The increased intensity with which the aesthetic choices outlined in the first chapter were adopted by filmmakers internationally around the turn of the century sparked a critical response regarding their status as markers of an autonomous movement. Could these choices be thought of as comprising a consistent, original, cinematic style? A multitude of critical baptisms ensued, inaugurated, most likely, by Michel Ciment’s ‘cinema of slowness’, a term first used in his address to the 46th San Francisco International Film Festival in 2003.1 Other denominations ranged from Beginning of page[p. 100] revisitations of older concepts, such as Mark Betz’s reworking of David Bordwell’s ‘parametric narration’,2 or Robert Koehler’s use of André Bazin’s ‘cinema of duration’,3 to adjectives that assigned primary constitutional significance to one or another trait or effect — physical,4 sensory,5 contemplative,6 and, of course, slow.
In the February 2010 issue of Sight & Sound, Jonathan Romney reverted to the latter term, with a capital ‘S’, to describe a ‘varied strain of austere minimalist cinema that has thrived internationally over the past ten years’.7 Shortly thereafter, his colleague Nick James adopted the same term in his unsympathetic editorials in the same publication.8 For James, Slow films are ‘passive-aggressive in that they demand great swathes of our precious time to achieve quite fleeting and slender aesthetic and political effects’.9 The Sight & Sound debate stirred a heated online discussion amongst critics and cinephile bloggers,10 which Beginning of page[p. 101] resulted in a follow-up editorial by James in July of the same year. In ‘Being Boring’, James, now on the defensive, asks: ‘Why pretend that contemplative cinema is immune to the usual pressures that success and ubiquity bring to art movements?’11 Despite his acerbic criticism, James thus became the first to officiate the demarcated territory of Slow cinema as that of a veritable art movement.
Intended at least partially as a critique, the term Slow cinema, in a kind of Fauvist twist, was the one to eventually become common currency in reference to the style, gradually making its way into criticism and beyond, despite its inherently knotty nature. In this sense, the most immediately apparent problem with the notion at the centre of the style was that slowness is a subjective value. Slow compared to what? In the installation Three Rooms (2008) by Jonas Dahlberg, the everyday objects furnishing three different rooms, filmed by fixed cameras and projected on three forty-six-inch screens, are shown slowly dissolving over twenty-six minutes of video. Here, time can be said to unfold both hopelessly slowly and infinitely fast: the movement of the annihilation is at a rate much slower than that of average perception, but it is also extremely accelerated from the point of view of geological time. Similarly, in Carlos Reygadas’s Silent Light (2007), a time-lapse shot of a sunrise can be said to be simultaneously accelerated and excruciatingly Beginning of page[p. 102] slow. As many have pointed out, the term is also negatively marked within the context of a financial capitalism that tends to assign positive value to speed and efficiency (we will come back to this). In this sense, it is noteworthy that one of the six subentries under the adjective slow in the 2012 edition of the Oxford American Writer’s Thesaurus is for a ‘slow movie’; the synonyms proposed are particularly revealing: ‘dull, boring, uninteresting, uneventful, tedious, tiresome, wearisome, monotonous, dreary, lackluster’. The suggested antonym is not fast, but exciting. A search for fast reveals no such negative bias (but it does serve to remind us that the Old English origin of the word connotes something ‘firmly fixed’). At the same time, the term slow raises questions of spectatorship (when and why do we perceive something as being slow?), invoking the opposition with an implied counterpart (slow/fast), and, for some, an equally presumed antagonist, the US film industry.
In a sense, Slow cinema can be said not to be slow at all: if it is true that its directors exhibit a preference for the long take, it is equally true that many frequently use techniques such as handheld cameras, jump cuts, and narrative ellipses — techniques, that is, often associated with a certain contemporaneous strand of Hollywood film productions and TV series. This type of mainstream production, defined as ‘post-continuity’ by Steven Shaviro, gradually became prominent around the turn of the century at approximately the same time as Slow cinema.12 Inversely, US productions may likewise employ time-stretching devices, such as dramatic slow motion, bullet-time photography (also known as dead time), or real-time models (24 [2001–10]).Beginning of page[p. 103]
This chapter considers the judgements we make when we talk about the flow of temporality in film. The peculiarities of Slow cinema proceed, to a large extent, from its distended durations. By positing duration as the constant upon which the various effects of Slow films are predicated, I will turn to Gilles Deleuze’s Cinema books to suggest that they help us see these effects enacted.
In his two volumes on cinema, Deleuze lays out his complex taxonomy of cinematic signs, structured around two forces: movement and time. The resultant movement- and time-images describe two different image orders, roughly corresponding to classical, or pre-war, and modern, or post-war cinema.13 Deleuze published the original French versions of Cinema 1: The Movement-Image and Cinema 2: The Time-Image in the mid-eighties. In 1997, D. N. Rodowick noted with stupefaction that ‘although both books were quickly translated into English, even today, over ten years later, Deleuze’s study of film has had comparatively little impact on contemporary anglophone film theory’.14 More than twenty-five years later, by contrast, one might dare say it has had too much.
