1992) and Through the Olive Trees (1994).
Rosenbaum suggests that Kiarostami's extreme long shots (derived, in Rosenbaum's distinction, from visual
strategies in Jacques Tati's Playtime [1967]) “make the world seem like a richer and more complex terrain
than any narrative could possibly contain,” and are designed to displace the space of narrative by retreating
to a 'cosmic' observation of landscape (Rosenbaum and Saeed-Vafa 2003: 21). The final series of shots in
Life and Nothing More... includes a cosmic long shot and dilatory reverse zoom that reveal, over the course
of several minutes, a vista in which a “puzzle-like” interaction takes place between the father (the film's
unnamed protagonist) and a wandering pedestrian [Fig. 41]. The gentle absurdity of the sequence, derived
from the slow action of the father's car impotently rolling down a hill, gives it a distinctive quality that
Rosenbaum refers to as 'Tatiesque' (Rosenbaum and Saeed-Vafa 2003: 22).
173
In Through the Olive Trees, a take of the same length renders the final interaction between Hossein
and Tahereh as barely distinguishable specks in the distance, in a manner similar to the focal sequence shot
of the Magi trudging across the desert in Serra's El cant dels ocells (described in my Introduction). Any
narrative action (their meeting and, presumably, a conversation) is obliterated by the remoteness of the
camera's perspective, which leaves the spectator to study the more distinguishable 'concrete realities' of
landscape instead: the movement of wind in the trees, and the rippling of the long grass, before Hossein
returns. This decentring of narrativity, and an insistence on a 'cosmic' extremity of scale and duration, is
indicative of the somewhat hyperbolic minimalism that Kiarostami would later adopt in Five.
A contemplative cinema: Five: 5 Long Takes Dedicated to Yasujiro Ozu
Kiarostami's Five modifies the device of the still life by extending its operation to comprise entire imagesound
situations. Five is an experimental film that retains nothing of the pragmatism of orderly, linear
narrative cinema, and breaks down any semblance of efficient sensory-motor relations to produce pure
instances of contemplation. The film consists of five minimalist sections that Kiarostami has referred to as
individual works in themselves: four single, unbroken shots that display places, objects, people, animals and
birds, and one composite section that seamlessly branches over twenty different takes to create the
impression of a continuous shot. These five sections, pithily and summarily titled Wood, Promenade, Dogs,
Ducks and Moon and Swamp, are punctuated by languorous fades, and, when taken together, impart an
overall ASL of 889 seconds. Hushed instrumental accompaniments are played in the fades between sections,
and all diegetic sound within each shot is carefully modulated.
The entirety of Five was filmed by, or near, the Caspian sea, during a period of two months in which
Kiarostami stayed in Northern Iran to write the narrative for Jafar Panahi's Crimson Gold (2003). As
Kiarostami recalls in Around Five (2005) (a series of reflections compiled by the filmmaker himself for the
French DVD release by mk2), this assignment gave him the opportunity “to spend a lot of time wandering
around and observing,” and the film was contrived when he encountered the piece of driftwood released to
the sea in the first shot, Wood. Using one of the small DV cameras on which he filmed Ten (2002),
Kiarostami was able to work alone during filming, but some final takes were produced with the help of an
174
assistant, Seifollah Samadian, a few friends (including Jafar Panahi), and associates such as the duck handler
who orchestrated the activity of the fourth shot. The apparent simplicity of visual presentation in Five is
quite deceptive, as every section has been organised and stylised as meticulously as the composition of
objects in the still lifes in Ozu's work. Tellingly, Kiarostami has disclosed that the filming, editing and final
mixing of image and sound for the final version took two years to complete, longer than for each of his
preceding narrative fiction features (Taste of Cherry [1997] and The Wind Will Carry Us [1999]). When Five
was first screened in London as part of the extensive Visions of the Artist retrospective in 2005, Kiarostami
noted that “it was the most difficult film I ever made, but it doesn't show on the surface” (Jeffries 2005).
Kiarostami's tribute to Ozu is more conceptual than Hou's. Although Five is clearly constructed as a
series of extended pillow-shots, there are no noticeable thematic or visual analogues between Kiarostami's
images of the sea and those in Ozu's cinema (which occur in, for instance, Record of a Tenement Gentleman
[1947], Tokyo Story [1953], Equinox Flower [1958] and, most prominently, in Floating Weeds [1959]). In
Around Five, Kiarostami suggests that his film “is not too far removed from Ozu's cinema, which is a kindly
cinema,” and that his intention was merely to echo the “simple subjects” and “everlasting and respectful”
long shots in Ozu's films. In this respect, his tribute is best understood in terms of its undramaticness and
hyperbolic application of duration: as Kiarostami states, “[Ozu's] films were not nervous and melodramatic,
which you find with today's montage facilities. Five, if it is not like Ozu's works, at least is not in contrast
with them—or, at least in contrast with the kind of cinema that Ozu simply avoided.”
By extending the observational capacity of the pillow-shot, Kiarostami bridges the gap between
Ozu's time-images and the tendency toward a protracted observation of landscape in recent durational
cinema. Whereas Bordwell has suggested that although Ozu's “rather fast” cuts give us “time to see
everything,” slow filmmakers test that assumption by making the temporal succession of changing states a
central component of the image (Bordwell 2011). As Peter Hutton has suggested in the context of his own
films, “in observing landscape, sometimes it takes a lot of time to comprehend what’s actually changing. In a
shorter shot you can’t perceive details, subtleties” (MacDonald 2008: 218). Hutton comments that in his
work “each shot becomes a small film,” and, in a quip that is indicative of the direction taken by durational
cinema during the last decade, reveals that “I’ve been thinking of making a film called One (for Kiarostami),
a single killer shot” (MacDonald 2008: 228).
175
Kiarostami's attention to places, objects and their duration is designed to activate a contemplative
spectatorial practice. After emptying out any practical coordination of sensory-motor relations signified by
ordered narrativity, Kiarostami's image-sound situations animate a renewal of relations between the mental
(thought) and the physical (the visible world). In line with Deleuze's contention that modern cinema is
primarily a “system of exchange between the imaginary and the real,” Kiarostami remarks in Around Five
that poetic films should “set our mind free about everything,” and aim to “extract the values that are hidden
in objects and expose them, in the observation of plants, animals and humans, everything” (Deleuze 2005b:
9). As Deleuze argues, the device of the still life performs this observational function by its signification not
of “the absence of a possible content,” but rather “the presence and composition of objects which are
wrapped up in themselves or become their own container” (Deleuze 2005b: 16). This isolation of a fullness
of objects and their duration opens the image to contemplation, and, in Deleuze's conception of modern
cinema, might comprise a renewed ability to think the visible world.
The durational still life's activation of instances of contemplation is, to a certain extent, prefigured in
Bazin's theory as well as Deleuze's philosophy. Bazin alludes to emergent modern relations between the
mental and the physical in his oft-quoted passage on the construction of narrative in Roberto Rossellini's
Paisà: that is, the way in which Rossellini's selection of image-sound situations fail to “mesh like a chain
with the sprockets of a wheel,” and force the mind “to leap from one event to the other as one leaps from
stone to stone in crossing a river.” In this set of false continuities, Bazin suggests that as the spectator is
prone to lose their footing and slip, “the mind does likewise. […] Facts are facts, our imagination makes use
of them, but they do not exist inherently for this purpose” (Bazin 2005b: 35). This absence of causal
association between sequences in modern cinema (shorn of an efficient sensory-motor schema) forms, in
Denis Lévy's contention, the necessary basis “for a gaze that shakes the barbarity of the imaginary in order to
reach the free exercise of thought” (Lévy 2011).
Bazin also suggests that this free exercise of thought is enabled by the modern sequence shot.
Whereas, in classical cinema, “analytical montage only calls for [the spectator] to let his attention follow
along smoothly with that of the director who will choose what he should see,” the deep-focus sequence shot
permits the spectator “to exercise at least a minimum personal choice.” This appropriation of perceptual
realism “implies, consequently, both a more active mental attitude on the part of the spectator and a more
176
positive contribution on his part to the action in progress” (Bazin 2005a: 35-36). Although it is clear that
Bazin assumed that this activation of the mental would be directed toward deciphering and enriching
narrative action, a further implication of his analysis points to a relation in which the meaning of the image
derives entirely from the spectator's attention and will. As I suggested in Chapter Two, Bazin's attention to
the contemplation of 'useless details', such as the materiality of objects in the kitchen sequence in Orson
Welles' The Magnificent Ambersons, points to a direction that is taken, with renewed observational rigour, by
contemporary slow cinema. With its dissolution of 'action' in favour of the extended observation of
landscape, alongside what Adrian Martin refers to as its minimal 'system of activities', contemporary
durational cinema directs mental activity toward spaces, objects, and their durations in ways that Bazin
prefigured, yet did not fully anticipate in practice.
It is perhaps worth noting here that Kiarostami has encouraged spectators of Five to exercise their
mental attention in ways that are usually rejected by commercial filmmakers. Five has been screened
publicly both as a gallery installation and as a standard feature in cinemas, but Kiarostami notes in Around
Five that limited theatrical and home screenings are his preferred modes of exhibition, not least because they
minimise any distractions from people entering and leaving the space of projection. Kiarostami has
suggested that an ideal screening should involve only a comfortable chair, a screen, and total (ambient)
darkness to provide conditions amenable to drifting off to sleep—to which he affirms: “I am not joking. I can
confidently say that you will not miss anything if you had a short nap.” Five's image-sound situations are
intended not to sustain a 'positive', pragmatic attention, but rather a spatial and temporal continuum that
permits the spectator to slip into the true realm of the imaginary: an emancipated state of dreaming or
reverie.74
As an alternative to these ideal conditions for narcoleptic spectatorship, Five has been screened in
galleries in two distinct ways that permit different modes of spectatorship. At the Abbas Kiarostami: Image
Maker installation at MoMA in New York in 2007, the film was divided into its individual sections and
looped in five separate partitioned spaces. However, at the extensive Erice—Kiarostami: Correspondences
exhibition, curated by Alain Bergala and Jordi Balló, at Centre de Cultura Contemporània de Barcelona in
74
Kiarostami's comments here recall Raul Ruiz's reflections on sleep in Poetics of Cinema, in which Ruiz suggestively
describes the periods of time that we spend awake as deftly constructed sequence shots: “It become clear to us that
the moments of real life functioned like a film, with segments spliced together so as to produce the illusion of
continuity” (Ruiz 2005: 118).
177
Fig. 42: Five: 5 Long Takes Dedicated to Yasujiro Ozu (2003)
2006 (which also travelled to La Casa Encendida in Madrid later in 2006 and Centre Pompidou in Paris in
2007), Five was projected in full in a small open room surrounded by other shorter video installations,
photographs and sculptural works (by both Kiarostami and Victor Erice: see Erice and Kiarostami 2006). The
former approach to public exhibition is probably more amenable to the structure of Kiarostami's film, as it
actualises Bazin's analogy of “stepping from stone to stone” within the actual space of the installation (Bazin
2005b: 35). Sharon Lockhart has also organised gallery screenings of Double Tide (2009)—a digital film that
I will describe shortly—in this manner, projecting each of the film's two shots in separate spaces that the
spectator must walk between to 'self-edit' their looping durations (such as at Galerie Jan Mot in Brussels in
2010).
178
Fig. 43: Five: 5 Long Takes Dedicated to Yasujiro Ozu (2003)
Five long takes
Although describing Five in considerable detail here might (quite reasonably) be deemed to be excessively
fastidious, I feel it is worth the effort to illustrate both the intricacy and fullness of 'concrete realities' in the
film's image-sound situations. One of the primary distinctions between contemporary slow cinema and other
recent forms of documentary, art or dominant cinema is a willingness to display content at the limits of
undramaticness that nevertheless rewards sustained attention. In experimental works such as Five, the
filmmakers' analysis of the undramatic field in time is no longer contingent on human presence, or activities
centred on games of fiction. Instead, it is given over entirely to insignificant notations that cannot adequately
be contained by more active narrational systems.
Five's first shot, Wood, lasts seven minutes and fifty-six seconds, and depicts a piece of barnacled
driftwood breaking in two on the shoreline of the Caspian Sea [Fig. 42]. The driftwood was selected by
Kiarostami for its resemblance to a weathered “sea creature,” and was carefully placed in the path of the
incoming tide before the camera was activated. After a minute, a wave dislodges the object's position and
179
forces it to roll back with the tide, a motion that the camera observes by continuously reframing, jerkily, as
the driftwood is buffeted by the variable current. After three minutes the driftwood breaks into two pieces,
and a large chunk of its body is drawn out to sea. The end piece remains immobile in the bottom left corner
of the frame, and the larger chunk is washed out of the image after four minutes, before bobbing gently back
into the top right corner a minute and a half later.
After the camera reframes to take in a slightly more expansive visual field divided by the diagonal
shoreline, the larger chunk of driftwood floats out of the top of the frame again after two minutes, and the
screen fades to black. All activity within the shot is contingent on the ebb and flow of the tide, and the
resilience of the wood to the strength of the current. In Around Five, Kiarostami quips that, in orderly
dominant cinema, an efficient filmmaker would have ensured that the driftwood splits at the right time by
detonating an explosive and using a boat to winch it out to sea. The composition and temporal structure of
the action in his shot, however, is dependant on an aleatory and unpredictable cooperation of non-human
phenomena: earth, wind and water.
The film's second take, Promenade, is a static planimetric composition that lasts ten minutes and
thirty-three seconds [Fig. 43]. The image comprises three planes: a seafront promenade, a line of white
railings that collide with waves breaking on the shore, and a flat blue expanse of sea. Throughout the shot,
men and women walk (and run) in both directions across the frame, at varying speeds and distances from the
camera. Joggers, dog walkers, surfers and students pass by, and a couple of pigeons strut around the bottom
left corner of the frame. The weather appears to be crisp, and many walkers carry umbrellas or canes. After
seven minutes of this (almost) bustling activity, two pairs of ageing men pause to greet each other at the right
side of the frame. As they talk, we quickly realise that their dialogue is inaudible, and that the sonic rhythm
of waves crashing against the shore comprises the only sound in the shot. After conversing for over two
minutes, the men part company, and this casual everyday event segues into a post-action lag that lasts over a
minute. During this time, no more people pass through the camera's field of vision so that the frame remains
uncannily empty—a clear signal that all visible activity is still being carefully coordinated by the 'invisible'
filmmaker. The image then slowly fades to white, and the noise of the waves is overlaid with non-diegetic
music.
The third shot, Dogs, submits a fixed composition of a pack of dogs resting on the beach to a
180
Fig. 44: Five: 5 Long Takes Dedicated to Yasujiro Ozu (2003)
systematic process of abstraction [Fig. 44]. Over the course of seventeen minutes and sixteen seconds, the
image is gradually bleached by the rising morning sun until its surface becomes a veneer of blazing white.
During the first half of the shot, Kiarostami creates a cavernous soundscape of rolling echoes produced by
breaking waves, accompanied by the distant interruptions of a barking dog. The camera is positioned at the
top of a beach, and the seascape is divided into three sectors: the light, almost teal, sky; a large, serene
expanse of sea; and, at the bottom of the frame, thick grey sand furrowed by a few visible trails. A few dogs
lie on the plain of sand, and are at first distinguishable only as amorphous black shapes. After four and a half
minutes two of the dogs, wagging their tails, rise to congregate with the others (apparently, in order to restrict
their movements, Kiarostami left food for them in strategic spots on the beach). Around this time it becomes
clear that the tonality, and temperature, of the image is gradually changing, and that the DV camera's
inability to register an intensity of light without protection is distorting the shade of sand and sea as the sun
rises.