Rodowick had already detected many of the problems and limitations that the use of Deleuze’s work on cinema would pose in the future. For instance, his eclectic choice of directors and his underwhelming use of close film analysis would leave future deployment of his theory open to the Beginning of page[p. 104] charge of cultural elitism and would often result in simplistic applications of his theory onto individual films.15 Much criticism also centred on what appeared to be a counterintuitive, if not directly self-contradictory use of history on Deleuze’s part.16 Critics noted that, in the Cinema books, Deleuze appears to describe a narrative of the transition from the movement- to the time-image via a crisis of the former, precipitated by the traumatic experience of the Second World War. This crisis, according to Deleuze, would cause the neat narratives and sense-making mechanisms of the movement-image to collapse in favour of a new regime of the image centred on the failure of meaningful sensory-motor action, and on the consequent foregrounding of time in itself. This account was criticized by some for leading to Eurocentric conclusions, by suggesting, for example, that the new time-image was born in the ‘ruins of Europe’, or by universalizing insights based on the analysis of mostly European and US films.17 It also presented an inexplicably teleological and progressivist historical narrative of the passage from classical to modern cinema that appeared to go decidedly against Deleuze and Guattari’s criticism of the concept of history in A Thousand Plateaus as a product of the state apparatus, against which they had posed the open-ended notion of ‘nomadology’.18
Despite these impasses, Deleuze’s work continues to be a source of highly relevant insights into the cinematic Beginning of page[p. 105] medium.19 In part, his work’s continuing relevance has been the outcome of efforts to address the lacuna between the two Cinema volumes, and to rethink Deleuze’s books on cinema in relation to his other work. As a result, new models have been set up, with varying degrees of success, by appealing to some of the main concepts in the two volumes of Capitalism and Schizophrenia, such as the kindred concepts of ‘schizoanalysis’ and nomadology.20
Rodowick himself suggests a way around this lacuna by positioning Deleuze’s intentions not as historical, but as philosophical. In Reading the Figural, he explains that Deleuze’s movement- and time-images are not chronologically consecutive, but are proposed as two separate philosophies of history. The movement-image, in this formulation, is Hegelian, whereas the time-image is Nietzschean. The former is characterized by an ‘organic representation marked qualitatively by a will to truth’, whereas the latter is ‘organized by “fabulation”, a falsifying narration defined not by representation but by simulacra whose qualities are “powers of the false”’.21 But whilst this reconfiguration Beginning of page[p. 106] rids the Cinema books of the problem of chronology, it still retains the sense of complete separateness, as well as the primacy of the time- over the movement-image, inasmuch as Deleuze is an anti-Hegelian par excellence, therefore maintaining a sense of progression or evolution from movement to time.
Rancière has similarly concerned himself with Deleuze’s classification of cinematic signs, but his interpretation avoids the pitfall of an implied hierarchy between the two regimes. In his Film Fables, Rancière recasts Deleuze’s cinema project by calling for an abolition of the consecutive aspect of his dual typology. He declares that ‘the rupture that structures the opposition between the movement-image and the time-image is a fictive rupture, so that we may do better to describe their relationship not as an opposition, but as an infinite spiral’.22 Rancière uses one of Deleuze’s most cited auteurs to problematize the latter’s diachronic construction of the moving image. Deleuze understands Robert Bresson’s use of the ‘fragmentation of bodies and shots’ as ‘the infinitization of the interval that disorients the spaces and separates the images’, which functions as an ideal demonstration of the weak sensory-motor links, irrational cuts, and any-space-whatevers that make up the pure optical and time-image. ‘But we could also see the fragmentation as doing the inverse,’ Rancière counters, ‘as intensifying the coordination between the visual and the dramatic: we seize with our hands, no need then to represent the whole body.’23 In this latter case, Bresson’s fragmented montage amounts to a brilliant economy of the action-image.Beginning of page[p. 107]
For all that Rancière’s argument does to debunk Deleuze’s fixed border between classical and modern cinema, his enterprise does not imply a rejection of the Deleuzian model, but the inverse. If Rancière claims the ‘near-total indiscernibility between the logic of the movement-image and the logic of the time-image’, he equally detects in Deleuze’s theory the capacity to ‘bring forth the dialectic constitutive of the cinema’.24 By this token, the Cinema books do not constitute a diachronic history of the image, but a synchronic one, ‘an infinite spiral’ indeed. Whilst Rancière’s approach to the movement-image and time-image is in part intended as a critique of Deleuze’s, it simultaneously salvages its usefulness by recasting it as a typology rather than a chronology. It is within this framework that the rest of this chapter looks at Slow cinema in relation to Deleuze’s work.
In many ways, Deleuze’s theorization of modern cinema is an obvious fit for this category of contemporary art films: Slow cinema continues to showcase and develop some of the most prominent aspects of the time-image to a remarkable extent. On the one hand, this comes down to the matching of Deleuze’s descriptions with some of Slow cinema’s staple techniques: for instance, the preference for the sequence shot and internal montage over conventional editing is, for Deleuze, one of the conditions of the direct time-image.25 On the other hand, with the release of Tarr’s Kárhozat (the film I have posited as one of the very first Slow films) in 1987, bookended by the publication of Deleuze’s books (1983 and 1985) and their translation into English (1986 and 1989), this is, in some sense, the Beginning of page[p. 108] first post-Cinema generation of filmmakers. Without wishing to make the unviable assumption that Slow directors are all necessarily familiar with Deleuze, it is nevertheless true that this is the first generation of art films born in a post-Cinema environment of filmmakers — and, more importantly, of critics and academics — increasingly aware of Deleuze’s vocabulary of cinematic images. Some do explicitly talk about Deleuze’s influence: Frammartino, for example, who also teaches film direction, often discusses his work in Deleuzian terms.26 I am not alone in paying attention to the historical conjuncture between the affinities of Slow cinema and Deleuze’s theory. Paul Schrader’s 2018 introduction to the republished edition of his influential monograph from the early 1970s, Transcendental Style in Film, explains the ‘hydra-headed creature we call slow cinema’ in similar terms. What happened, he asks, to transcendental style, the mode of filmmaking whose main exponents he identified in Ozu, Bresson, and Dreyer, that morphed it into Slow cinema? ‘What happened? Gilles Deleuze happened.’27
To begin with, therefore, I will trace some of the ways in which Slow cinema and Deleuze’s cinematic concepts correspond with each other and work together. Gradually, however, it will become more and more apparent that this correspondence is neither absolute nor adequate. If, for instance, Deleuze posited that modern cinema was ‘completely political’ in its day,28 the mutations and intensifications Beginning of page[p. 109] of capitalism during the last decades of the twentieth century and the first decades of the twenty-first require different modes of being political in cinema. Understanding in what ways Slow cinema can be thought of as responding to the contemporary political moment is therefore an important question. Before turning to these issues, I will begin by sketching how the cinematic apparatus can be seen as corresponding to the mechanisms of memory as expounded by Bergson, in order to suggest how, combined with long duration, it can effect an expanded spectatorial experience.
In the chapter devoted to the ‘crystal-image’ in Cinema 2, Deleuze defines the process of crystallization of the image as such: ‘the real object is reflected in a mirror-image as in the virtual object which, from its side and simultaneously, envelops or reflects the real: there is a “coalescence” between the two.’29 The most self-evident mode of a crystal-image is the mirror. In the case of the mirror, the indiscernibility of actual and virtual is an inherent characteristic of an image that is by its nature double: ‘the mirror-image is virtual in relation to the actual character that the mirror catches, but it is actual in the mirror which now leaves the character with only a virtuality and pushes him back out-of-field.’30 For Deleuze, the ideal example of this type of crystal is, unsurprisingly, the climactic hall of mirrors sequence in Orson Welles’s The Lady from Shanghai (1947).
Needless to say, if we were to stop at this very specific, literal extension of the crystal it would hardly be malleable enough to hold any importance as a figure for film analysis.Beginning of page[p. 110] The mirror is Deleuze’s starting point for a Bergsonian exploration of the actual–virtual circuit and its two correlative dualities, limpid–opaque and seed–environment. What ensues in the second part of his chapter is an extensive conceptual analysis of the ontology of the crystal, which corresponds to the indivisibility and indiscernibility of actual and virtual, present and past.
Deleuze identifies four states of the crystal: the perfect crystal (in Max Ophüls), the imperfect crystal (in Jean Renoir), the crystal caught in its formation (in Federico Fellini), and the crystal in the process of its decomposition (in Luchino Visconti). All of his examples, disappointingly, move away from the image as such and eventually plunge into considerations of narrative and symbolism. One of Rancière’s main points of critique of Deleuze’s Cinema books is, in fact, that he falls into the trap of discussing plot and character development instead of the cinematic image itself. According to Rancière, Deleuze commits the same mistake, for example, in his discussion of Hitchcock as the privileged example of the crisis of the movement-image: ‘the “paralysis” of each of these characters’, Rancière says, ‘is actually only an aspect of the plot. […] Hitchcock’s camera is not paralyzed by Scottie’s vertigo.’31 The problem most poignantly arises in Deleuze’s discussion of the ‘states of the crystal’. The perfect crystal’s round impermeability is instanced in Lola Montès (Lola Montez) (1955), when ‘the pitiless M. Loyal […] keeps on thrusting the failing heroine back on the stage’;32 Renoir’s crystal manifests its broken circuit in La Règle du jeu (The Rules of the Game) (1939) in the gamekeeper, who, being ‘the only person who does not have a double or reflection, […] is the one Beginning of page[p. 111] who breaks the circuit, who shatters the cracked crystal with rifle shots’;33 in Visconti the decomposing crystal manifests symbolically in ‘the rotting of Ludwig II’s teeth, […] the debasement of Ludwig II’s love-affairs; and incest everywhere, […] everywhere the thirst for murder and suicide’.34
Instead, I suggest it might be fruitful to attempt to transpose the crystal-image from the filmic to a limit position between the filmic and pro-filmic. I see a tendency of the crystal-image to assume this placement in Deleuze’s theoretical discussion (‘the crystal always lives at the limit, it is itself “the vanishing limit between the immediate past which is already no longer and the immediate future which is not yet”’35), a tendency which is swiftly halted in his examples of the crystals of time. If we were, however, to act upon this tendency of the crystal for liminal positioning, we would find that situating it between the pro-filmic event and the filmed event most adequately enacts that tendency. Most straightforwardly, this operation would be parallel to the Deleuzian mirror-image: the filmed image of an object is virtual in relation to the filmed object itself, but when projected it is an actual film image, as opposed to the absent, thus virtual, object. So, if I film a vase, the filmed representation of the vase is virtual in relation to the object itself, but when I project the filmed image for a spectator, they see the actual image of a virtual vase that is absent.
In Bergson’s famous cone schema, point S stands for the actual present, and the various AB, A’B’ sections are the virtual circuits that contain all of our past to a greater or lesser degree of contraction, depending on how far from Beginning of page[p. 112] point S we move. Thus, point S is, in fact, not a point but the smallest circuit, where the contraction of the past is absolute and crystallizes with actual experience in the present moment. If we replace for the present at point S the moment of projection, then the ‘smallest circuit’ where the actual crystallizes with its own virtual image can be represented by the trinity: projection–actual projected two-dimensional image–virtual pro-filmic past moment of shooting. In other words, if the present S is the moment of projection, then this moment is actualized in the filmed image, in which, however, is simultaneously inscribed the moment of registration, or the ‘memory’ of the pro-filmic event. The projection, it must be noted, is not an instance of a recollection-image, which Deleuze does not recognize as a pure crystal state. It is not a recollection-image because the projection is not the actualization of a former past in a new present, a different present from the one in which it was actualized. Instead, it is its own virtual image, the image of pro-filmic-event-plus-camera, which is by nature a virtual image, as it is actualized nowhere, except in the moment of its projection. Thus, the moment of projection becomes the actual image of its own virtual image and crystallizes with it. This is a perfect circuit, because the actual image of the present projection is also the virtual, repeatable image of a pro-filmic event that once belonged to the real. Here we see that the extension of the crystalline argument is that cinema, as a medium, is itself a crystal. Like a mirror, the cinema has an actual and a virtual pole. More than that, it presents multiple layers or folds of time, which crystallize with the simultaneous memory of the pro-filmic actual. The screen, as Deleuze puts it, is itself ‘the cerebral membrane where immediate and direct confrontations take place between that past and the future, the inside Beginning of page[p. 113] and the outside, at a distance impossible to determine’.36 The cinema itself thus performs like memory.37 Thinking about cinema-as-crystal reveals how well the concept of the crystal can interact with the cinematic medium as a whole. As crystal, cinema becomes a machine for the production, not of the present or the past, but of temporal potentiality or virtuality. What we see in the crystal, as Deleuze says, is time in its pure state, constantly splitting into actual and virtual. This attribute of the cinema therefore finds its perfect correlate in the presentation of long durations without dramatic action, where time draws attention to itself and to its passing.