After one of the dogs pads out of the frame, the tonal distinction between sectors softens, and the
colours of sky and sea begin to merge. The eye soon has to strain to make out the line of the horizon, and the
181
Fig. 45: Five: 5 Long Takes Dedicated to Yasujiro Ozu (2003)
bodies of the dogs begin to smear into their immediate environment. As a bird flies across the skyline, the
shape of waves rolling into the shore, in Kiarostami's description, “gradually turn into sea creatures.” This
abstraction slowly escalates until, after fifteen minutes, the dogs become indistinct silhouettes, and the entire
image is discoloured beyond recognition. During the final minutes of the shot, the sound of the waves is
purposely muted, and by seventeen minutes everything has been “completely annihilated in front of our
eyes.” In Around Five, Kiarostami suggests that the gradual obscuring of information within the shot
provides a visual actualisation of duration, and that the image's alterations “take place so slowly that we
actually cannot see any change. But after a few minutes we feel that everything is moving towards an
absolute brightness, then it melts and is annihilated.” Our intermittent awareness of the technological
impairment of the image's content marks the passage of time in the shot—a realism of duration rather than a
strictly comprehensible documentation of the material world.
The fourth shot, Ducks, lasts six minutes and thirty seconds, and presents a raft of ducks traversing
the visual field from left to right, and back again, in a static set-up facing the incoming tide [Fig. 45]. The
composition is similar to that of Dogs, but adopts a non-human level of perspective that is much closer to
Ozu's set-ups. David Bordwell has described the axiomatic adoption of low positioning in Ozu's style as
182
“non-anthropocentric” (referring to Masahiro Shinoda's statement that “the reason [Ozu] placed the camera
so low was to prevent it from having a human viewpoint”), and that distinction seems particularly
appropriate here (Bordwell 1988: 79). There is no pre- or post-action lag in Ducks, and the birds make their
way across the frame immediately: some waddle slowly, some hastily, and some hesitate before quickening
their pace. Kiarostami mutes the cacophonous quacking that would have arisen from the concentration of
eight hundred birds within range of the microphone (a recording of which is glimpsed in Seifollah
Samadian's footage in Around Five), and reduces the sound track to a modulated patter of webbed feet. After
four minutes, a gaggle of birds attempt to keep pace with a white duck that has sprinted across the frame
until, directed by an experienced handler, they all suddenly reverse direction. After a couple more minutes,
Kiarostami ends the shot, the shortest (and funniest) in the film, with a swift fade to black.
The fifth section of Five, Moon and Swamp, depicts the moon's reflection on the surface of swamp
water before, during and after a thunderstorm [Fig. 46]. The section lasts twenty-seven minutes and fifty-one
seconds, and consists of a composite of multiple takes, filmed over the course of four months, that depict a
progression from a still night to the onset of a thunderstorm and the breaking of dawn. Even more so than in
the preceding four shots, the soundscape of Moon and Swamp is the section's primary operative element, and
is carefully mastered to conceal the discontinuities in the image track.
The full 'take' begins with the sounds of a lone dog barking and a quiet chorus of croaking toads. A
blurred, silvery reflection of the moon slowly appears, disappears, and reappears over the course of the first
few minutes, partially obscured by limpid pondweed. The call and response rhythm of croaking intensifies
during this period, and after four minutes a thunderclap is heard. After six minutes, a brighter moon emerges
from behind a cloud, and illuminates the edges of the frame for the first time. This illumination reveals that
the position of the camera and configuration of weeds on the pond's surface has unsubtly altered, and the
temporal shift has been disguised by the image's sporadic lapses into total darkness. This cycle of sounds and
details is repeated for the next eight minutes, until the soundtrack becomes ominously quiet, and the croaking
repeatedly quickens and slows, as if sensing a meteorological disturbance to come.
After fifteen minutes the reflection of the moon is rendered invisible by an onset of torrential rain
and bursts of lightning. Only a lone dog howls on the soundtrack, and the image is obscured for long
stretches that are punctuated by abstract flashes. This 'drama' lasts for three minutes until calm is restored,
183
Fig. 46: Five: 5 Long Takes Dedicated to Yasujiro Ozu (2003)
and a gentle rain falls to dissipate the moon's reflection on the water again. After twenty minutes have passed
the symphonic croaking returns, and a bright, clear composition is held for over three minutes, during which
a tiny pond skater traverses the frame. After twenty-five minutes a cock crows and the frame begins to
lighten: voices of animals and birds greet the dawn, and in the space of two minutes the image mutates from
a monolithic blackness to the rich teal and purple hue of the swamp's calm surface. A title caption reads
“November 2003,” and the film abruptly ends without a customary fade-out.
Five and digital cinema
As Kiarostami comments in Around Five, “the making of [Moon and Swamp] is more like a normal
cinematic work. The film looks as if it comprises only one shot, but there are as many shots as there are
patches of darkness in the film.” The intermittent filtering of darkness and light throughout the section
seamlessly masks a proliferation of cuts that were necessitated by practical eventualities, then shaped into a
symphonic time-image during post-production. During filming, Kiarostami found that the moon's position
would shift inside a static set-up after only a couple of minutes, and disrupt the immobility and balance
184
necessary for an extended still life. This contingency was compounded by two larger factors: the location of
the pond and the cyclical lunar phase itself, meaning that Kiarostami had only two hours a night, for two
nights a month, to film the moon's reflection in unpredictable weather conditions. As a result, Kiarostami
returned to the same location multiple times over the course of several months (in turn travelling over
400km), to compile a set of takes that could only later be worked into a desirable sequence shot.
In this respect, Moon and Swamp is a wholly 'digital depiction', in the sense that the final sequence
actively replaces, on the level of both input and output, what D.N. Rodowick refers to as the “(analogical)
transcription” of film with “(digital) conversion or calculation” (Rodowick 2007: 110, 116). Every aspect of
the production of Five is illustrative of the impact of digital technology on contemporary film over the past
fifteen years, a trend that is most evident in the field of slow cinema on the level of independent practice and
attenuated durational style. In this period, technological innovation and the competitive pressures of
globalisation have increased the access to (and portability of) DV cameras, data storage and editing software
whilst decreasing their cost, and digital software has introduced graphic techniques that have completely
altered montage relations by increasing the capacity for duration in actual or apparently unbroken shots.
For Kiarostami, the ease and portability of digital cameras enabled him to work on Five without the
distraction of a crew, and served to relieve fiscal pressure from the process of image-sound production. For a
contemporary filmmaker, the equipment necessary to produce a depopulated observational work can now be
reduced to only a DV camera (if an appropriate aesthetic option), adequate data storage space (for instance, a
set of multiple-terrabyte hard drives), editing software, and material support for a period of filming (which, if
necessary, might come from another occupation, project grants, residency commissions, or revenue streams
from other artworks or exhibitions). With these advantages, digital technology frees time, reduces costs and
grants independence to the filmmaker, in turn enabling an actively 'unproductive' practice that can withdraw
from the capture of what Lyotard refers to as “the synthesis of good movement” that increasingly defines the
visual language of dominant cinema (Lyotard 1986: 353),
These particular technological advantages had, of course, been sought by filmmakers prior to digital
innovation, and it is worth noting that James Benning, before converting to digital in 2009, used 16mm
rather than 35mm film for a similar purpose.75
In Circling the Image (2003), Benning notes that the electric
75
Accounting, of course, for standard analogue limitations of processing costs and the restriction of reel lengths. In
this context, 16mm signifies only a step closer to the advantages of digital for independent filmmakers.
185
Bolex camera and Nagra tape recorder that he used since 1975 fulfilled two practical requirements: firstly,
the capacity to be transported by car and moved around on location as quickly as possible (in response to
potential changes in ambient light or unexpected activity), and, secondly, to be operated independently. The
latter factor is critical for the contemplative aspect of Benning's practice, which he describes as “a need to
reflect on what I capture on film without other people being around me.” This sentiment is echoed by
Kiarostami in relation to Five, when he discloses that the film's production had been his most pleasurable
“because I was able to preserve my solitude during the making of it.”
Graphism in Five and contemporary cinema
The form of Kiarostami's Moon and Swamp is symptomatic of analogue cinema's convergence with new
media's practice of 'graphism', which is described by Rodowick as “the ability to erase and efface, to add and
subtract, to alter perceptual values in a painterly or pictorial way” (Rodowick 2007: 105). In The Language
of New Media, Lev Manovich notes that, in dominant cinema, techniques such as the seamless compositing
of images in Moon and Swamp are referred to as 'invisible effects', as they tend, more often than not, to
simulate aspects of classical film language rather than point to a new aesthetic (Manovich 2001: 254). In this
sense, it is possible to suggest that the seamless branching of cuts in Moon and Swamp comprises a more
refined version of similar methods in classical film, such as Alfred Hitchcock's first long-take work, Rope
(1948). In Rope, five of the cuts between the film's ten long takes (which last an entire reel each) are
camouflaged to maintain a performative tension of temporal continuum. Hitchcock attempts to disguise edits
by tracking into the backs of characters and the surface of objects before cutting in a slightly unwieldy
fashion, but the purpose of the rudimentary operation is clear: that 'concealed' edits are necessary to sustain
an analysis of the dramatic field in time, whether the spectator notices their presence or not.76
In contemporary commercial cinema, this practice is of course considerably more refined, and can
76
In Hitchcock's film, only five cuts are camouflaged because the necessary reel changes are met with conventional
reframings at alternate edits. Jean-Pierre Coursodon, in a fascinating essay on the film, describes Hitchcock's use of
extended takes as a provocation of fetishistic pleasure to compensate for the film's lack of a heroine. Coursodon
compares the tension of long-take disclosure of narrative action to sexual tension itself, both in relation to shooting
(in which “the concentration required to bring a long take safely to its end creates a suspense similar to the one
engendered, in sexual intercourse, by the concentration that postpones orgasm in order to prolong pleasure”) and
spectatorship: that “the jouissance of the long take” makes the pleasure last, unlike short takes that tend “to suggest
premature ejaculation” (Coursodon 2004).
186
now extend to a covert simulation of what Manovich refers to as 'spatial montage': the appearance of
different images on the screen at the same time rather than in succession. David Fincher, for instance,
constructs single images from sections of multiple shots (for the most ordinary as well as spectacular scenes)
that are seamlessly branched entirely in spatial terms. For The Girl with the Dragon Tattoo (2011), Fincher
routinely shot dozens of takes for scenes, and his editor, Kirk Baxter, later 'glued' together either two or three
different sections of images (with their favourite elements or performances) in the editing room (see Edwards
2011). In this practice, as Manovich suggests, “the logic of replacement, characteristic of cinema, gives way
to the logic of addition and co-existence. Time becomes spatialised, distributed over the surface of the
screen” (Manovich 2001: 270, 272). Fincher is quite clear about the purpose of this graphic manipulation,
celebrating the fact that digital montage enables a seamless compression of time in order to give “the illusion
of continuity, even as we take out the 'boring' bits” (Edwards 2011). In slow cinema, 'graphic montage'
techniques are employed solely for temporal purposes, in order to extend the impression of duration in a
manner closer to Hitchcock's experiment in long-take style.
Sharon Lockhart's Double Tide (2009) contains a particularly sophisticated instance of graphic
montage in the field of slow cinema. Lockhart's film is a portrait of Jen Casad, a clam digger, harvesting at
dawn and dusk on coastal mudflats near Tremont, Maine in North America. Double Tide was filmed directly
after Lockhart finished shooting Lunch Break and Exit (both 2008) in two factories in the area, and consists
of two static shots that last forty-three and fifty-six minutes each. These shots display, in a subtle simulation
of 'real time', Casad performing the arduous labour of digging for clams, and wading the distance of a cove
whilst the sun rises (in the first shot) and sets (in the second shot) [Figs. 47-48].
The film's title refers to the seasonal period, during summer solstice, when clam diggers work at both
dawn and dusk to increase their revenues by taking advantage of the daily double tides. Casad, Lockhart and
her crew filmed for five days in three different locations, and the takes used for the final film were shot in
Seal Cove (near Tremont) on the third day (which is credited in the closing titles as “July 22, 2008”). As the
tide would usually be out for three or four hours during both periods of the day, Lockhart gathered footage
over a longer time span than is depicted in the final pair of takes. The initially analogue footage was shot on
16mm with 400ft reels, meaning that the complete film actually consists of ten separate shots that, as in
Kiarostami's Moon and Swamp, have been seamlessly merged to create the illusion of an unbroken
187
Fig. 47: Double Tide (2009)
performance of labour.
The placement of edits in the final version is entirely imperceptible (or, at least, I have only spotted
one during four viewings on two different formats) and, in order to preserve the visual integrity of continuity
during takes, the film can only be projected on video.77
As Lockhart notes, “there was nothing really digital
about [the film] except the final output. Like NO [2003], the cuts were hidden in order to get a seemingly
continuous take of an activity” (Holte 2010: 2). As Lockhart rarely operates the camera herself, Double Tide
was shot with a small crew in the manner of a conventional analogue production—a practice Lockhart sticks
to as “there’s nothing casual about it: it heightens the performance.” At the first of the film's screenings at
London Film Festival in 2010, Lockhart reflected on the unusual nature of the production by joking that,
77
As a result, Double Tide has also been projected as a video installation, with the two takes divided into separate
screenings. At Galerie Jan Mot in Brussels in 2010, the two shots of 'sunrise' and 'sunset' were projected in separate
spaces about thirty feet from each other, so the spectator was able to walk between the two 'volumes' of space and
time to self-edit them.
188
during an earlier screening in Berlin, “everyone laughed when 'Editor' [May Rigler] came up as a credit.”78
The first 'shot' of the film begins as soon as Casad enters the frame, pulling a sled in her wake. A
mile of mud separates the shore and the position of the camera, which is set at a considerable distance from
the incoming tide. Initially, the visual field is light with mist and grey in tone, and the soundtrack is animated
by birdsong and faint ambient traffic noise. During the first ten minutes of the take, the mist thickens to
obscure the image's depth of field, concealing the outline of trees in the distance. Casad's indirect path across
the mudflat is determined by the location of clams under the surface, and she repeatedly doubles over to
plunge an arm into the thick mud to extract the clams as she wades across, in turn producing a thick suction
sound.
Casad, also an artist, habitually digs with her left arm in order to protect her right hand for painting at
other times of the day. In the second take, however, she injures her left hand, and temporarily switches to
harvest with her right in order to preserve her digging hand. This seemingly insignificant notation, to use
Barthes' term, repays the spectator's close attention to her repetition of labour over time. After twenty-five
minutes, a distant foghorn is heard, and the balance of the composition begins to alter: the mist clears to
reveal the mouth of the cove, as well as an outcrop of green firs atop a foundation of pale stone in the
distance. The image then transforms into a 'cosmic' long shot during its second half, as Casad retreats to the
very rear of the frame near the incoming tide, and ends after forty-three minutes when she returns to her sled
close to the camera.
At the beginning of the second shot, at dusk, the camera is positioned closer to the outgoing tide. The
lack of dawn chorus foregrounds Casad's noisy extraction of clams, as well as a brief bout of coughing and
sneezing. As in Kiarostami's Dogs, it takes a substantial amount of time to register the failing quality of light
as time passes, which becomes obvious here only after ten or fifteen minutes have passed. After twenty
minutes, the slick surface of the mud takes on a steely blue-grey tone that reflects the discoloured sky, and
the outgoing tide is reduced to a thin pool of ink in the centre of the frame. Sunlight begins to drain from the
clouds, which turn white, then pink, after half an hour. A luminous pink hue is cast upon the liquid skin of the
78
A recording of this post-screening conversation with Lockhart is available at http://vimeo.com/20341992 (part I) and
http://vimeo.com/20394440 (part II) (thanks to Revista Lumière). Also, some information that I have noted about the
production of Double Tide is taken from an unpublished interview that I conducted with Lockhart in October 2011.
A couple of details in later paragraphs about Lav Diaz's method also comes from similar conversations, with both
the filmmaker and acquaintances.