In fact, whilst the time-image is most often associated with Neorealism, some of Deleuze’s most significant and elegant passages on the matter are devoted to Yasujirō Ozu’s cinema of the quotidian. It is in relation to the long, fixed shot of the vase in Late Spring, for instance, that we find its most oft-quoted poetic definition (‘This is time, time, “a little time in its pure state”’).38 Here, also, Deleuze praises the merits of duration in Ozu’s portrayal of everyday banality. For Deleuze, an emancipation of the senses happens through this type of image:Beginning of page[p. 114]
In everyday banality, the action-image and even the movement-image tend to disappear in favour of pure optical situations, but these reveal connections of a new type, which are no longer sensory-motor and which bring the emancipated senses in direct relation with time and thought.39
In duration, and particularly when faced with stretches of temps morts that do not advance the narrative, the spectator is called upon to make an effort of attention. For Bergson, ‘the progress of attention results in creating anew not only the object perceived, but also the ever-widening systems with which it may be bound up.’40 The effort of attention or ‘degree of tension which our mind adopts’ will thus expand the circuits of memory into larger, deeper strata.41 These new connections, made possible in and through time, are prompted by and enabled through extended durations that exhaust narrative and denotative meaning. The long takes in Slow cinema are most often autonomous, that is, they do not look forward to or follow from previous ones, but are events in themselves or, as Steven Marchant says in relation to Tarr’s work, the event in the shot coincides with the event of the shot.42 In such cases, attention cannot look for connections outside of the shot, but has to distend inside it. This provides another way of looking at the loosening of motor connections — not just within the film, but in the construction of sequences. In Slow films, we see accumulations of qualities that are absorbed by the image itself, rather than being extended either into narrative or editing Beginning of page[p. 115] action. These qualities Deleuze links to the concepts of the face and the affection-image, as we will now see.
Already in A Thousand Plateaus, Deleuze, with Guattari, speaks of the close-up of the human face in relation to landscape:
The close-up in film treats the face primarily as a landscape; that is the definition of film, black hole and white wall, screen and camera. […] Close-ups animate and invent all of their correlations. So, is your mother a landscape or a face? A face or a factory? (Godard.) All faces envelop an unknown, unexplored landscape; all landscapes are populated by a loved or dreamed-of face, develop a face to come or already past. What face has not called upon the landscapes it amalgamated, sea and hill; what landscape has not evoked the face that would have completed it, providing an unexpected complement for its lines and traits?43
In the circuit that is formed between face and landscape, the face mirrors and contracts the attributes of the environment, which in its turn duplicates this process — the one acting as a virtual mirror-image of the other, ‘the little crystalline seed and the vast crystallizable universe’.44 The landscape is the virtual, unexplored, and opaque and enveloped in the actual image of the face that will ‘amalgamate’ it, whilst simultaneously the landscape becomes the actual image or the milieu of which the past or future face (‘a face to come or already past’) is the virtual mirror.
As an example, we can see this amalgamation of relations in the last cinematic image produced by Portuguese Beginning of page[p. 116] director João César Monteiro before his death. This is a ten-minute-long shot of an extreme close-up of his eye (Monteiro was also the protagonist in most of his films) in the closing scene of his final film, Vai e Vem (Come and Go) (2003). The gigantic eye is preceded by an image of the century-old cypress that dominates the Principe Real square in Lisbon, whose canopy reaches twenty-three metres in diameter. The image of the tree dissolves into that of the eye and the two coexist for a few moments. This is an unusual choice for Monteiro, who otherwise almost exclusively uses the simple cut. The transition from tree to eye thus wants us to notice it, to see the connection that it describes, because Monteiro’s idiosyncratic cinematic universe is embedded and summarized in these two consecutive/overlapping images. The single most important principle of organization and the guiding force of the Portuguese director’s work is his obsession with the observation and pursuit of young (often underage) women, a pursuit that unfailingly ends in disaster, as in Recordações da Casa Amarela (Recollections of the Yellow House) (1989), A Comédia de Deus (God’s Comedy) (1995), As Bodas de Deus (The Wedding of God) (1999), and Vai e Vem.
In a slightly earlier shot, the tree is inhabited by a mysterious female presence, Daphne. The film subsequently cuts to a long shot of the protagonist, Monteiro himself, sitting on a bench just underneath the tree. When we return to the cypress in the next shot, we find it is now empty: a close-up of the empty tree, a ‘landscapified’ shot. The ‘lack’ of the second tree is inscribed as a consequence of the woman’s disappearance. The connection that is then made between the empty tree and the eye mirrors the inscription and transfers it to the latter. The object of desire, enveloped in the shadows of the enormous tree, entangled in its intricate, serpentine branches, is extracted, transforming the Beginning of page[p. 117] latter into pure libidinal force. This objectless, pure passion is then duplicated in the overwhelming eye of João, ‘the black hole of subjectivity as consciousness or passion, the camera, the third eye’ (Fig. 2.1).45

The tree and the eye, landscape and face, the ‘deterritorialized world’ and ‘deterritorialized head’. The production of faces and landscapes, according to Deleuze and Guattari, is a consequence of deterritorialization and reterritorialization. The face is produced when the head has been decoded and is reterritorialized as face; whereas the ‘collapse of corporeal coordinates or milieus implies the Beginning of page[p. 118] constitution of a landscape’.46 This violent dislocation effectuated in the extreme close-up produces a double image: the internal point of origin (the disappeared girl) and the envelope that encompasses all (the desire), the seed and universe of Monteiro’s world in one image with two distinct but indiscernible facets.
The facialization of João’s eye, however, brings with it the horrors of the face. More than just the passional consciousness of the black hole of subjectivity, João’s eye institutes a system of power, becoming the despotic, eroticized male gaze. It is not the face of a white man, but, as Deleuze and Guattari say, ‘White Man himself, […] the average sensual man, in short, the everyday Erotomaniac’.47 In A Thousand Plateaus, ‘facialization’, the process of the production of a face, is defined as ‘the semiotic of capitalism’, a fixing of identity.48 The abstract machine of faciality has two functions, according to Deleuze and Guattari: computation of normalities, and deviance detection. The first functions to produce and transform ‘concrete individualized faces’ on the basis of ‘an x or y’ distinction (a man or a woman). The second performs a judgement on the concrete face, on the basis of an equally binary ‘yes–no’ distinction. The machine ‘rejects faces that do not conform, or seem suspicious’ (‘A ha! It’s not a man and it’s not a woman, so it must be a transvestite […]. At any rate, you’ve been recognized’).49 As gridded, normalized face, João is the heteronormative, hegemonic male gaze; as perverted old man who lusts after underage girls, João is a deviant. In either case, he has been facialized, recognized by the machine.Beginning of page[p. 119]
The machine of faciality, triggered by an apparatus of power, does not recognize an outside of itself, and that is why, for Deleuze and Guattari, it is imperative that we try to escape the face, to dismantle it, in order to become clandestine.50 The first step in this project of breaking away from subjectivity, however, is to know the face.51 Just as it is the locus of power, the face is itself also the promise of a beyond, the transformation of identity into difference. The close-up in film is a form of facialization, a mask that assigns identity to a character,52 such as that of the femme fatale (Garbo) or the saint (Falconetti). On the other hand, what better way to know the face than by its deterritorialization and expansion into the close-up? In this sense, the close-up can come close to that other tendency of the face, to create ‘lines of flight’ away from facialization and towards a rejection of subjectivity. This other tendency is closely linked with what Deleuze, in Cinema 1, calls the affection-image.
The face as affection-image, as described in the Cinema books, is not the same as the face of A Thousand Plateaus, the social production of face. As Amy Herzog points out, the two concepts have in common a decontextualized, not-necessarily-human character. However, they also represent two different tendencies of the face, one towards subjectivity and the other towards indeterminacy.53Beginning of page[p. 120]
As with the face in A Thousand Plateaus, Deleuze’s affection-image is closely associated with the close-up, to such an extent that the two appear interchangeable: ‘the affection-image is the close-up and the close-up is the face’.54 The affection-image is a power or quality considered for itself, as expressed,55 situated between the perception-image and the action-image. Perception and action, according to Deleuze, are inseparable, the former relating to objects of perception and the latter to acts of a subject, but the passage from one to the other is imperceptible. It is in that imperceptible interval that the affection-image is situated. Deleuze says that ‘there is inevitably a part of external movements that we “absorb”, that we refract, and which does not transform itself into either objects of perception or acts of the subject’. That is the habitat of the affection-image. When some part of action has become, for whatever reason, momentarily impossible, affection is the ‘effort’ that replaces it, ‘a motor effort on an immobilised receptive plate’.56
Similarly, the face is a ‘reflecting surface’, because it has sacrificed most of its motoricity: it has lost its movement of extension, and movement has become movement of expression. Its function therefore coincides with that of the affection-image, in that it brings forth a pure quality, rather than externalizing its affect by extending it into action. In Béla Tarr and Ágnes Hranitzky’s Werckmeister harmóniák (2000), for example, a sequence in which a mob brutally attacks hospital patients ends with the camera slowly retracing the protagonist, János, who has witnessed Beginning of page[p. 121] the scene, and staying fixed on a close-up of his shocked, gawk-eyed, unchanging expression for over thirty seconds (Fig. 2.2).

It is easy to draw parallels between the fixed shot of János’s face and his lack of motor response to the violence surrounding him. Yet Deleuze is careful to point out that these characteristics extend beyond the specificity of the human face. The same procedure can occur with shots that are the equivalents of a face, images that have been facialized — ‘visagéifiée’57 — or, in other words, ‘any type of shot that works to isolate and extract some kind of expressive quality in and for itself, any shot that has a “faciality”, any Beginning of page[p. 122] shot with an expressive intensity’.58 These can be faces, a multitude of faces, parts of the body, objects, or landscapes (‘any-space-whatevers’).59
There are, accordingly, different types of shots that can achieve the status of close-up, by opening up a synapse between perception-image and action-image to become, as Gregory Flaxman and Elena Oxman put it, ‘a hesitation, an affective intensity that prolongs itself in an expression in relation to something “as yet” to be seen or a “not yet” decided’.60 Accordingly, they interpret this hesitation as a moment of indetermination, before the fixing of a face through the process of facialization by the abstract machine. Deleuze does point to such a ‘line of flight’ when he suggests that ‘the close-up does indeed suspend individuation’.61 Although it becomes clear from the rest of his analysis that this is not a general principle of the affection-image, but one possibility amongst others, it is worth exploring this radical potential of the close-up, and the role duration occupies in it. Flaxman and Oxman posit that,
in as much as the affection-image enjoys a special relationship to the face, abstracting it from determinate milieus and expressing its affective singularities in unprecedented durations, it enjoys the occasional power to go through the face — to tear the face from its signification and thus to elude its own abstract-machine.62
Cinema, therefore, itself a machine for the production of faces, can also sometimes create lines of flight from which Beginning of page[p. 123] to escape or tear through the face. In Slow cinema, despite the rare use of literal close-ups, the kinds of shots that retain their expressiveness without being extended into action are plentiful, so that certain Slow films can be said, paradoxically, to be made up almost entirely of close-ups. This effect is enabled by the immobility of the camera, and a particular kind of slow tracking shot that evokes what Dreyer called ‘the flowing close-up’, placing the focus on the photographic qualities of the image.63 More crucially, however, this effect is also enabled by long durations. By this token, in Serra’s El cant, the extreme long shot of the Magi’s trek over the hill and back is a close-up, because it satisfies both poles of the affection-image: the infinite landscape is an immobile surface on which are inscribed the micro-movements of the Magi and the changes of light from the slow movement of the clouds.