189
Fig. 48: Double Tide (2009)
mud until, after thirty-five minutes, the image returns to a richer metallic tone. The hooting of an owl and
chirping of crickets signals the final setting of the sun, and a stork enters the right side of the frame to linger
just beyond Casad's path. Towards the end of the shot, Casad's work is momentarily interrupted a couple of
times, as she pauses to look up at a plane passing overhead, and catch the sound of children playing behind
the camera (the film's only other implied human presence). A light orange hue finally sets in above the trees
as Casad returns to her sled, and scrubs and rinses the clams in the muddy water. Light then begins to
disappear quite quickly, and the shot ends when Casad drags her sled out of the frame after fifty-six-minutes.
During filming, Lockhart used sticks and stones as markers to indicate the edge of the visual field to
Casad, and contain her movements within the parameters of the frame. The path of Casad's harvesting was
otherwise not choreographed except for Lockhart's need to prepare the graphic transitions that would be
completed during post-production. After the film's screening at London Film Festival in 2010, Lockhart
190
disclosed that Casad carried a walkie-talkie whilst digging so she could be informed when each reel of film
was coming to an end. As each eleven-minute endpoint approached, Lockhart instructed Casad to find a
comfortable position and pause whilst the crew halted filming, reloaded the magazine, and activated the
camera again.79
The motionless images were then 'glued' together in the editing room, and the points at which
Casad paused during filming are not detectable: at times we see her wipe her brow with her arm, or stiffen to
stretch her back, but otherwise her 'performance' is intensive and fully continuous—there are no periods of
temps mort, or inclusions of pre- or post-action lags. The labour of Casad's performance is, ultimately,
overwhelmed by both distance and duration in the final version, as Double Tide comprises a hyperbolic
distension of even slow cinema's standard long-shot long-take aesthetic. Here the aesthetic is shorn of
narrativity in any guise, and applied to a pair of image-sound situations that are designed solely to 'open' the
world to the spectator's gaze.
It is worth noting that Lockhart employs similar strategies to those in Double Tide for NO (2003), a
thirty-one minute film of two Japanese farmers piling and raking fifteen stacks of hay across an otherwise
barren field. The activity of the film is structured with almost mathematical precision: firstly, the farmers
deposit stacks of hay in five rows from the rear of the visual field to the front, with three piles placed in each
line [Fig. 49]. After reaching a position close to the camera, they then rake the stacks across the dark earth in
the opposite direction from which they came [Fig. 50]. The film is recorded on four 400ft reels, and, just as
in in Double Tide, the farmers had to delay their labour (this time, whilst fetching hay off-screen) at the end
of reels until the camera was reactivated with a fresh magazine. Instead of using sticks and stones as markers
to direct movement, however, Lockhart placed oranges in the earth that were visible to the farmers but not to
the camera.
NO is, first and foremost, a structural experiment in the dialectical relationship between spatial
perspective and time. James Benning, in a short tribute to the film published by Cinema Scope in 2010,
notes that Lockhart's camera is deliberately set up to capture the visual field as a trapezoid which “is then
79
Lockhart also used the walkie-talkie to contact Casad if she strayed too far to the edge of the frame, which nearly
happens towards the end of the second shot. Albert Serra uses the same tactic in El cant dels ocells, in the sequence
shot of the Magi trudging across the desert that I described in my Introduction. To coordinate this shot, Serra gave
walkie-talkies to his actors and delivered unintelligible instructions intended to confuse them when they reached the
horizon. This caused them to hesitate after five minutes and walk back toward the camera and crew to confirm
Serra's instructions, thus completing a cyclical journey to desired effect. This somewhat unconventional method,
characteristic of Serra's attitude toward his actors, is detailed in Mark Peranson's excellent documentary of the film's
production, Waiting for Sancho (2008).
191
Fig. 49: NO (2003)
mapped into the film frame as a rectangle.” In order to balance sectors of the image, Lockhart ensured that
the farmers made haystacks at the rear larger than those in closer proximity to the camera. Due to its
trapezoidal shape, the front half of the visual field necessarily has a smaller area than the rear, meaning that
the farmers deposited less hay on the stacks and were able to cover the ground in less time. As Benning
notes, “since the trapezoid appears as a rectangle in the film frame, the two halves appear equal,” and
“because of this illusion, time seems to slow down as the workers go deeper into the frame” (Benning 2010:
96).
In the first half of the film, as the farmers build the stacks from the rear to the front, their activity
appears to quicken slightly as they reach the camera, but when they reverse direction the time taken to
complete each line is deliberately delayed. Here we are reminded of Chantal Akerman's dictum for the
performance of activity and labour in durational film (spoken to Delphine Seyrig whilst filming Jeanne
192
Fig. 50: NO (2003)
Dielman): “I don't want it to look real, I don't want it to look natural, but I want people to feel the time that it
takes, which is not the time that it really takes” (M. Rosen 2004: 125). In light of Akerman's statement,
Lockhart has suggested that her films are designed to “unravel the expected trajectory of time,” and distend
the spectator's experience of its passage: “People are used to experiencing time as money, as labour. They
don’t slow down and contemplate the everyday. I’m interested in creating that opportunity for
contemplation” (Holte 2010: 2). NO's composite take preserves the singularity of the time of its performance
(both of the farmers and of the mathematical task), and ends as soon as its system of activity is complete. At
the close of the film, we realise that the surface raked by the farmers has, in direct proportion with the visual
field projected on the screen, formed a perfect trapezoid of hay on the ground in front of the camera.
193
Contemporary slow cinema and the digital transition
Over the course of the last fifteen years, certain filmmakers who started working on 16mm or 35mm film
have increasingly switched to digital production, and it is worth reflecting on this transition in some detail
here. Alongside Kiarostami, who made the break after The Wind Will Carry Us (1999), other prominent
filmmakers who have converted to digital, for both fiscal and aesthetic purposes, include Pedro Costa (who,
after Ossos [1997], has used a small Panasonic DV camera and a crew of one or two), Jia Zhang-ke (after
Platform [2000]), Chantal Akerman (in Sud [1999], De l'autre côté [2002], Là-bas [2006] and Tombée de
nuit sur Shanghai [2007]), Aleksandr Sokurov (in most of his work since Whispering Pages [1994]), and
James Benning. Benning's recent transition to digital after three and a half decades of 16mm production has
been quite remarkable, both in terms of the adaptability of his practice and a marked acceleration of
production. After his final series of 16mm works (Ten Skies [2004], 13 Lakes [2004], casting a glance [2007]
and RR [2007]), Benning's films have been distinguished by freer temporal structures, an increased emphasis
on focus in depth (enabled by HD), and an even greater openness to the observational potential of extended
duration. Technological liberation has been matched by a major compositional shift in his work: by replacing
the resolution of 16mm with the higher definition of HD (almost equivalent to 35mm), Benning has also
exchanged a standard 4:3 framing for an anamorphic 16:9 ratio, in turn modifying the typical ordered
'squareness' of his compositions.
After his first digital work, Ruhr (2009), Benning has produced a series of films that comprise either
single or multiple shots in durations that vary between a few minutes and an hour and a half. These works,
worth itemising in full to illustrate the range of their image-sound situations, include Pig Iron (2009), a
thirty-one-minute shot of trains collecting slag at the HKL Steelworks in Duisburg; Twenty Cigarettes (2011),
which comprises twenty sequence shots of a series of performers (including filmmakers Sharon Lockhart and
Thom Andersen) smoking an entire cigarette each; Nightfall (2011), a single take of its titular event in a
forest in the Sierra Nevada; Two Cabins (2011), two fifteen-minute takes filmed from the windows of
Benning's timber reconstructions of Henry David Thoreau and Theodore John Kaczynski's cabins that he
built in the Sierra Nevada in 2008; small roads (2011), featuring forty-seven shots (filmed over the course of
two years) of lanes and minor roads across the states of California, Missouri, Nebraska, Wisconsin and
194
Wyoming; two radical 'remakes' of American films from the late 1960s, Faces (2011) and Easy Rider (2012),
in which Benning isolates all the (literal) images of faces in John Cassavetes' film to shape a new work, and
replaces the content of Dennis Hopper's film with a series of static takes of its locations filmed over forty
years later; Stemple Pass (2012), a two hour work composed of four static shots of landscape next to
Benning's reconstruction of Kaczynski's cabin in the Sierra Nevada, filmed from the same position during
different seasons (and accompanied by Benning reading Kaczynski’s texts on the soundtrack); and the war
(2012), a fifty-five-minute structural essay on the Russian activist collective Voina (the “war” of the title)
and the recently imprisoned Pussy Riot, featuring unsubtitled video (much of it no longer easily accessible)
culled from the internet, which is followed by a belated translation of audible dialogue against a black
screen.
Benning's conversion to digital from analogue production has, more recently, been repeated by JeanMarie
Straub after the passing of Danièle Huillet. Although Straub has not been nearly as prolific as
Benning, his last three textual adaptations—O somma luce (2010), Un Héritier (An Heir) (2011) and
Schakale und Araber (2011)—have each been filmed on HD video with no detectable alteration in style to
Straub and Huillet's preceding 16mm and 35mm work. In Straub's films, one recording tool is simply
replaced by another, and the set of (ostensibly modernist) formal properties that distinguished his and
Huillet's analogue cinema have remained broadly intact.
The broad base of slow cinema is, on the whole, divided between 'native' digital filmmakers (whose
independent cinema would not have been possible without the development of the format), filmmakers who
have steadfastly continued to work on film (predominantly 35mm), and those who strategically work in both
formats. The former group includes Liu Jiayin, Albert Serra and Lav Diaz (in his independent work after
Evolution of a Filipino Family [1994-2004], which also incorporates 16mm footage shot before Diaz bought
a DV camera). The second group comprises, amongst others, Lisandro Alonso, Hou Hsiao-hsien, Peter
Hutton, Fred Kelemen, Tsai Ming-liang, and Béla Tarr, and it is perhaps telling that the output of some of
this group appears to have slowed in recent years, most likely due to the inability to secure funding for the
increased cost of film stock and processing. For instance, Kelemen has not made a film since Fallen (2005),
and Tarr has signalled that A torinói ló (2011) will be his final directorial work. Peter Hutton, unlike
Benning, has resisted taking up digital and continues to shoot on 16mm with an electric Bolex: despite
195
working on a number of projects in recent years (in Detroit, Ethiopia, and filming The Poor Stockinger, The
Luddite Cropper and the Deluded Followers of Joanna Southcott [2012] with Luke Fowler in England), he
has not completed a solo film since At Sea (2007).
Filmmakers who incorporate both formats include, for example, Apichatpong Weerasethakul and
Raya Martin. Apichatpong prefers to work on 35mm and Super 16mm for fiction features (adopting the latter
format for Uncle Boonmee Who Can Recall His Past Lives [2010] as a result of funding constraints), and has
disclosed that he finds digital inadequate for features (in a similar manner to comments that have been made
by Kelemen, Tarr and Patrick Keiller): “I don't think the technology is there yet. The image quality of film is
much better still. But we will see what kind of video cameras would be available in two or three years”
(Chang 2011). Apichatpong uses SD and HD video for work that necessitates a greater “spontaneity” of
observation during filming, an advantage that outweighs the depreciation of material registration in the
image—for instance, in short films (Mobile Men [2008]), art installations (Haunted Houses [2001], Morakot
[2007], Phantoms of Nabua [2009], the seven Primitive videos [2009] and Mekong Hotel [2012]), and
commercial commissions (such as Vampire [2008], which was funded by Louis Vuitton). Raya Martin,
similarly, has shot features on 35mm when grants are available (namely, A Short Film about the Indio
Nacional (or The Prolonged Sorrow of Filipinos) [2005] and Independencia [2009]) and reserves video for
documentaries (The Island at the End of the World [2005]), short abstract sketch films (Track Projections
[2007], Ars colonia [2011]), and experimental features that operate at, and beyond, the limits of narrativity
(Autohystoria [2007], Next Attraction [2008] and Now Showing [2008]).
The field of slow cinema, then, continues to incorporate both digital and analogue technologies for
different aesthetic purposes. I would contend, however, that cinema itself is now first and foremost a digital
medium—on the level of craft, technology, distribution and exhibition—and that recent durational cinema, as
a result, displays properties that are quite unique to the convergence between analogue and digital that has
taken place over the last twenty years.
196
Digital transitions and filmic ontology
As I have previously proposed that André Bazin's (analogue) model of modern cinema can serve as a key
basis for understanding the role of duration in particular contemporary films, the displacement of analogue
filmic ontology signified by digital technology remains a pressing issue. In drawing upon Bazin in such
detail, I hope not to have already implied that the ontological specificity of photochemical relations between
image and world remain supreme, or that digital technology has not fully encompassed the capture,
distribution and exhibition of image-sound production. In general, films that are shot on film are now
scanned and edited digitally (only experimental practice might provide consistent exceptions to this rule),
and all finished films are increasingly, for better and worse, projected in theatres from DCP (or lesser media
such as Blu-ray), and circulated both online and offline through a variable deluge of formats, platforms and
codecs.
The introduction of digital has mostly been considered by theorists to signify a transformation of
Bazin's conception of photographic ontology, an approach that is best summarised by Steven Shaviro's claim
that, in today's mediascape, “even the most mimetically faithful images are artificial and fictive. There is no
longer any ontological distinction between a 'true' image and a 'false' one” (Shaviro 2007: 65). In Bazin's
theory, the ontological distinction between a 'true' photochemical image and a 'false' painting, etching or
sculpture (for example) provides the foundation for his conception of realism, but, as I have already
suggested, should by no means be constrained by it.
For Bazin, analogue photography always entails a “transference of reality from the thing to its
reproduction,” thanks to the scientific circumstance that first “made it possible to recover and print directly
the luminous rays emitted by a variously lighted object” (Bazin 2005a: 14). The photographic index, as
defined by Peter Wollen in the context of C.S. Peirce's semiology, constitutes a sign “by virtue of an
existential bond between itself and its object” (Wollen 1969: 122). In analogue photography, the index forms
a dynamic (and spatial) connection with a visible object and the senses or memory of whoever sees it: hence,
in Peirce's definition, indexicality denotes “a real connection” between object and sign that asserts nothing,
and the fact of the dynamic connection is not declarative but simply is (Peirce 2001: 109, 111). This
interpretation of the unique relation of the index to the real has formed an adaptable basis not only of Bazin's
197
theory but concepts such as Paul Willemen and Christian Keathley's 'cinephiliac moment' and Kristin
Thompson's notion of 'cinematic excess': that is, the visible registration of a 'concrete reality' or material
trace to which realist cinema turns its attention, whether it is captured by analogue technology or not.
As Rodowick argues, however, “digital acquisition” collapses the analogue function of the
photographic index by quantifying the world “as a manipulable series of numbers,” and reducing the
impression of light taken by the camera to an “abstract symbolization” that is sampled and encoded rather
than sensitised and printed (Rodowick 2007: 116). The digital image is a graphic 'assemblage' subject to both
numerical processing and potentially infinite manipulation, and comes into being only in correspondence
with memory locations on hard drives, and the pixel count of electronic screens. Hence, as Lev Manovich
contends, “once live action footage is digitised (or directly recorded in a digital format), it loses its privileged
indexical relationship to pro-filmic reality. […] Live action footage is reduced to be just another graphic, no
different than images which were created manually” (Manovich 2001: 254).
Some theorists have countered this potentially overbearing assumption that quantification has
entirely supplanted indexicality, however. As Philip Rosen questions, if we are to assign a totality to digital
utopia, “how is it that a digital camera can be sold not as a displacement, but as a replacement for a
conventional still camera?” Rosen proposes that the digital camera “does not necessarily exclude all the
operations of a photochemical camera; it even uses a lens to gather light,” and therefore “retains indexical
import as a light-sensing device” (P. Rosen 2000: 308).80
In Rosen's estimation, with which I concur, the
digital image cannot be considered to be non-indexical as it retains the critical automatic aspect of analogue
image production. Rosen's contention is perhaps validated by the inference of Hollis Frampton's claim that,
as analogue “photographic materials are, in the thumbprint sense, intractable,” it is impossible “to make a
photographic image that is a picture of nothing” (Frampton 2009: 72). Whereas the graphic manipulability of
digital media means it is now possible to create a digital image from nothing, it is still quite difficult to take a
light-sensing image of the world that is of nothing.81
80
To support this claim, Rosen also lists the methods by which it is possible to fake photographs without resorting to
digital manipulation, thus downplaying the impact of graphism on the production of 'false' images (see P. Rosen
2007: 320-321).
81
Although, I should note, “quite difficult” is not the same as impossible. The passage of light through a lens onto film
stock, even if it does not register as a representation of anything other than that process, formulates a material
impression that is explored in modes such as experimental and structural/materialist film (see Gidal 1978). Digital,
lacking this analogue basis, cannot accommodate a materialist practice, as a digital film does not (simultaneously)
form a physical “record (not a representation, not a reproduction) of its own making” (Gidal 1978: 2).
198
Alongside a diluted indexical function, the digital image also retains another key aspect of Bazinian
ontology: the automaticity of image-sound production that grants freedom from human interventions that
only “cast a shadow over the image” (Bazin 2005a: 12). Stanley Cavell describes this distinction between
manual creation and machinic indexicality as a distinction between the fact that whereas “a painting is a
world; a photograph is of the world”—a dictum that is echoed by Susan Sontag's comment that as “the
painter constructs, the photographer discloses” (Cavell 1979: 24; Sontag 2002: 92). Digital presents no
challenge to the basic automaticity of image production, meaning that, as Shaviro notes in his valuable essay
on this subject, “Emotion Capture”, “to the extent that digital technologies reinforce the automaticity of the
cinematic apparatus, one can imagine Bazin welcoming them” (Shaviro 2007: 66).
Rather than clinging to a 'true' indexicality that would prescribe the ontological integrity of analogue
media, the field of slow cinema mostly continues to be characterised by Bazin's notion of a subjective
striving for realism on behalf of both the artist and spectator. This notion of striving is extended by Deleuze's
conception of modern cinema, in its emphasis on a 'thinking' of the world, rather than an analogue
transcription of it. For this reason, Dudley Andrew directs the Cahiers axiom that I cited earlier—that the
cinema has a fundamental rapport with reality and that the real is not what is represented—against
proponents of what he calls “an overconfident discourse of the digital” (D. Andrew 2010: 5). In order to
overturn the ontological primacy of analogue/digital discourse, Andrew's implication is that we must recall
Bazin's dictum that “every realism in art was first profoundly aesthetic” (Bazin 2005b: 25). The recent
technological convergence in the field of slow cinema seems to have been unified by this notion of realist
aesthetics, and digital has been accepted, above all, as a potentially superior tool that can be used by artists to
respond to the world through filmic practice.
I am inclined to suggest, then, that dwelling on a radical shift in the ontology of media in relation to
contemporary durational cinema risks overstating the effects that this shift might have had on the outcomes
of filmmakers' practice and style. In contemporary cinema, digital and analogue media co-exist within a
digital regime that has enabled the establishment of new durational forms and new methods of observational
practice. As Lav Diaz has stated: “Practice is the real medium ultimately. It is up to the maker, the creator, or
the artist. You qualify 16mm by application. You qualify 35mm by application. And you qualify digital by
application. […] There is only cinema” (Wee 2005).
199
In light of this, it is worth noting Rodowick's objection in The Virtual Life of Film that “one feels or
intuits in digital images that the qualitative expression of duration found in photography and film is missing
or sharply reduced” (Rodowick 2007: 118). Rodowick feels this to be the case in the work of contemporary
artists who have chosen to replace the analogue preservation of passing time with the cyclical 'processing' of
duration found in digital media. This means that, when Rodowick turns to the contemporary durational film
under discussion here—that is, “a fascination with the choreography of quotidian acts deployed in a
continuous duration”—he draws attention to artists who have continued to adopt 16mm and 35mm film as an
'artisanal' practice: Sharon Lockhart's work from Goshogaoka (1997) to NO, and the work of Sam TaylorWood,
Jeroen de Rijke and Willem de Rooij (see Rodowick 2007: 158-159).
Rodowick senses, rightly, that the allegiance to 16mm and 35mm in these filmmakers' works “could
also express a countervailing desire—that is, the yearning for duration and uninterrupted time, for perceptual
depth, and for a sensuous connection to physical reality in a universe dominated by simulation and
information saturation” (Rodowick 2007: 158). I would be inclined to add the work of Tacita Dean (whose
recent Film [2011], first projected onto a thirteen-metre white monolith in the Turbine Hall at the Tate
Modern in London for five months, is a bold plea for the safeguarding of 16mm production) in this context,
as well as works that preserve the singularity of passing time in extended takes, such as Straub-Huillet's
Itinéraire de Jean Bricard (2008). However, as with prior assumptions regarding the non-indexical capacity
of digital cinema, Rodowick tends to overstate the primacy of ontological concerns in contemporary
durational film. The aesthetics of the cinema under discussion here suggest that the artist and spectator's
yearning for the 'uninterruption' of time, and striving for perceptual depth (rather than a lack of indexicality),
can be equally desired and replicated (if not strictly on indexical grounds) in a fully digital regime. It is
perhaps telling that Rodowick summons only Sokurov's slyly spectacular Russian Ark (2003) in his
discussion of contemporary digital cinema, in turn bypassing other non-analogue durational films (prior to
the book's publication) that I have mentioned here.
200
Digital transitions in observational practice I
Analogue technology places both desirable and undesirable restrictions on the duration of sequence shots and
observational long takes. Filmmakers who have used small, inexpensive analogue equipment, such as Hutton
and Benning, have tended to integrate the temporal limitations of their technology (whether entirely
deliberate or not) into a material basis for the production of images. Since the 1970s, Hutton has used 16mm
Bolex cameras with the capacity to load a hundred feet of film, or 2.77 minutes of unbroken duration at 24
fps. Hutton has noted that, alongside considerations of cost, shooting with such limited capacity forces him
to gather footage sparingly whilst exploring environments, and he often shoots a ratio of only 2:1 when
making a short work, or sometimes 3:1 or 4:1 if fiscal circumstances permit. The exploratory activity of
gathering “raw” images of place is always the primary concern of Hutton's practice, and his editing process
solely consists of 'distilling' up to a couple of hours of footage into a fifteen to thirty-minute film to “give the
audience a more concentrated hit” (MacDonald 2008: 224-225).
Benning, when shooting his 'California Trilogy' of El Valley Centro (1999), Los (2000), and Sogobi
(2001), uses the same technological constraint as a structural device. By choosing to fill a 100ft reel for
every shot in each film, Benning subjects his practice of observing the Californian wilderness and Greater
Los Angeles area to a mathematical contingency. As he recalls, “I was interested in seeing what activities
take place in that two and a half minute period. It's the time it takes a rodeo champion to tie up three goats, a
cotton-picker to traverse half a field, and an empty goods train to pass straight through the shot” (Hebdige
2011). Each film of the California Trilogy consists of thirty-five shots, and many of them are structured by
ordered systems of activities similar to those that Benning describes: the measured traversal of vehicles
(bulldozers and harvesters) across the frame; a cargo ship gliding along the horizon that appears to be
overtaken by a car in the foreground; and a well-timed sequence that records, from a distance, roll being
called in one of the Valley's many prisons. Because of the structural precision of the disclosure of activity in
these shots, it is tempting to assume that certain events are choreographed to conform to the length of time
that the 100ft reels permit. However, Claudia Slanar notes that “Benning denies this. According to him, it is
only a matter of waiting for the right moment, which reveals itself through careful observation” (Slanar
2007: 169).
201
It is worth noting that Andy Warhol also used an electric Bolex for his seminal Screen Tests (1964-
1966), alongside other works such as the Kiss series and Mario Banana (nos. 1 and 2) (both 1964). Warhol
would load a separate hundred feet of film for each individual test (producing over two hundred in total),
and, when necessary, loaded reels over and over again to construct longer works such as Eat and Blow Job
(both 1964). Warhol instructed sitters for the Screen Tests to simply remain in place until the end of the reel
(regardless of the outcome, which provokes displays of restlessness, boredom, or tears), so that both the
duration and accidental structure of the films would be wholly dictated by the limitations of the apparatus.82
Warhol then slowed the films from 24 fps to 16 fps during projection, adding an extra 50% of running time
to decelerate the visible activity of the image (see Angell 2006: 21-22).
For his films that followed the California Trilogy, Ten Skies and 13 Lakes (started before, but
finished after, One Way Boogie Woogie: 27 Years Later [2004]), Benning loaded his electric Bolex with 400ft
magazines to extend the duration of shots to ten minutes (usually trimming a minute from the maximum
capacity of eleven during editing). After switching to digital, however, Benning, as previously documented,
has been able to shoot up to two (continuous) hours of HD video on a single memory card—a capacity that
can be increased by either inserting an extra card (or two), or adapting the camera to connect to an external
hard drive (an option that is easier with SD or MiniDV than HD, and produces a duration limited only by the
battery life of the camera). Benning has so far only worked within the “manageable” two-hour capacity of a
single memory card, and has commented that:
I love the luxury of being able to do the two-hour shot. [...] If I don't like it, I just erase the card and
use it again. So there's this luxury of recording shots that are much much longer than I was able to do
in the past because—first of all—I was limited by the reel size, but I was also limited by money
more than anything (Guillen 2010).83
As mentioned above, digital has enabled Benning to shift his attention from systems of activities that
82
Warhol called these films “stillies,” and they probably constitute the first modern single-shot films, a tradition that
extends to Kiarostami's Five, Lockhart's NO and Benning's Nightfall (MacDonald 2008: 241).
83
Benning switched to digital for another key reason: a need to counteract declining standards in 16mm lab
processing, and the cost of errors made by projectionists (such as damaging a fresh, irreplaceable print during only
its second or third screening) (see Peranson 2010: 56).
202
last minutes (as in the California Trilogy) to ones that take considerably longer, such as the filling and
shunting of a slag train between two tracks in Pig Iron (a process that takes half an hour), and the complete
transition from dawn to dusk that takes an hour and a half in Nightfall. It is worth noting that some
contemporary filmmakers have lamented analogue film's inability to permit the representation of events in
such extended, uninterrupted duration: for instance, Béla Tarr has complained that no shot in his work lasts
much longer than “ten minutes and twenty seconds,” because “Kodak cannot make [a 35mm magazine]
longer than three hundred metres.” Tarr's films do not contain many sequence shots that occupy the full
length of a reel, and the longest unbroken takes in his work are a mobile close-up of Irimias' speech in the
seventh chapter of Sátántangó, which lasts ten and a half minutes, and the opening sequence of Werckmeister
harmóniák (2000), in which Valuska stages an eclipse with his drunken comrades. However, Tarr has joked
that the time limit of “this fucking Kodak” constitutes “a kind of censorship,” and his only experiment with
video yielded a film of two shots: Macbeth (1982), in which the second take runs for sixty-seven minutes
(Schlosser 2000).
Some contemporary filmmakers have chosen to continue using analogue technology for its material
discipline. Ben Rivers, who has recently entered the informal slow canon with Two Years at Sea (2011),
filmed his early experimental work (up to Ah, Liberty! [2008] and I Know Where I'm Going [2009]) with a
1970s clockwork Bolex. Like Hutton, Rivers has noted that the thirty-second time constraint delivered by the
Bolex (after being wound) requires a particularly attentive, discerning observational practice: “It creates
mini-rules and more concentration and consideration are needed. It makes filming less arbitrary, it helps you
think about what you're doing” (Corless 2008). Patrick Keiller has echoed this sentiment in relation to the
third film in his 'Robinson Trilogy', Robinson in Ruins (2010), for which he travelled across England to
capture static shots of built environments (in Oxford and London) and bucolic landscapes (of military sites or
agricultural land, mostly in Oxfordshire and Berkshire). Keiller, as with all marginal filmmakers, invokes the
cost of purchasing and processing film stock as a primary practical constraint, but also suggests that it
necessarily sharpens the production of image and sound. Keiller notes that, when using a camera magazine
that holds only 122m of stock (just over four minutes at 25 fps), shooting on 35mm “tends to involve a
greater commitment to an image before starting to turn the camera, and there is pressure to stop as soon as
203
possible, both to limit expenditure and to avoid running out of loaded film” (Keiller 2009: 412).84
It is interesting, then, to note that Kiarostami takes an approach entirely to the contrary after adopting
digital in Five. Kiarostami does set up (or, more accurately, prepares) the field of (in)activity in image-sound
situations, but, after activating the camera, retreats from intervention by flaunting the lack of pressure to stop
filming (which is no longer restricted by fiscal considerations). In Around Five, Kiarostami describes his
method of filming the seventeen-minute duration of Dogs: “It was an opportunity for me to be the audience.
My duty as the director ends precisely when I start the camera. Normally, the director's role should start
when shooting begins, but there was no input from me during the making of it. I switched the camera on and
then I went to sleep.” Digital media's loosening of temporal constraints replaces the need for active
intervention with an accumulation of duration, as long as the artist continues to trust in the automaticity of
the camera's recording mechanism.
Kiarostami's creative act in Dogs involves little more than an establishment of “the right conditions”
in which (in)activity might unfold—in this case, the registration of passing time, and the rising sun that
bleaches the frame beyond recognition. Kiarostami's disengagement whilst filming suggests a hyberbolic
modification of the Bazinian mode of encountering the real—in which, as Shaviro notes, “the cinematic artist
must [...] step aside from the image that he/she helps to bring into being, letting the machine do its work
without imposing subjective interpretations on it” (Shaviro 2007: 66). Observational filmmakers, after
converting to digital, appear to be retreating from the materialist discipline of filmmakers like Keiller, in
order to heed an implication expressed in one of Stanley Cavell's automatisms: that of drawing on the
moving image's potential “to let the world happen, to lets its parts draw attention to themselves according to
their natural weight” (Cavell 1979: 25; my emphasis).
84
Keiller also reveals that he also chose 35mm for the purpose of material conservation, and argues that “as electronic
image and digital file formats are prone to rapid obsolescence,” digital images “require continual management. [...]
In comparison, a photographically originated ciné frame is a reassuringly visible, physical original in a format that
has survived for over 110 years and seems likely to continue, if only as a basis for successive electronic copies”
(Keiller 2009: 413).
204
Digital transitions in observational practice II
As well as enabling an increased capacity for duration within the sequence shot, the impact of digital
technology on contemporary cinema has also given rise to a radically new mode of observational practice,
and a resultant 'hybrid' aesthetic described by Robert Koehler as the 'cinema of in-between-ness' in my first
chapter. Pedro Costa's cinema of the last decade (between In Vanda's Room [2000] and O Nosso Homen
[2010]) is a clear instance of this shift in production, which is exemplified by Costa's statement that “going
to video changed my way of seeing. It's exactly the same problem of filmmaking, but with more freedom and
time” (Peranson 2006: 13). Before In Vanda's Room, Costa's first three fiction features were shot on 35mm,
and his transition to digital was directly inspired by the circumstances and experience of his last analogue
production, Ossos (1997).