The creation of close-ups in Slow cinema is aided by the lack of conventional narrative constraints, and those constraints being expressed through coverage: very rarely do we get reaction shots, eye-line matches, or cuts on action, for instance. Naturally, this is not always the case, and a film that is quite radically anti-dramatic, such as Lisandro Alonso’s La libertad (2001), can also make use of continuity editing in ways that undermine the close-up-ness of its images. In other films, such conventional devices can serve, instead, to accentuate the loss of face: the close-up shot of the dying goatherd in Le quattro volte, following a blurry point-of-view shot of a goat at the foot of his bed, signals the becoming-animal of the wandering soul (Fig. 2.3).

In Tarr’s Kárhozat, close-ups of surfaces such as walls, floors, doors, or windows are used to bookend sequences. These surfaces all relate to the element of water. Kárhozat is Beginning of page[p. 124] a film bathed in water, not so much for symbolic purposes but as an instigator of a play with light and shadow. The wet surfaces render the image a reflecting surface, to the extent that the film in its entirety becomes such a surface: Kárhozat is a mirror. Tarr and Hranitzky, who edited the film, cleverly use these close-ups as internal montage devices that contribute to the film’s idiosyncratic pacing. In one striking sequence, a close-up of a wet wall becomes a very slow tracking shot that reveals a silent crowd waiting, apparently endlessly (for the rain to stop? For the party to begin?). The camera moves slowly so that we have time to observe the constellation of faces before it reaches the next wall. As the camera tracks onwards, the wall closes in on the scene like a theatre curtain, only to open onto a different set of faces of bodies on its other side. The walls here function Beginning of page[p. 125] like the equivalent of a horizontal wipe into a black screen. The duration of the shot and the slow movement of the camera give us time to see the changes in the frame, and to focus on each face and body individually, turning the mass of people into a series of close-ups. As the camera slowly changes its position relative to the crowd, the focus shifts from one face to another, and a different body takes centre stage every few seconds. Tarr therefore translates what would have traditionally been a series of close-ups divided by simple cuts into an elegantly orchestrated sequence shot. There are no actual close-ups in this shot — and yet the entire shot is a close-up. Without extending into action of any kind, the expressive intensity of this palimpsest of Pasolinian features and bodies becomes a quality or virtual possibility. The ubiquitous element of water — the wet wall, the rain pouring down in front of the crowd like yet another wall, the sound of rain in the soundtrack — emphasizes the mirror-quality of the image, to the extent that the entire shot becomes a ‘reflecting surface’ with ‘intensive micro-movements’, in other words, a close-up (Fig. 2.4).64


How does the spectator feature in all of this? Tampering with duration in Slow cinema can foster a participation of the spectator that involves not only the temporalities on screen, but an interaction between those temporalities and the time-space of off-screen reality (in terms perceived as either positive or negative — boredom, for example). Issues of reality–virtuality, the crystalline aspects of cinema according to Deleuze, arise poignantly in the extended temporal experience of images of the quotidian.Beginning of page[p. 126]
Another way in which the schizoanalytic project of ‘losing face’ is enacted in Slow cinema is through the figure of the actor, most prominently the non-professional actor. In order to demonstrate how the close-up can escape subjectivity, Deleuze asks us to recall ‘that the actor himself does not recognise himself in the close-up’. Famously, he explains, during the editing of Ingmar Bergman’s Persona (1966), Liv Ullmann mistook a close-up of herself for the other protagonist, Bibi Andersson.65 In Slow cinema, the ambiguity sparked within the figure of the actor is not so much the result of a structured agenda of indeterminacy (after all, Bergman’s aim was precisely to create an amalgam of the two women through a calculated use of mise en scène and editing), as it is the result of an expansion of time and space, a tightening of attention and a loosening of control that allows for slippages and fissures. Serra’s Honor, for instance, oscillates between the strong,Beginning of page[p. 127] fixed, fictional frame of the Don Quixote narrative and a naturalized filmic universe completely stripped down to its essentials. Indications of the grand narrative the film visualizes are restricted to the names of the protagonists and their elemental costumes. Shot on location, without the use of artificial light or extradiegetic sound, and with minimal montage, improvised dialogue, and no narrative thrust, the film is essentially centred on the observation of the two non-professional actors and the Catalan landscape that surrounds them through a series of long takes, in an obsessive inspection of their presence. An extended close-up on Quixote’s face becomes ambiguous by virtue of this oscillation, because it transposes us from one space of signification to another, much as obsessive concentration on a word transposes us from the field of semantics to that of phonetics.
This ambiguity is not dissimilar to that which Laura Mulvey explores in her peculiar ode to the pause button, Death 24x a Second.66 Mulvey suggests that digital media enable a game of time that consists in the shift from fictional time to the time of registration of a filmic event. The halting of the flow of film, which became possible with home viewing technologies,
splits apart the different levels of time that are usually fused together. In detaching the time of the index from the time of fiction, the delayed cinema dissolves the imaginative power of the fiction, as well as the forward drive.67
When the image is paused, ‘the fictional world changes into consciousness of the pro-filmic event. As […] disbelief is Beginning of page[p. 128] no longer suspended, “reality” takes over the scene, affecting the iconic presence of the movie star.’68
In the context of the star system, this oscillation between fiction and reality is more pronounced, since ‘due to the star’s iconic status, he or she can be grafted only tangentially onto a fictional persona’ to begin with — the ‘pose’ becomes ‘pause’ and vice versa.69 But a similar effect can be achieved through an aesthetic of slowness rather than through literal fixedness, and through the unidentifiable face of a non-professional actor rather than the marked face of the film star. As we saw in the first chapter, when cinematic pacing is slowed down dramatically and narrative vectors deactivated, the same shift from film time to time of registration is enabled. If, moreover, we are not familiar with the screened face and thus not transported by the connotations of a star’s symbolic status, the new face we are confronted with can invite discovery and observation. Serra’s long takes take us from Don Quixote to an old man and back; we are looking at the same face, but we are forced to dwell on it so forcefully as to extrapolate the physicality of a human being from the fixed fictional bracketing of Quixote. Is this not the Deleuzian ‘smallest circuit’ in a single close-up shot (Fig. 2.5)?