On returning to Lisbon after filming his second fiction feature, Casa de Lava (1994), in Cape Verde,
Costa delivered a number of letters and gifts (wine, tobacco, and so on) from residents of the island to Cape
Verdean immigrants living on the outskirts of Lisbon. The Cape Verdean immigrant community in Lisbon
was, at the time (and, largely, still is), concentrated in the neighbourhood of Fontainhas, in a slum estate
(since demolished) that housed about five thousand people subsisting in tight, cramped, often destitute
spaces. After delivering the letters and gifts from friends and relatives, Costa began to travel to the
neighbourhood regularly, hanging out and drinking with the residents. After a year or so, he started work on
Ossos, which was primarily shot in the neighbourhood (only glimpsing the centre of Lisbon) with a few of
the residents, including Vanda and Zita Duarte, who star in all three Fontainhas films (Ossos, In Vanda's
Room and Colossal Youth [2006]). Ossos was a standard art film production in every sense, bringing with it a
tight budget, sizeable professional crew, assistants, cars, floodlights and metres of camera tracks. After facing
the difficulties of shooting in the neighbourhood's extremely narrow corridors, stairwells and alleyways, as
well as rooms that were less than six feet wide, Costa later admitted that “the neighbourhood refused this
kind of cinema, it didn’t want it” [Fig. 51]. Cyril Neyrat notes that, during the filming process, Costa's
impression was that there was “too much squalor and despair in front of the camera; too much money,
equipment, and wasted energy behind it. And too much light shining in the night of a neighbourhood of
manual labourers and cleaning women who got up at 5:00 a.m.” (Neyrat 2010: 12).
205
Fig. 51: Ossos (1997)
Ossos was filmed in a controlled long-shot long-take style, and Costa has since stated that he found it
difficult to maintain the necessary rigour of observation during production:
I saw only 20% of the things that I should have been seeing every day because my eyes were
attracted to the guys in the crew or whatever; the means and the ends weren't thought through
correctly. So I thought to myself I had to do things another way. And this led me to think that the
normal way of making films was all wrong. We should rethink all of it. What are second, third
assistants doing on the set? (Peranson 2006: 12).
A year later, Costa abandoned the equipment and procedure of professional production, and returned to
Fontainhas with no money and a small, cheap Panasonic DV camera. After spontaneously gathering footage
of the lives of heroin addicts Vanda and Zita for an unplanned documentary, Costa, over the course of six
months, began to completely alter his observational approach: “I was there trying to catch things with my
camera, and then I slowly realised I was there to lose moments, not to catch them. Video is very good for
206
that. It's made to go slower and slower, and not to try and catch reality, but to try and lose reality in a way”
(Peranson 2006: 13; my emphasis).
Costa filmed Vanda and Zita (as well as Ventura, star of Colossal Youth, and other friends and
addicts) for another six months, setting the camera against a wall or in the corner of a small room, and letting
it register the passage of time as they talk, laugh, cough, and shoot up for hours, sometimes from day until
night. The patience of Costa's practice derives from its immobility of observation, in which both the
filmmaker and camera slowly take in routines, behaviours, and systems of everyday activities (such as the
times at which Vanda, Zita and friends wake, eat, work, shoot up, sleep, visit friends and family, or simply
move from room to room in domestic spaces). The digital footage compiled during this time became In
Vanda's Room, and Costa employed the same observational strategy for his next film, Où gît votre sourire
enfoui? (2001), which documents Jean-Marie Straub and Danièle Huillet's editing of Sicilia! (1999) at the
Studio National des Arts Contemporains in Le Fresnoy, Tourcoing.
In Où gît votre sourire enfoui?, Costa primarily films Straub and Huillet's practice by fixing a camera
(one of two) in the corner of their editing suite, which points toward a Moviola editing table in the centre of
the frame, and a door that is rarely closed to its left. As Huillet works at the table, Straub repeatedly paces in
and out of the doorway, staging a series of entrances and exits that strategically punctuates his dialogue, and
self-edits his performance within the restricted space of the room and corridor. Costa, accompanied by an
assistant with a second camera, filmed Straub and Huillet for five weeks (from nine o'clock in the morning
until seven o'clock in the evening), and produced over a hundred hours of footage that were carefully and
'dishonestly' (Costa's term) edited to give the impression of ordered dialogue and activity (see Kasman
2009).
For In Vanda's Room, Costa accumulated 140 hours of footage (only half as much as for Colossal
Youth), and excised “137 hours of rubbish” in the space of a year to complete the film (Peranson 2006: 13).
The patient observation and production of surplus footage in Costa's digital practice is analogous with
Straub-Huillet's own tendency of filming entire reels whether the purpose of the shot has been exhausted or
not, piling up time that is then painstakingly filtered later. Digital technology facilitates this process with far
greater ease, and at an immensely reduced cost, than 16mm or 35mm production, as filming is restricted only
by the length of a camera's battery life or the amount of data that can be held on a memory card. The
207
slowness of digital that Costa describes, more than anything else, is its potential for what Lyotard refers to as
'useless expenditure': the ability to lose moments rather than maximise them, by passing and 'wasting' as
much time as possible during filming.85
The ability to produce a surplus of digital image-sound situations for little material cost has other
advantages for contemporary filmmakers. Lav Diaz's adoption of digital after Hesus Rebolusyunaryo (2002)
has led to a series of radical experiments in distended narrative structures, and his long-form works have
become somewhat notorious for their extended running times: Evolution of a Filipino Family (ten and a half
hours), Heremias Book I: The Legend of the Lizard Princess (2006; nine hours), Death in the Land of
Encantos (2007; nine hours), Melancholia (2008; eight hours), Century of Birthing (2011; six hours) and
Florentina Hubaldo, CTE (2012; six hours). In Diaz's work, this 'excessive' length in itself functions as a
reflection of acinema—as he noted in a roundtable discussion in Bangkok in 2009, “convention tells you that
a film has to be two hours, mainly for commercial purposes. If you can screen a film seven times in a day,
that’s maximum profit. But I don’t have anything to do with commerce or the marketplace, I just make my
films” (Diaz et al 2011).
Diaz is an elder member of the Philippine New Wave, a contemporary independent cinema that, over
the course of the last decade, has revitalised the nation's film practice.86
Alexis Tioseco, in an important early
overview of the movement published in Inter-Asia Cultural Studies in 2007, observes that Filipino
filmmaking has been transformed by the affordability of digital technology (as well as innovations in editing
software and ease of distribution): “Today’s filmmaker, no longer working under the pretences of a
controlling martial law regime and graced with the tools that make working entirely outside the system a
genuine possibility, need only to battle themselves [...] in order to make the films they want to make.”
Tioseco adds that, in the last decade, “many younger filmmakers, witnessing the difficulty their fellow
directors have endured when attempting to work in the star-driven studio system, have chosen to go the
independent route, sacrificing financial security in the name of artistic integrity” (Tioseco 2007: 298).
85
Lisandro Alonso, a staunch defender of 35mm production, has echoed this notion by complaining that “you can
spend hours and hours and hours” filming on video, rather than understanding “what it means to use the focus and
the timing and the lighting. Video is not the same thing” (Anderson 2001: 37).
86
If not, as the late Alexis Tioseco lamented, its film culture, which he accused of being “much weaker than the films
are at this point. A lot of interesting films are being made, but there is not a culture which supports them” (Diaz et al
2011). In 2008, Tioseco made a set of demands to resolve this set of affairs entitled Wishful Thinking for Philippine
Cinema: http://alexistioseco.wordpress.com/2009/03/15/wishful-thinking-for-philippine-cinema/
208
This younger generation of filmmakers includes, amongst others, Raya Martin, Adolfo B. Alix Jr.,
Khavn De La Cruz, Sherad Anthony Sanchez and John Torres, as well as Diaz himself, after his
abandonment of studio production in precisely the circumstances that Tioseco describes.87
Diaz's first
independent production, Evolution of a Filipino Family, was financed solely by himself and Paul Tañedo
(Diaz's friend, and the film's producer), and took eleven years to shoot without a fixed script or planned
structure (during which time three of the original actors died).88
Alix Jr. and Khavn have been the most
prolific filmmakers of the New Wave, shooting multiple films a year (Khavn has recently completed the first
documentary on the movement, Philippine New Wave: This is Not a Film Movement [2010]), but Diaz has
probably produced more footage for exhibition if measured on an hourly basis.
The production of duration is central to Diaz's digital cinema in every sense: on the level of narrative
(which usually takes the form of a network narrative that integrates memories of the past with stories of the
present and future); in the undramatic presentation (or refusal) of content within individual shots; and in
terms of the time that it takes for spectators to see the films when projected—time that must be extracted
from other everyday activities, or from the tight schedules of international film festivals that screen the films
in full.
Diaz insists on projecting individual works without a break so that the films occupy a sense of actual
time for the spectator, and programmers of major festivals (such as Toronto and Rotterdam, who habitually
screen and, in the case of the latter, fund Diaz's work) accommodate this request by screening the films in an
open installation space where food and drink are provided, and where people are encouraged to come and go
during the projection. Alternatively, programmers and projectionists tend to intervene by positioning an
interval when DigiBeta tapes are changed, which occurs every two hours. Diaz is always careful to structure
the narration of his films to conform to the two-hour capacity of the DigiBeta L Tape, which explains the
evenness of his films' hourly running times, and means that each two-hour segment tends to act as a discrete
chapter of the narration (as well as helping to ensure that no long takes are interrupted or curtailed by the
87
Diaz's first five films were studio productions, a mode that he abandoned after Hesus Rebolusyunaryo and a
particularly disastrous experience with Batang West Side—the negative of which is still embargoed by the film's
AWOL producer (meaning that the only copy Diaz now has access to is a degraded festival screener, complete with
ticker and watermark, that originated on VCD).
88
Diaz experienced a notable pitfall of digital production whilst completing Evolution: when adding subtitles to a
rough final cut of the film in July 2004, a hard drive crashed and wiped the entire edit. Diaz had to start all over
again, and only completed another final cut in December 2004. The complete film was premièred at the University
of the Philippines in December 2004, then screened at Rotterdam Film Festival (IFFR) in January 2005.
209
Fig. 52: Death in the Land of Encantos (2007)
transition between tapes). Hence, the content of Diaz's works are constructed with the duration of their
unique exhibition in mind, and are designed to be quite literally lived with (either through endurance or
distraction) as they are projected.
After Evolution of a Filipino Family, Diaz's video works disclose activity in very few, often static,
unbroken shots. For instance, Death in the Land of Encantos and Melancholia contain only 166 and 145
shots each, and possess respective ASLs of 196.1 and 180.2 seconds. Their narrational patterning
consistently foregrounds sequence shots of characters (always seers) simply waiting for melodramatic
incident or walking substantial distances (for hours, in the case of Heremias Book I); static long shots of
extended monologues or conversations; and inexorable displays of visceral emotion (often of grief or mania
—a persistent theme in Encantos, Melancholia and Century of Birthing) [Fig. 52].
One of Diaz's habits is to stage offhand dialogue scenes that exhaust any putative narrative function.
Encantos features a number of these scenes, and the longest lasts nineteen minutes: a sequence shot in which
Benjamin, Teodoro and Catalina sit around a table at night, dimly lit by candles, to discuss art, philosophy,
religion and nationalism. The off-the-cuff flow of dialogue initially gives the impression that the scene is
210
improvised, but they are always scripted in full by Diaz, who tends to write drafts the night before filming
and prepare them with his actors in a morning before shooting in the afternoon. There is, as a result, an
intentional rawness to both the actors' performance and ideas discussed, which Diaz will record in only a
single take or sometimes a ratio of 2:1 or 3:1 (see Tioseco 2006). Diaz does not predetermine the length of
such conversational scenes for either want or need: his practice is to fix a point of observation, and let the
dialogue exchange play out for as long as it must take. This laxity of planning extends to the shaping of a
complete work, as Diaz has stated:
I don't really think about length when I make films. I'm a slave to the process, following the
characters and the story and where they lead. It's a very organic process for me, I just keep shooting
and shooting once there's an idea. When I watch the footage later, if I think there’s still more to be
done I have to shoot it. I don’t think, 'oh, it’s already seven hours,' or 'there’s already fifty hours of
footage' (Diaz et al 2011).
The reduced cost of digital production also enables Diaz to reshoot entire films at will. Diaz's most
recent (at the time of writing) long-form work, Florentina Hubaldo, CTE, is a comprehensive HD reshoot of
an SD work originally titled Agonistes, Myth of a Nation, of which a (supposedly final) five-hour version was
completed in 2009. Agonistes was screened at an extensive retrospective of Diaz's work in Bangkok in 2009
(entitled Death in the Land of Melancholia, organised by a group of Thai cinephiles known as the Filmvirus
network), and an essay written by one of the attendees of that screening, Wiwat Lertwiwatwongsa, comprises
what will likely now be the only detailed record of it (see Lertwiwatwongsa 2012).
Equally, Diaz takes advantage of the freedom of digital by starting and delaying projects for long
periods of time, such as his sequel to Heremias Book I (2006), Book II: The Legend of Tagabulag Island
(2008), which was mostly shot in the second half of 2006 but then postponed due to the absence of Ronnie
Lazaro, who played the character of Heremias in Book I. Heremias Book II is occasionally screened in its
current form as a two-hour work, but Diaz has intended to restart production (or, more likely now, reshoot)
when Lazaro becomes available for filming again. This tendency to accumulate a surplus of image-sound
situations for overlapping and unfinished projects leaves Diaz with material that he often integrates into
211
long-form works in innovative ways: for instance, Encantos incorporates footage of an unfinished short film
shot in Zagreb (in 2003) in the form of a woozy flashback that mirrors Benjamin's gradual loss of sanity, and
Century of Birthing incorporates an older SD project as the set of rushes that the main protagonist, a
filmmaker himself, is struggling to edit. In Century of Birthing, Diaz stages a couple of long takes in which
we observe the filmmaker painstakingly edit this work-in-progress, struggling to whittle down a vast amount
of footage to a meaningful complete film. The reflexivity of this joke does not escape the spectator, even if
Diaz's vision in this work is, altogether, unremittingly bleak: by the end of the film, the filmmaker's attempt
to escape from the project's incompleteness is one of the factors that drives him to lose his sanity entirely.
I have focussed on the shift from analogue to digital production in detail here to give a sense of the
transitional state of contemporary durational cinema, and to address the centrality of practice to emergent
forms of observational realism. This latter factor is often easily overlooked, but, I would suggest, must be
considered integral to a realist aesthetics as modest and subjective as Kiarostami's Five and James Benning's
recent work, or as contingent on the actuality of (collaborative) practice as Pedro Costa's hybrid cinema. In
light of the emergence of these new modes of practice (now fully encompassed by the digital regime), one
aspect of Bazin's theory of durational realism appears to be as relevant as ever: that any definition of the
filmic medium and its realist aesthetics must be formed through the artistic responses to the world that it
makes possible, rather than a material history that it aspires to simply transcribe.
212
CONCLUSION
My aim in this thesis has been to map the network of stylistic devices, observational tendencies and
(broadly) realist aesthetics that comprise the basis of 'slow cinema'. In my introduction, I concurred with
Tiago Magalhães de Luca that the ambitious grouping of international films and filmmakers under discussion
here could not be defined as a “structured film movement” that transcends boundaries of national (or
transnational) production, or the divergent aesthetic concerns of individual artists and collectives (de Luca
2011: 24). My aim has, equally, been to suggest that all films considered here as 'slow' cannot be defined
solely by their seeming opposition to dominant cinema, and that they do belong to quite specific geographic,
cultural and economic traditions. To this end, I have primarily organised the structure of this thesis around a
selection of related yet diverse filmmaking practices, and a variety of critical, theoretical and historical
approaches that, I feel, might be productively employed in order to make sense of the artworks under
discussion. For the purpose of concluding this study, then, I will briefly summarise the dominant concerns of
my approach.
Implicit in the organisation of this thesis is the conviction that a new cinematographic form can be
located across a range of otherwise divergent national contexts and cinemas, and that the primary connection
between such 'slow' films can be thought to arise from their common aesthetic sympathies. In order to point
to this shared commonality, I have employed a number of terms that foreground the key mutual aspects of
slow cinema: 'durational', 'observational', 'undramatic', and so on. The qualitative similarities between films
that display these formal, narrational and realist aspects do not stem from a shared industrial base or national
context of production, and are therefore best defined on the level of concept and style. The application of
Bazinian criticism to my case studies of individual films and shots in Chapter Two, and more detailed
consideration of observational style in both narrative and non-narrative film in Chapter Four, should,
hopefully, have demonstrated this sense of aesthetic commonality derived from a collective application of the
long take.
In Chapter One, I offered a tentative map of the contours of slow cinema by specifying its most
distinctive qualities: the merging of canonically divergent modes of address, and a visible overlapping of
four established modes of representation—the fiction, documentary, art and experimental film. Whilst the
213
most obvious factors that unify these modes in slow cinema are the application of the long take and a
tendency toward undramatic activity or narrativity (if narrativity is present at all), a large part of the main
body of this thesis has attempted to locate an 'art of realism' (to borrow Denis Lévy's term) as a dominant
tendency in the field of slow cinema. This attention to realism is addressed directly in Chapter Two, as well
as in my brief analysis of undramatic narrative and everyday films (including Tsai Ming-liang's Vive l'amour
[1994], Lisandro Alonso's La libertad [2001], and Liu Jiayin's formalist Oxhide [2005] and Oxhide II [2009])
in Chapter Three, and the distinctive style of Hou Hsiao-hsien's Café Lumière (2003) in Chapter Four. My
interpretation of the notion of an 'art of realism' (which, as Lévy notes, was “contrarily maintained” in the
transition between classical and modern cinema) has comprised three primary aspects: a realism of duration
(that is, of an impression of actual time within the shot), an unmediated or materialist mode of observation,
and a rigorous attention to the everyday (in terms of both narrativity and the depiction of objects in the
world) (Lévy 2011).
My first chapter also addressed the significance of the notion of 'the wind in the trees' in relation to
the realism of slow cinema—that is, the particular resonance of halted time, and rigorous observation of
body and place, that arises when more familiar conventions of classical or intensified continuity style are
overturned. Foregrounding this tendency to dwell upon depictions of material phenomena and everyday
gestures serves to bind together both the core observational tendencies of slow films, and the structure of my
analyses throughout this thesis. It is perhaps a prerequisite of formal analysis in this field that the attention
such events are afforded by artists should, and must, be replicated by the writer in response.
Throughout the main body of the thesis, it has also been my aim to map post-war durational cinema
by departing from some perceived limitations of canonical models, and to directly address the issue of the
production of duration in contemporary film and culture. This emphasis on temporality unifies the otherwise
divergent central aspects of Chapters Two, Three and Four: in my second chapter, I turned to the durational
application of the Bazinian sequence shot in contemporary films, whereas my third chapter surveyed slow
cinema's production of a 'wasteful' excess of duration in relation to the intensified spaces and flows of
neoliberal culture. In my fourth chapter, I suggested that slow cinema visibly embodies Gilles Deleuze's
conception of the time-image, in which time rises up to the surface of the screen to enable a restoration of the
dislocated link between man and world in the post-war era. Slow cinema is, I would suggest, uniquely
214
concerned with this latter aspect, and can, without reservation, be said to manifest a belief in the world that
aspires to “reconnect man to what he sees and hears” (Deleuze 2005b: 166).
To anchor this attention to the production of duration in post-war film and culture, I have also
attempted to trace, in a somewhat piecemeal manner, the strictly historical evolution of slow cinema. In
Chapter Two, I employed Bazin's theory as a guide to the development of long-take style in pre- and postwar
modern cinema (discussing its roots in the work of, amongst others, Kenji Mizoguchi, Orson Welles and
William Wyler), whilst also proposing a strong link between Bazinian realism and the style of contemporary
filmmakers such as Jia Zhang-ke and Lav Diaz. Jia and Diaz are familiar with both Bazin's critical legacy
and his idea of realist cinema—a conceptual debt that is fully evident in even their most radical work. In
Chapter Three, I suggested that a notable association might be located between the moral commitment of
post-war Italian neorealism (in Cesare Zavattini's writing and collaborations with Vittorio de Sica) and the
“in-between” realist cinema of the present (such as Lisandro Alonso's 'outsider trilogy'—best described using
Robert Koehler's term, as outlined in my first chapter). As a counterpart to this historical legacy, the
remainder of Chapter Three's consideration of post-continuity style hopefully serves not only as a conceptual
tool to understand the nature of temporality in two poles of contemporary acinema, but as an account of key
developments in dominant and marginal film style in recent years. To complete this process, Chapter Four
deliberately links two contemporary tributes to Ozu with the modernity of their filmic precedent: the specific
formal device of the still life, which is thought by Deleuze in philosophical rather than formal terms, as an
image that discloses a little stretch of “time in its pure state” (Deleuze 2005b: 16).
I also hope to have illustrated, alongside this film-historical basis, that the conception of 'slow
cinema' is a useful framework not only for analysing the intersection of trends in contemporary cinema, but
also for the study of visual style and staging within individual durational works. In that respect, I have
inevitably been subject to considerable restraint, both in terms of the scale of reference that can be afforded
here, and in the space that has been available. I have, necessarily, been extremely selective with regard to
representative case studies in individual chapters, and many potential subjects for extended analysis (from
both art and experimental cinema) are noticeably lacking. A more comprehensive survey would turn to a
much fuller range of exemplary durational works, such as the minimalist narrativity of Paz Encina's
Paraguayan Hammock (2006) (a film that draws on a scenario reminiscent of Samuel Beckett's Waiting for
215
Godot [first performed in full in 1953] by prioritising the gentle bickering between two characters, and
perceptible recession of daylight, over active narrativity), the de-dramatised art cinema of José Luis Torres
Leiva’s The Sky, the Earth, and the Rain (2008) (cited by James Quandt in my first chapter as an archetypal
'festival film', replete with attenuated takes, long tracking and panning shots of depopulated landscapes and
isolated people walking, dollies to windows and open doors framing nature, and “a preponderance of
Tarkovskian imagery”), or a work as stark and monumental as Wang Bing's Crude Oil (2008), a fourteenhour
video installation that details oil extraction in a remote landscape in Qinghai, China, for a duration that
approximates the actual time of a working day (Quandt 2009b: 76).
Despite these inevitable omissions, however, the purpose of this thesis has been, above all, to show
that a wider convergence is taking place in the production of duration in contemporary film and culture, and
that this has, in turn, formed new aesthetic, critical and even socioeconomic effects. The evolution of
durational cinema over the course of the last three decades has been one of the most significant
developments in contemporary international cinema, and it is a dynamic that will, on the basis of the research
presented here, most likely continue to strengthen in years to come.
216
WORKS CITED
Adorno, Theodor W. and Horkheimer, Max. Dialectic of Enlightenment. Trans. John Cumming. London;
New York: Verso, 1997. Print.
Albera, François. “'Sickle and Hammer, Cannons, Cannons, Dynamite!': Conversation with Jean-Marie
Straub and Danièle Huillet.” Die Früchte des Zorns und der Zärtlichkeit: Jean-Marie Straub / Danièle
Huillet und ausgewählte Filme von John Ford. Vienna: Österreichisches Filmmuseum, 2004. 41-50.
Print.
Andermann, Jens. New Argentine Cinema. London; New York: I.B. Taurus, 2012. Print.
Andersen, Thom. “Against the Grain.” Film Comment 45.4 (2009): 25-27. Print.
----. “Painting in the Shadows.” Film Comment 43.2 (2007): 56-61. Print.
Andersen, Thom and Costa, Pedro. “No Secrets, Just Lessons: A Dialogue Between Pedro Costa and Thom
Andersen on Danièle Huillet and Jean-Marie Straub.” Cinema Scope 29 (2007): 40-45. Print.
Anderson, Jason. “Taste of Armadillo: Lisandro Alonso on La libertad.” Cinema Scope 9 (2001): 36-38.
Print.
Andrew, Dudley. André Bazin. New York: Oxford UP, 1978. Print.
----. What Cinema Is!: Bazin's Quest and its Charge. Chichester: Wiley-Blackwell, 2010. Print.
Andrew, Geoff. “Interview: Béla Tarr.” Sight and Sound 17.7 (2007): 19. Print.
Angell, Callie. Andy Warhol Screen Tests: The Films of Andy Warhol Catalogue Raisonné Volume 1. New
York: Abrams; Whitney Museum of American Art, 2006. Print.
----. The Films of Andy Warhol: Part II. New York: Whitney Museum of American Art, 1994. Print.
Angelopoulos, Theo. Theo Angelopoulos: Interviews. Ed. Dan Fainaru. Mississippi: Mississippi UP, 2001.
Print.
Antonioni, Michelangelo. The Architecture of Vision: Writings and Interviews on Cinema. Chicago: Chicago
UP, 2007. Print.
Arnold, Liz. “Interview with Jonas Mekas, Cameraman for 'Empire'.” 18 February 2011. WNYC Culture.
Web. 19 February 2011. <http://culture.wnyc.org/articles/features/2011/feb/18/interview-jonas-mekas-
empire-cameraman/>
Arrighi, Giovanni. The Long Twentieth Century: Money, Power and the Origins of Our
Times. London; New
York: Verso, 2010. Print.
Arthur, Paul. “No Longer Absolute: Portraiture in American Avant-Garde and Documentary Films of the
Sixties.” Rites of Realism. Ed. Ivone Margulies. Durham and London: Duke UP, 2002. 93-118. Print.
Balsom, Erika. “Saving the Image: Scale and Duration in Contemporary Art Cinema.” CineACTION 72
(2007): 23-31. Print.
Barthes, Roland. Camera Lucida. Trans. Richard Howard. London: Vintage, 2000. Print.
217
----. “The Reality Effect.” Trans. R. Carter. French Literary Theory Today. Ed. Tzvetan Todorov. Cambridge:
Cambridge UP, 1982. 11-16. Print.
----. “The Third Meaning: Research Notes on Some Eisenstein Stills.” Trans. Richard Howard. Artforum 11.5
(1973): 45-64. Print.
Bazin, André. Bazin at Work: Major Essays and Reviews from the Forties and Fifties. Ed. Bert Cardullo.
Trans. Allan Piette and Bert Cardullo. New York: Routledge, 1997. Print.
----. “Cinema and Television: Jean Renoir and Roberto Rossellini interviewed by André Bazin.” Sight and
Sound 28.1 (1958-9): 26-30. Print.
----. “Death Every Afternoon.” Trans. Mark A. Cohen. Rites of Realism. Ed. Ivone Margulies. Durham;
London: Duke UP, 2003. 27-31. Print.
----. French Cinema of the Occupation and Resistance. Trans. Stanley Hochman. New York: Frederick
Ungar, 1981. Print.
----. Jean Renoir. Ed. François Truffaut. Trans. W. W. Halsey II and William H. Simon. Da Capo Press, 1992.
Print.
----. Orson Welles. Trans. Jonathan Rosenbaum. Los Angeles: Acrobat Books, 1991. Print.
----. What is Cinema? Volume I. Trans. Hugh Gray. Berkeley; London: California UP, 2005. Print.
----. What is Cinema? Volume II. Trans. Hugh Gray. Berkeley; London: California UP, 2005. Print.
Beasley-Murray, Jon. “Whatever Happened to Neorealism?—Bazin, Deleuze, and Tarkovsky's Long Take.”
Iris 23 (1997): 37-52. Print.
Benjamin, Walter. “The Work of Art in the Age of Mechanical Reproduction.” Trans. J.A. Underwood. Oneway
Street and Other Writings. London: Penguin, 2009. 228-259. Print.
Benning, James. “NO.” Cinema Scope 42 (2010): 96. Print.
Benning, James and Lockhart, Sharon. “James Benning Interviews Sharon Lockhart.” Sharon Lockhart:
'Lunch Break'. Ed. Sabine Eckmann. Chicago: Chicago UP, 2010. 100-109. Print.
Benning, James and Nord, Cristina. “Ruhr: James Benning im Gespräch.” CARGO, 12 November 2009.
Web. 15 January 2010. <http://www.cargo-film.de/thema-reihe/us-indie/ruhr-james-benning-im-
gesprach/>
Berry, Chris. “Where is the Love? Hyperbolic Realism and Indulgence in Vive l'amour.” Island on the Edge:
Taiwan New Cinema and After. Eds. Chris Berry and Feii Lu. Hong Kong: Hong Kong UP, 2005. 89-
100. Print.
Berry, Michael. Speaking in Images: Interviews with Contemporary Chinese Filmmakers. New York;
Chichester: Columbia UP, 2005. Print.
Betz, Mark. “Beyond Europe: On Parametric Transcendence.” Global Art Cinema: New Theories and
Histories. Eds. Rosalind Galt and Karl Schoonover. Oxford; New York: Oxford UP, 2010. 31-47. Print.
----. Beyond the Subtitle: Remapping European Art Cinema. London; Minneapolis: Minnesota UP, 2009.
Print.
Bickerton, Emilie. A Short History of Cahiers du Cinéma. London; New York: Verso, 2009. Print.
218
Bingham, Adam. “The Romance of Certain Old Clothes, or They Don’t Make ‘em Like That Anymore.”
CineACTION 75 (2008): 34-45. Print.
Bíró, Yvette. “The Fullness of Minimalism.” Rouge 9, 2006. Web.
<http://www.rouge.com.au/9/minimalism.html>
----. Turbulence and Flow in Film: The Rhythmic Design. Trans. Paul Salamon. Bloomington; Indianapolis:
Indiana UP, 2008. Print.
Bogue, Ronald. Deleuze on Cinema. New York; London: Routledge, 2003. Print.
Bordwell, David. “The Art Cinema as a Mode of Film Practice.” Film Criticism 4.1 (1979): 56-64. Print.
----. “Beyond Asian Minimalism: Hong Sangsoo’s Geometry Lesson”. Korean Film Directors: Hong
Sangsoo. Ed. Huh Moonyung. Seoul: Seoul Selection and Korean Film Council, 2007. 19-29. Print.
----. “Do filmmakers deserve the last word?” Observations on film art, 10 October 2007. Web. 11 October
2007. <http://www.davidbordwell.net/blog/2007/10/10/do-filmmakers-deserve-the-last-word/>
----. Figures Traced in Light. Berkeley; London; Los Angeles: California UP, 2005. Print.
----. “Good and Good For You.” Observations on film art, 10 July 2011. Web. 11 July 2011.
<http://www.davidbordwell.net/blog/2011/07/10/good-and-good-for-you/>
----. Narration in the Fiction Film. London: Routledge, 1986. Print.
----. On the History of Film Style. Cambridge; London: Harvard UP, 1997. Print.
----. Ozu and the Poetics of Cinema. London: BFI, 1988. Print.
----. “Pierced by Poetry.” Observations on film art, 13 May 2009. Web. 14 May 2009.
<http://www.davidbordwell.net/blog/2009/05/13/pierced-by-poetry/>
----. The Way Hollywood Tells It: Story and Style in Modern Movies. Berkeley; London; Los Angeles:
California UP, 2006. Print.
Bordwell, David and Lee, Kevin. “Slow Food.” Moving Image Source, 28 April 2011. Web. 30 April 2011.
<http://www.movingimagesource.us/articles/slow-food-20110428>
Bordwell, David; Staiger, Janet and Thompson, Kristin. The Classical Hollywood Cinema: Film Style and
Mode of Production to 1960. London: Routledge and Kegan Paul, 1985. Print.
Burch, Noël. Life to Those Shadows. Berkeley; Los Angeles: California UP, 1990. Print.
----. Theory of Film Practice. London: Secker and Warburg, 1973. Print.
----. To the Distant Observer. Berkeley; Los Angeles: California UP, 1979. Print.
Burnett, Colin. “Parametric Narration and Optical Transition Devices: Hou Hsiao-hsien and Robert Bresson
in Comparison.” Senses of Cinema 33, 2004. Web.