Deleuze does consider the implications of the actual–virtual duality in the ontology of the actor:
The actor is bracketed with his public role: he makes the virtual image of the role actual, so that the role becomes visible and luminous. […] But the more the virtual image of the role becomes actual and limpid, the more the actual image of the actor moves into the shadows and becomes opaque.70Beginning of page[p. 129]
In his examples, however, it soon becomes clear that Deleuze is referring to the actor as film character, to opacity as a narrative theme. As in Tod Browning: ‘there will be a private project of the actor, a dark vengeance, a strangely obscure criminal or justice-bringing activity.’ In The Unknown (1927), for example, ‘a fake limbless man takes to his role and has his arms really cut off, for the love of the woman who could not bear men’s hands, but tries to recover his dignity by organising the murder of a rival who is whole.’71 Deleuze’s examples therefore move away from what he sets up as a relationship between the character and the actor in his theoretical exposition.Beginning of page[p. 130]
What happens instead if we turn to the crystalline aspects of the actor in relation to the pro-filmic? Robert Bresson’s model acting method was famously based on draining repetition and the precise direction of gesture and movement in order to reduce his non-professional actors’ performances to a monotone, automatic rendition that would supposedly allow for the model’s pure qualities to come to the surface. The non-professional actor who finds him or herself inside the brackets of a ‘public role’ in the Bressonian method appears to me to enact just this liminal positioning of the crystalline circuit: the actual image of a body (the limpid side of the crystal) is projected into virtuality through the bracketing that the enactment of a character requires. The crystal circuit completes itself through the use of the virtual (the film character) as a tool that grinds the acting down through exhaustive repetition to a transcendental quality that belongs to the actual (the person portraying the character). Bresson himself, in his Notes sur le cinématographe, evokes this paradoxical double capacity of the model for limpidity and opaqueness.72 The model lacks naturalness, but he or she does not lack nature, he says: ‘Draw from your models the proof that they exist with their oddities and enigmas.’73 His objection to the use of trained actors is best and most beautifully captured in a short parenthetical note: ‘(without shadows)’.74 The straightforward intelligibility of the trained actor, a lack of shadows, elegantly resonates with the Deleuzian opaque. The trained actor ‘is not there’;75 they are absent by virtue of being in control of their performance. Non-professional Beginning of page[p. 131] actors, according to this logic, are there despite themselves: ‘it is [their] non-rational, non-logical “I” that your camera records.’76
Working without a script, Albert Serra often models his sequence shots on the basis of his actors’ confused improvisation, in ways that recall Bressonian terminology:
When an actor said something I liked, I could ask him during shooting to repeat it over and over. I realised that this method would result in strange behaviours, ambiguous gestures, without signification. And the film was constructed on the basis of this aberrant, meaningless matter.77
The virtuality of the mythic Quixote is stripped bare through the extended take, the naturalized rendition, and Bressonian acting method, which occasionally generates veritable fissures in the fiction: momentary glances at the camera, or repeated words that announce themselves as different takes of a particular line. At the same time, this peek into the limpidity of the pro-filmic is, in its turn, in the process of becoming opaque once again, constantly rewoven into the fiction.
Much of this becomes possible thanks to duration. As Mark Le Fanu says of Tarkovsky’s Stalker, the face of the homonymous protagonist becomes shrouded in ambiguity by virtue of its relentless exposure to our vision. Tarkovsky’s camera, in Le Fanu’s words, looks at ‘Kaidanovsky (Stalker) not as though he were an actor declaiming portentous lines, but as though he were, somehow, a landscape: unique, weathered, sculpted and natural. […] In Beginning of page[p. 132] sum, he is opaque and ungraspable’.78 These two facets, natural and virtual, limpid and opaque, become indiscernible and crystallize in the face-landscape of Tarkovsky’s Stalker, Serra’s Quixote, Monteiro’s João, Costa’s Vanda.
An actor’s crystalline aspects can create connections between the actual and the virtual, or, using Mulvey’s terms, an oscillation between the fictional and the pro-filmic that incites ambiguity. For Mulvey, in the cinema of delay, such as that of Abbas Kiarostami’s Taste of Cherry (1997), narrative flow is suspended from within the film, and replaced with the observation of ‘absences in representation’. These absences are conventionally displaced by the dogma of narrative continuity, a system of ordering that is imposed on the image externally.79 The long undramatic takes of Slow cinema, contrarily, extract or deterritorialize blocs of time and space from the grid of strict narrative flow. The temporal insistence, exposing these ‘close-ups’ of time-space for durations that far exhaust signification, reminds us of that childhood game of extracting a single word from speech and repeating it to exhaustion. As MacDougall describes,
When we repeated a familiar word over and over again — a word like ‘hippopotamus’ — sign and referent begin to separate until the sign became an unrecognizable phonetic pattern. It then became subject to the mispronunciations that occur with tongue-twisters. A kind of verbal searching led to a play on alternative stress patterns (hippopotamus), picking out new signs previously hidden in the word — hip and pot, for example.80Beginning of page[p. 133]
The pleasure derived from such a children’s game lies in undoing that other grid of signification, ‘the linguistic codes of adults’.81 In a similar way, in Honor, the obsessive durations and loosened control, allowing fissures and failures to leak out, can escape the apparatus of signification. The face of Quixote escapes itself by cheating on its own fictionality, its own foundational gridding as Quixote. By eluding its own identity and becoming different (or difference), it eludes our grasp as well, no matter the intensity of our gaze.
Lisa Åkervall sees the structural closeness between the affection-image and the time-image as an important dimension of Deleuze’s cinematic project. The implied liminal tension common to both opens up a space of ‘the unbearable’ that is central in the designation of the Deleuzian conception of modern cinema. The pure optical and sound situation, Deleuze says, makes us ‘grasp something intolerable and unbearable’, as in Antonioni.82 Åkervall sees this as transforming the cinematic experience from one of watching to one of ‘enduring’.