<http://archive.sensesofcinema.com/contents/04/33/hou_hsiao_hsien_bresson.html>
Byg, Barton. Landscapes of Resistance: The German Films of Danièle Huillet and Jean-Marie Straub.
Berkeley; Los Angeles; Oxford: California UP, 1995. Print.
Campany, David. “When to be Fast? When to be Slow?” The Cinematic. Ed. David Campany. London;
Cambridge: Whitechapel and The MIT Press, 2007. 10-17. Print.
----. Photography and Cinema. London: Reaktion Books, 2008. Print.
219
Cardinal, Roger. “Pausing Over Peripheral Detail.” Framework 30/31 (1986): 112-130. Print.
Cavell, Stanley. The World Viewed (Enlarged Edition). Cambridge; London: Harvard UP, 1979. Print.
Chandler, Raymond. Raymond Chandler Speaking. Ed. Dorothy Gardiner and Kathrine Sorley Walker.
Berkeley; Los Angeles: California UP, 1997. Print.
Chang, Dustin. “Cinema Reborn: Apichatpong Weerasethakul Interview.” Twitch, 26 February 2011. Web. 25
August 2012. <http://twitchfilm.com/interviews/2011/02/memories-and-reincarnation-apichapong-
weerasethakul-interview.php>
Cherchi Usai, Paolo; Francis, David; Horwath, Alexander and Loebenstein, Michael, eds. Film Curatorship:
Archives, Museums, and the Digital Marketplace. Vienna: SYNEMA Publikationen
(FilmmuseumSynemaPublikationen), 2008. Print.
Corless, Kieron. “Ah, Liberty! Ben Rivers at the Edge of the World.” Sight and Sound, November 2008. Web.
25 August 2012. <http://old.bfi.org.uk/sightandsound/feature/49491/>
Costa, Pedro and Gorin, Jean-Pierre. “Conversación entre Pedro Costa y Jean-Pierre Gorin: Onde Jaz o Teu
Sorriso?” Trans. Marta Montiano, Jean-Michel Bergère and Francisco Algarín Navarro. Lumière 2
(2009): 46-48. Web.
Coursodon, Jean-Pierre. “Desire Roped In: Notes on the Fetishism of the Long Take in Rope.” Rouge 4,
2004. Web. <http://www.rouge.com.au/4/rope.html>
Cummings, Doug. “Contemplative Cinema and Honor of the Knights.” Film Journey, 8 January 2007. Web.
12 February 2008. <http://filmjourney.weblogger.com/2007/01/08/>
Daney, Serge. “Cinemeteorology.” Trans. Jonathan Rosenbaum, 8 November 1982. Web. 25 August 2012.
<http://www.jonathanrosenbaum.com/?p=21944>
----. L’exercice a été profitable. Paris: POL, 1993. Print.
----. “Falling Out of Love.” Sight and Sound 2.3 (1992): 14-16. Print.
----. “The Screen of Fantasy (Bazin and Animals).” Trans. Mark A. Cohen. Rites of Realism. Ed. Ivone
Margulies. Durham; London: Duke UP, 2003. 32-41. Print.
Davis, William Darrell and Yeh, Emilie Yueh-yu. Taiwan Film Directors: A Treasure Island. New York:
Columbia UP, 2005. Print.
Deleuze, Gilles. Cinema 1: The Movement-Image. Trans. Hugh Tomlinson and Barbara Habberjam. London:
Continuum, 2005. Print.
----. Cinema 2: The Time-Image. Trans. Hugh Tomlinson and Robert Galeta. London: Continuum, 2005.
Print.
de Luca, Tiago Magalhães. “Carnal Spirituality: the Films of Carlos Reygadas.” Senses of Cinema 55, 2010.
Web. <http://sensesofcinema.com/2010/feature-articles/carnal-spirituality-the-films-of-carlos-
reygadas-2/>
----. Realism of the Senses: A Tendency in Contemporary World Cinema. PhD thesis. University of Leeds,
2011. Print.
Diaz, Lav; Chulphongsathorn, Graiwoot; Ingawanij, May Adadol; Lertwiwatwongsa, Wiwat; Tioseco,
220
Alexis. “Lav Diaz in Conversation.” Lumen 1, March 2011. Web. <http://lumenjournal.org/i-
forests/conversation-diaz/>
Dorsky, Nathaniel. Devotional Cinema. Berkeley, California: Tuumba Press, 2005. Print.
Duménil, Gérard and Lévy, Dominique. The Crisis of Neoliberalism. Cambridge, Massachusetts: Harvard
UP, 2011. Print.
Eckmann, Sabine. “Times and Places to Rest.” Sharon Lockhart: 'Lunch Break'. Ed. Sabine Eckmann.
Chicago: Chicago UP, 2010. 23-29. Print.
Edwards, Dan. “Family Life One Moment at a Time: Liu Jiayin on Oxhide.” The Beijinger, 24 February
2010. Web. 30 February 2010. <http://www.thebeijinger.com/blog/2010/02/24/Family-Life-One-
Moment-at-a-Time-Liu-Jiayin-on-Oxhide>
Edwards, Gavin. “Four Minutes With The Girl with the Dragon Tattoo.” The New York Times, 15 December
2011. Web. 29 December 2011. <http://www.nytimes.com/2011/12/18/magazine/david-fincher-girl-
with-the-dragon-tattoo.html>
Ellickson, Lee. “Preparing to Live in the Present: An Interview with Hou Hsiao-hsien.” Cinéaste 27.4
(2002): 13-19. Print.
Elsaesser, Thomas. European Cinema: Face to Face with Hollywood. Amsterdam: Amsetrdam UP, 2005.
Print.
Erice, Victor and Kiarostami, Abbas. Erice—Kiarostami: Correspondences. Ed. Núria Aidelman. Barcelona:
Centre de Cultura Contemporània de Barcelona, Institut d'Edicions de la Diputació de Barcelona and
Actar, 2006. Print.
Falicov, Tamara L. The Cinematic Tango: Contemporary Argentine Film. London: Wallflower Press, 2007.
Print.
Fiant, Antony. “Des films gueles de bois: Notes sur le mutisme dans le cinéma contemporain.” Trafic 50
(2004): 522-539. Print.
Fieschi, Jean-André. “Jean-Marie Straub.” Cinema: A Critical Dictionary. Trans. Michael Graham. Ed.
Richard Roud, New York: Viking, 1980. 969-973. Print.
Fisher, Mark. Capitalist Realism. Winchester; Washington: Zero Books, 2009. Print.
----. “Solidarity not Identification: Straub-Huillet's Radical Cinema.” Film Quarterly 64.1: 46–52. Print.
Flanagan, Matthew. “The Act of Seeing, Synthetically: James Benning's 'Ruhr'.” Mubi Notebook, 25 January
2010. Web. <http://mubi.com/notebook/posts/the-act-of-seeing-synthetically-james-bennings-ruhr-
2009-usa>
----. “Towards an Aesthetic of Slow in Contemporary Cinema.” 16:9 29, 2008. Web. <http://www.16-
9.dk/2008-11/side11_inenglish.htm>
Fox, Dan. “Slow, Fast, and Inbetween.” Frieze (blog), 29 May 2010. Web. 31 May 2010.
<http://blog.frieze.com/slow_fast_and_inbetween/>
Frampton, Hollis. On the Camera Arts and Consecutive Matters: The Writings of Hollis Frampton. Ed.
Bruce Jenkins. Cambridge, Massachusetts: The M.I.T. Press, 2009. Print.
221
Gallagher, Tag. The Adventures of Roberto Rossellini. New York: Da Capo Press, 1998. Print.
Geldzahler, Henry. “Some Notes on 'Sleep'.” Film Culture 32-35 (1964-1965): 13. Print.
Gidal, Peter. Blow Job. London: Afterall, 2008. Print.
----. “Theory and Definition of Structural/Materialist Film.” Structural Film Anthology. Ed. Peter Gidal.
London: BFI, 1978. 1-21. Print.
Godard, Jean-Luc. Godard on Godard. Trans. Tom Milne. Eds. Jean Narboni and Tom Milne. New York: Da
Capo Press, 1986. Print.
Goodman, Ezra. The Fifty-Year Decline and Fall of Hollywood. New York: Simon and Schuster, 1961. Print.
Gozlan, Gérard. “In praise of André Bazin.” The New Wave. Ed. Peter Graham. London: Secker and
Warburg, 1968. 52-71, 114-136. Print.
Groys, Boris. “Comrades of Time.” e-flux 11, December 2009. Web. 3 March 2011. <http://www.e-
flux.com/journal/comrades-of-time/>
Gruyaert, Saskia; Herrera, Aliosha; Lheureux, Christelle; Martin, Raya; Neyrat, Cyril; Renzi, Eugenio and
Thirion, Antoine. “Une conversation avec Apichatpong Weerasethakul et Wallapa Mongkolprasert.”
Independencia, June 2010. Web. 3 July 2010.
<http://www.independencia.fr/SERIES/ITW_JOEI_WALLAPA_SOMMAIRE.html>
Guillen, Michael. “Darkest Americana and Elsewhere: Ruhr—A Few Evening Class Questions For James
Benning.” The Evening Class, 2 March 2010. Web. 3 March 2010.
<http://theeveningclass.blogspot.co.uk/2010/03/darkest-americana-elsewhere-ruhr-few.html>
Halle, Randall. “Offering Tales They Want to Hear: Transnational European Film Funding as NeoOrientalism.”
Global Art Cinema: New Theories and Histories. Eds. Rosalind Galt and Karl
Schoonover. Oxford; New York: Oxford UP, 2010. 303-319. Print.
Hardt, Michael and Negri, Antonio. Empire. Cambridge, Massachusetts; London: Harvard UP, 2000. Print.
Harvey, David. A Brief History of Neoliberalism. Oxford, New York: Oxford UP, 2005. Print.
----. The Condition of Postmodernity: An Enquiry into the Origins of Cultural Change. Oxford: Blackwell,
1989. Print.
Hasumi, Shigehiko. “The eloquence of the taciturn: an essay on Hou Hsiao-hsien.” Inter-Asia Cultural
Studies 9.2 (2008): 184-194. Print.
Hebdige, Dick. “Reeling in Utah: The Travel Log Trilogy.” East of Borneo, 8 January 2011. Web. 16 March
2011. <http://www.eastofborneo.org/articles/29>
Henderson, Brian. A Critique of Film Theory. New York: E.P. Dutton, 1980. Print.
Highmore, Ben. Everyday Life and Cultural Theory. London; New York: Routledge, 2002. Print.
Holte, Michael Ned. “A Few Questions about Place and Time: Sharon Lockhart and Michael Ned Holte in
Conversation.” Newspaper Jan Mot 73 (2010): 1-2. Print.
Hughes, Darren. “Albert Serra Interviewed on El Cant dels ocells.” Senses of Cinema 50, 2009. Web.
<http://sensesofcinema.com/2009/50/albert-serra-interview/>
James, Nick. “Passive Aggressive.” Sight and Sound 2.4 (2010): 5. Print.
222
Jameson, Fredric. The Cultural Turn: Selected Writings on the Postmodern, 1983-1998. London; New York:
Verso, 1998. Print.
----. “The End of Temporality.” Critical Inquiry 29.4 (2003): 695-718. Print.
----. “Reflections in Conclusion.” Aesthetics and Politics. London; New York: Verso, 2007. 196-293. Print.
----. Postmodernism, or the Cultural Logic of Late Capitalism. Durham: Duke UP, 1991. Print.
----. “Theo Angelopoulos: The Past as History, The Future as Form.” The Last Modernist: The Films of Theo
Angelopoulos. Ed. Andrew Horton. Trowbridge, Wiltshire: Flicks Books, 1997. 78-95. Print.
Jeffries, Stuart. “Landscapes of the mind.” The Guardian, 16 April 2005. Web. 25 August 2012.
<http://www.guardian.co.uk/film/2005/apr/16/art>
Jones, Kent. “Cinema with a roof over its head.” Film Comment 35.5 (1999): 46-51. Print.
----. “Here and There: The Films of Tsai Ming-liang.” Movie Mutations: The Changing Face of World
Cinephilia. Eds. Jonathan Rosenbaum and Adrian Martin. London: BFI, 2003. 44-51. Print.
Josipovici, Gabriel. What Ever Happened to Modernism? London; New Haven: Yale UP, 2011. Print.
Kasman, Daniel. “Digital Fugitive: Interview with Pedro Costa.” Mubi Notebook, 16 June 2009. Web. 25
August 2012. <http://mubi.com/notebook/posts/a-digital-fugitive-interview-with-pedro-costa>
Keathley, Christian. Cinephilia and History, or The Wind in the Trees. Bloomington and Indianapolis:
Indiana UP, 2006. Print.
----. “Pass the salt... and other bits of business.” Screen 52.1 (2011): 105-113. Print.
Keiller, Patrick. “Landscape and cinematography.” Cultural Geographies 16 (2009): 409-414, Print.
Klevan, Andrew. Disclosure of the Everyday: Undramatic Achievement in Narrative Film. Trowbridge,
Wiltshire: Flicks Books, 2000. Print.
Koch, Stephen. Stargazer: The Life, World and Films of Andy Warhol. New York; London: Marion Boyars,
1991. Print.
Koehler, Robert. “Agrarian Utopias/Dystopias: The New Non-Fiction.” Cinema Scope 40 (2009): 12-15.
Print.
----. “Cinephilia and Film Festivals.” Dekalog 3: On Film Festivals. Ed. Richard Porton. London:
Wallflower, 2009. 81-97. Print.
----. “El cant dels ocells.” Cinema Scope 35 (2008): 54-56. Print.
----. “The Exiles.” Cinéaste 33.4, 2008. Web. 14 January 2009. <http://www.cineaste.com/articles/the-
exiles.htm>
Kovács, András Bálint. Screening Modernism: European Art Cinema, 1950-1980. London; Chicago:
Chicago UP, 2007. Print.
Kracauer, Siegfried. Theory of Film: The Redemption of Physical Reality. New Jersey: Princeton UP, 1997.
Print.
Latour, Bruno. We Have Never Been Modern. Trans. Catherine Porter. Cambridge, Massachusetts: Harvard
UP, 1993. Print.
Lee, Mark Ping-Bing. Mark Lee Ping-Bing: A Poet of Light and Shadow. Taipei: Garden City Publishers,
223
2009. Print.
Lehman, Peter and Triplett, William. “Money and Art: An Interview with Independent Filmmaker Jon Jost.”
Wide Angle 3.3 (1980): 28-33. Print.
Lertwiwatwongsa, Wiwat. “Agonistes.” Criticine 6, 16 May 2012. Web. 30 May 2012.
<http://criticine.com/review_article.php?id=26>
Lévy, Denis. “Modern Cinema.” Trans. Edwin Mak and Nikolaus Vryzidis. Lumen 1, March 2011. Web.
<http://lumenjournal.org/i-forests/modern-levy/>
Lim, Dennis. “Blurring Reality’s Edge in Fluid China.” The New York Times, 20 January 2008. Web. 12 July
2008. <http://www.nytimes.com/2008/01/20/movies/20lim.html?>
----. “It’s Actual Life. No, It’s Drama. No, It’s Both.” The New York Times, 20 August 2010. Web. 21 August
2010. <http://www.nytimes.com/2010/08/22/movies/22hybrid.html??>
Lim, Song Hwee. “Positioning Auteur Theory in Chinese Cinemas Studies: Intratextuality, Intertextuality
and Paratextuality in the Films of Tsai Ming-liang.” Journal of Chinese Cinemas 1.3 (2007): 223-245.
Print.
Lim, Song Hwee and Hee, Wai Siam, “'You must believe there is an author behind every film': An interview
with Tsai Ming-liang.” Journal of Chinese Cinemas 5. 2 (2011): 181–191. Print.
LiPuma, Edward and Lee, Benjamin. Financial Derivatives and the Globalization of Risk. Durham: Duke
UP, 2004. Print.
Lyotard, Jean-François. “Acinema.” Trans. Paisley N. Livingston and Jean-François Lyotard. Narrative,
Apparatus, Ideology: A Film Theory Reader. Ed. Philip Rosen. New York; Oxford: Columbia UP,
1986. 349-359. Print.
MacCabe, Colin. “Theory and Film: Principles of Realism and Pleasure.” Screen 17.3 (1976): 7-28. Print.
MacDonald, Scott. A Critical Cinema: Interviews with Independent Filmmakers. Berkeley; Los Angeles:
California UP, 1988. Print.
----. A Critical Cinema 2: Interviews with Independent Filmmakers. Berkeley; Los Angeles: California UP,
1992. Print.
----. A Critical Cinema 3: Interviews with Independent Filmmakers. Berkeley; Los Angeles: California UP,
1998. Print.
----. Adventures of Perception: Cinema as Exploration. Berkeley; Los Angeles; London: California UP, 2009.
Print.
----. Avant-garde Film: Motion Studies. Cambridge: Cambridge UP, 1993. Print.
----. The Garden in the Machine. Berkeley; Los Angeles; London: California UP, 2001. Print.
MacKenzie, Donald. “How to Make Money in Microseconds.” London Review of Books 33.10 (2011): 16-
18. Print.
Manovich, Lev. The Language of New Media. Cambridge, Massachusetts; London: MIT Press, 2001. Print.
Marcus, Millicent. Italian Film in the Light of Neorealism. Princeton; Guildford: Princeton UP, 1986. Print.
Margulies, Ivone. Nothing Happens: Chantal Akerman’s Hyperrealist Everyday. Durham, London: Duke UP,
224
1996. Print.
Marrati, Paola. “The Catholicism of Cinema: Gilles Deleuze on Image and Belief.” Religion and Media. Eds.
H. de Vries and S. Weber. Stanford: Stanford UP, 2001. 227-240. Print.
Martin, Adrian. ”Unfinished Diary Las Palmas '09.” Rouge 13, 2009. Web.
<http://www.rouge.com.au/13/diary.html>
----. “What is Modern Cinema?” 16:9 38, 2010. Web. <http://www.16-9.dk/2010-09/side11_inenglish.htm>
----. “The View from Australia.” Dekalog 3: On Film Festivals. Ed. Richard Porton. London: Wallflower,
2009. 98-106. Print.
----. “What's Happening? Story, Scene and Sound in Hou Hsiao-hsien.” Inter-Asia Cultural Studies 9.2
(2008): 258-270. Print.
Matthews, Peter. “Divining the Real.” Sight and Sound 9.8 (1999): 22-25. Print.
Michelson, Annette. “Introduction.” Theory of Film Practice. Noël Burch. London: Secker and Warburg,
1973. v-xv. Print.
Mitchell, Tony. “Animating Dead Space and Dead Time.” Sight and Sound 50.1 (1980): 29-33. Print.
Mulvey, Laura. Death 24x a Second: Stillness and the Moving Image. London: Reaktion Books, 2006. Print.
Navarro, Francisco Algarin. “Aquí y en otro tiempo.” Lumière: Internacional Straub-Huillet, 2011. Web. 25
August 2012. <http://www.elumiere.net/exclusivo_web/internacional_straub/textos/troptot.php>
Neale, Steve. “Art Cinema as Institution.” Screen 22.1 (1981): 11-39. Print.
Needham, Gary. “Ozu and the Colonial Encounter in Hou Hsiao-hsien.” Asian Cinemas: A Reader and
Guide. Eds. Dimitris Eleftheriotis and Gary Needham. Edinburgh: Edinburgh UP, 2006. 369-383.
Print.
The New Schwelgern Coke Plant. Dortmund: ThyssenKrupp, 2012. Print.
Neyrat, Cyril. “Rooms for the Living and the Dead.” Booklet. 10-17. Letters from Fontainhas: Three Films
by Pedro Costa. Dir. Pedro Costa. Criterion Collection, 2010. DVD.
Nichols, Bill. “The Fact of Realism and the Fiction of Objectivity.” Post-war Cinema and Modernity. Eds.
John Orr and Olga Taxidou. Edinburgh: Edinburgh UP, 2000. 188-207. Print.
Peirce, Charles S. The Philosophy of Peirce: Selected Writings. Ed. Justus Buchler. London: Routledge,
2001. Print.
Peranson, Mark. “Pedro Costa: An Introduction.” Cinema Scope 27 (2006): 6-15. Print.
----. “Ruhr.” Cinema Scope 41 (2010): 56-57. Print.
Peranson, Mark, Villamediana, Daniel V. and Arroba, Alvaro. “Battles of Honour and Humanity: Albert
Serra’s Quixotic Experiment.” Cinema Scope 29 (2007): 20-23. Print.
Perez, Gilberto. The Material Ghost; Films and their Medium. Baltimore; London: Johns Hopkins UP, 1998.
Print.
Perkins, Victor. The Magnificent Ambersons. London: BFI, 1999. Print.
Picard, Andréa. “Like a Total Eclipse, or a Supernova: Tatlin, Warhol, Taanila.” Cinema Scope 29 (2007): 66-
68. Print.
225
Quandt, James. “Resistant to Bliss: Describing Apichatpong.” Apichatpong Weerasethakul. Ed. James
Quandt. Vienna: SYNEMA Publikationen (FilmmuseumSynemaPublikationen), 2009. 13-100; 252-
253. Print.
----. “The Sandwich Process: Simon Field Talks About Polemics and Poetry at Film Festivals.” Dekalog 3:
On Film Festivals. Ed. Richard Porton. London: Wallflower, 2009. 53-80. Print.
Quintín. “The Alonso Murder Case.” Cinema Scope 21 (2005): 38-40. Print.
Rancière, Jacques. Film Fables. Trans. Emiliano Battista. Oxford; New York: Berg, 2006. Print.
----. “The Saint and the Heiress: A propos of Godard's Histoire(s) du cinéma.” Discourse 24.1 (2002): 113-
119. Print.
Rapfogel, Jared. “Still Lives in Times of Change: An Interview with Jia Zhang-ke.” Cinéaste 33.2 (2008):
44-47. Print.
Reichert, Jeff and Syngle, Erik. “Ghost Writer.” Trans. Shujen Wang. Reverse Shot 10, 2004. Web. 25
August 2012. <http://www.reverseshot.com/legacy/winter04/tsai.html>
Reynaud, Bérénice. A City of Sadness. London: BFI, 2002. Print.
Rist, Peter. “Interview with Liu Jiayin.” Offscreen 13.12, 31 December 2009. Web. 16 January 2010.
<http://www.offscreen.com/index.php/pages/essays/interview_liu_jiayin/>
Rivette, Jacques. “Letter on Rossellini.” Trans. Tom Milne. Cahiers du Cinéma: The 1950s. Ed. Jim Hillier.
Cambridge; Massachusetts: Harvard UP, 1985. 192-204. Print.
Rodowick, D.N. The Virtual Life of Film. Cambridge, Massachusetts: Harvard UP, 2007. Print.
Rohmer, Eric. “André Bazin’s 'summa'.” Trans. Carol Volk. A Taste for Beauty. Ed. Jean Narboni.
Cambridge: Cambridge UP, 1989. Print.
Römers, Holger. “Creating His Own Cinematic Language: An Interview with Apichatpong Weerasethakul.”
Cinéaste 30.4 (2005): 42-47. Print.
Romney, Jonathan. “In Search of Lost Time.” Sight and Sound 20.2 (2010): 43-44. Print.
Rosen, Miriam. “In Her Own Time.” Artforum 42.8 (2004): 122-127. Print.
Rosen, Philip. Change Mummified; Cinema, Historicity, Theory. Minneapolis; London: Minnesota UP, 2001.
Print.
----. “Screen and 1970s Film Theory.” Inventing Film Studies. Ed. Lee Grieveson and Haidee Wasson.
Durham and London: Duke UP, 2008. 264-297. Print.
Rosenbaum, Jonathan. “A Place in the Pantheon.” Chicago Reader, 9 May 1996. Web. 25 August 2012.
<http://www.jonathanrosenbaum.com/?p=6749>
----. Essential Cinema: On the Necessity of Film Canons. Baltimore: Johns Hopkins UP,
2004. Print.
----. Film: The Front Line 1983. Denver: Arden Press, 1983. Print.
----. “Historical Meditations in Two Films by John Gianvito.” Film Quarterly 62.2 (2008): 26-32. Print.
----. “Is Ozu Slow?” Senses of Cinema 4, 2000. Web. <http://www.sensesofcinema.com/2000/feature-
articles/ozu-2/>
226
Rosenbaum, Jonathan and Saeed-Vafa, Mehrnaz. Abbas Kiarostami. Urbana; Chicago: Illinois UP, 2003.
Print.
Roud, Richard. Straub. New York: Viking, 1972. Print.
Ruiter, Bram. “Audio Interview: Lav Diaz and Hazel Orencio.” Salon Indien, 14 February 2012. Web. 16
February 2012. <http://www.salonindien.nl/2012/audio-interview-lav-diaz-hazel-orencio/>
Ruiz, Raul. Poetics of Cinema. Paris: Dis Voir, 2005. Print.
Salmon, Felix. “Chart of the day, HFT edition.” Reuters (blogs), 6 August 2012. Web. 6 August 2012.
<http://blogs.reuters.com/felix-salmon/2012/08/06/chart-of-the-day-hft-edition/>
Salmon, Felix and Stokes, Jon. “Algorithms Take Control of Wall Street.” Wired, 27 December 2010. Web. 6
August 2012. <http://www.wired.com/magazine/2010/12/ff_ai_flashtrading/all/1>
Schlosser, Eric. “Interview with Béla Tarr.” Bright Lights Film Journal 30, 2000. Web. 25 August 2012.
<http://www.brightlightsfilm.com/30/belatarr1.php>
Schoonover, Karl. “Wastrels of Time: Slow Cinema’s Laboring Body, the Political Spectator, and the Queer.”
Framework 53.1 (2012): 65-78. Print.
Schrader, Paul. Transcendental Style in Film: Ozu, Bresson, Dreyer. Berkeley: California UP, 1972. Print.
Shaviro, Steven. The Cinematic Body. Minneapolis: Minnesota UP, 1993. Print.
----. “Emotion Capture: Affect in Digital Film.” Projections 1.2 (2007): 63-82. Print.
----. “Gamer.” The Pinocchio Theory, 15 December 2009. Web. 16 December 2009.
<http://www.shaviro.com/Blog/?p=830>
----. “Post-Cinematic Affect: On Grace Jones, Boarding Gate and Southland Tales.” Film-Philosophy 14.1
(2010): 1-102. Web.
----. “Post-Continuity: full text of my talk.” The Pinocchio Theory, 26 March 2012. Web. 26 March 2012.
<http://www.shaviro.com/Blog/?p=1034>
----. “Slow Cinema Vs Fast Films.” The Pinocchio Theory, 12 May 2010. Web. 13 May 2010.
<http://www.shaviro.com/Blog/?p=891>
Shklovsky, Victor. “Art as Technique.” Russian Formalist Criticism: Four Essays. Trans. Lee T. Lemon and
Marion J. Reis. Lincoln, Nebraska; London: Nebraska UP, 1965. 3-24. Print.
Sicinski, Michael. “James Benning, one of America's finest filmmakers, trains his attention on the railroad in
RR.” The Nashville Scene, 8 April 2010. Web. 9 May 2010.
<http://www.nashvillescene.com/nashville/james-benning-one-of-americas-finest-filmmakers-trains-
his-attention-on-the-railroad-in-rr/Content?oid=1492701>
----. “2010 Toronto International Film Festival: Uncle Boonmee Who Can Recall His Past Lives.” The
Academic Hack, 14 September 2010. Web. 25 August 2012.
<http://academichack.net/TIFF2010.htm#Apichatpong>
Singer, Ben. “Jeanne Dielman: Cinematic Interrogation and 'Amplification'.” Millennium Film Journal 22
(1989): 56-75. Print.
Sitney, P. Adams. “Landscape in the Cinema: the Rhythms of the World and the Camera.” Landscape,
227
Natural Beauty and the Arts. Ed. Salim Kemal. Cambridge: Cambridge UP, 1993. 103-126. Print.
----. Visionary Film: The American Avant-Garde 1943-2000 (Third Edition). Oxford: Oxford UP, 2002. Print.
Slanar, Claudia. “Landscape, History and Romantic Allusions.” James Benning. Eds. Barbara Pichler and
Claudia Slanar. Vienna: SYNEMA Publikationen (FilmmuseumSynemaPublikationen), 2007. 169-180.
Print.
Sontag, Susan. On Photography. London: Penguin Classics, 2002. Print.
----. Against Interpretation and Other Essays. London: Penguin Classics, 2009. Print.
Tarkovsky, Andrei. Sculpting in Time. Austin: Texas UP, 1987. Print.
Teo, Stephen. “Cinema with an Accent—Interview with Jia Zhangke, Director of Platform.” Senses of
Cinema 15, 2001. Web. <http://sensesofcinema.com/2001/feature-articles/zhangke_interview/>
Tioseco, Alexis. “A Conversation with Lav Diaz.” Criticine 2, 30 January 2006. Web. 25 August 2012.
<http://criticine.com/interview_article.php?id=21>
----. “Shifting Agendas: The Decay of the Mainstream and Rise of the Independents in the Context of
Philippine Cinema.” Inter-Asia Cultural Studies 8.2 (2007): 298-303. Print.
Thompson, Kristin. Breaking the Glass Armor: Neoformalist Film Analysis. Princeton; Guildford: Princeton
UP, 1988. Print.
----. “The Concept of Cinematic Excess.” Narrative, Apparatus, Ideology: A Film Theory Reader. Ed. Philip
Rosen. New York: Columbia UP, 1986. 130-142. Print.
Vaughan, Dai. For Documentary. Los Angeles; Berkeley; London: California UP, 1999. Print.
Walsh, Martin. “Brecht and Straub/Huillet: The Frontiers of Language.” Afterimage 7 (1978): 12-31. Print.
Wee, Brandon. “The Decade of Living Dangerously: A Chronicle from Lav Diaz.” Senses of Cinema 34,
2005. Web. <http://sensesofcinema.com/2005/34/lav_diaz/>
Willemen, Paul and King, Noel. “Through the Glass Darkly: Cinephilia Reconsidered.” Looks and Frictions:
Essays in Cultural Studies and Film Theory. London: BFI, 1994. 223-257. Print.
Williams, Evan Calder. “To Sit on a Throne of Teeth, Graced With a Crown of Teeth.” The New Inquiry, 29
September 2011. Web. 29 September 2011. <http://thenewinquiry.com/essays/to-sit-on-a-throne-of-
teeth-graced-with-a-crown-of-teeth/>
Wollen, Peter. Signs and Meaning in the Cinema. London: Secker and Warburg; BFI, 1969. Print.
Wong, Silvia. “Evolution of an Epic.” Screen International 1525 (2005): 11. Print.
Yardley, Jim. “Chinese Dam Projects Criticized for Their Human Costs.” The New York Times, 19 November
2007. Web. 4 April 2010. <http://www.nytimes.com/2007/11/19/world/asia/19dam.html?>
Yue, Genevieve. “The Light Outside the Window.” Reverse Shot 23, 2008. Web. 25 August 2012.
<http://reverseshot.com/article/café_lumiere>
Zavattini, Cesare. “A Thesis on Neo-Realism.” Springtime in Italy: A Reader on Neorealism. Ed. and Trans.
David Overbey. London: Talisman, 1978. 67-78. Print.
----. “Some Ideas on the Cinema.” Film: A Montage of Theories. Ed. Richard Dyer MacCann. New York:
E.P. Dutton, 1966. 216-228. Print.
228