The concept of ‘enduring’ is linked to an insistence inherent in the affection-image that is ‘the non-actualizable part in experience, the part which transcends any actualizability and ordinary experienceability and thus always exists, or better: insists in a state of virtuality, of always unactualized potentiality’.83 Åkervall’s notion of ‘enduring’ is also informed by a sense of the insistence of time, or the Beginning of page[p. 134] persistence of something to be endured through time. Extended duration makes palpable qualities that are situated beyond the dramatic locus of action, and induces a type of vision that is both obsessive and investigative, but which also has to be endured, due to its temporal hyperbole.
Endurance is, in fact, a term that frequently arises in relation to Slow cinema. But the ‘too beautiful’ or ‘too terrible’ of Deleuze’s unbearable is joined by a ‘too slow’ or ‘too long’ in Slow films. Reviewers and critics often refer to films as endurance tests, with Malin Wahlberg calling Pedro Costa’s Vanda ‘a screen event to be endured both physically and morally’.84 The intolerable in Vanda therefore consists both in our uneasy position as spectators of intimate, harsh realities, as well as in the physical difficulty of sitting through a film that tracks an eventless dailyness for three hours. As Glyn Davis rightly notes, to endure as a spectator of a Slow film often involves enduring along with the characters in the films.85 They are often waiting, sometimes endlessly, or engaging in repetitive or menial activities, often without so much as the relief of speech or interaction, therefore grafting the intolerable into the everyday. Deleuze sees in Ozu, that great ‘critic of daily life’, someone who ‘picks out the intolerable from the insignificant itself’, from the banal and the quotidian. Sensory-motor situations, on the other hand, ‘no matter how violent, are directed to a pragmatic visual function which “tolerates”’. The time-image introduces a ‘new type of character for a new cinema’, one who knows ‘how to extract from the event the part that cannot be reduced to what Beginning of page[p. 135] happens’, a visionary.86 This function is not limited to the character alone, but to the viewer as well: ‘The important thing is always that the character or the viewer, and the two together, become visionaries.’87
Deleuze reminds us that this type of post-war cinema, of the seer rather than the agent, was accused by Marxist critics of being politically problematic. They saw it as being too passive, bourgeois, or marginal, because it replaced action and agency with a confused vision, and because the characters were unconcerned, even by their own fates. Deleuze countered that it is precisely in this loosening of connections that modern cinema becomes ‘completely political’. The visionaries’ stunted motor-capacity proceeds from the acknowledgement that what happens to them does not belong to them. In normal vision, according to Bergson, we always perceive less than the entire image, because we perceive what it is in our interest to perceive, by virtue of our economic interests, ideological beliefs, and psychological demands. In other words, we perceive clichés, or sensory-motor images of things.88 However, when the sensory-motor schemas jam or break, the visionary is exposed to pure optical situations, things in themselves and in their unbearable beauty or horror. Of course, this does not always suffice, and Deleuze factors in the possibility of such images creating their own clichés.89 The post-war character does not act, therefore, inasmuch as he ‘sees better and further than he can react’.90
There is an ongoing discussion about the Deleuzian seer in contemporary art cinema, and in fact it seems like Beginning of page[p. 136] a fit too perfect to resist. According to Julian Reid, the political relevance of this figure is now questionable. For Deleuze and Guattari, capitalism is an ‘immanent machine’ of deterritorialization and decoding,91 where capital has become itself the flow of exchange. Capitalism’s ability to subsume everything, from avant-garde art to intersectional feminism, is well understood.92 In such a context, the seer is no longer a figure of radicalness or of self-understanding, because vision itself has been co-opted. Reid proposes instead that we ‘start by presupposing the existence not simply of a modern but a postmodern cinema in which the reterritorialization of the seer by the state is the very problem at stake’.93 As he points out, the reterritorialization of the seer by capital and the state has been thematized more and more frequently in recent moving image production. He gives the example of Lee Tamahori’s Next (2007), in which a man with prophetic capacities (the ability to ‘see’ the future) is recruited by the FBI, against his will, to assist with an anti-terrorist mission. Another example might be Charlie Brooker’s TV series Black Mirror (2011–). In the episode White Bear (2013), the official state punishment of a woman who watched and filmed her boyfriend kill a child is given the form of a spectacle, in which families are invited to watch and film the woman on their phones as she goes through terrible ordeals. In the ‘if you see something, say something’ society of the twenty-first century, Reid appears to be saying, the figure of the seer has become wholly recuperated, itself a cliché. In fact, none of the senses, neither Beginning of page[p. 137] vision nor tactility, are now adequate — ‘that cinema, of the senses, is now dead’, he proclaims. I tend to agree with Reid. However, in Davis’s notion of endurance as a central trope of Slow cinema, both in its bodily connotations and as an intensity of vision,94 a particular mutation of the Deleuzian seer might still prove to be germane to contemporary cinema.
In Béla Tarr and Ágnes Hranitzky’s late films, seeing is a central conceit, at once an everyday activity, an expression of endless waiting, a way of being in the world, a method of cinematic construction, and a reminder of and comment upon our viewing position. The camera, uninterrupted by montage, enacts a kind of special vision, slipping slowly from one character, object, landscape, or situation to another, making visible the connections between them, like points on a common surface. In Kárhozat, as in Sátántangó, A londoni férfi (2007), and A torinói ló, protagonists engage in frequent and intent sessions of looking out of the window. In the opening sequence of Kárhozat, we are introduced to a fixed shot of an empty and silent landscape, with the exception of the monotone, lethargic movement and accompanying screeching sound of coal buckets moving back and forth on cable cars overhead. This fixed shot turns into an excruciatingly slow retreating camera movement that reveals, in turn, a frame-within-the-frame in the form of a window, then puffs of smoke prefiguring a human presence, and eventually the backlit outline of the bust of the protagonist seen from behind, where the camera settles (Fig. 2.6). We are left to look at this operation of looking for many seconds. Halfway through the film, the same man, Karrer, tells his lover:Beginning of page[p. 138]
I sit by the window and look out in vain. For years and years I’ve been sitting there. […] Yet I don’t cling to anything. I don’t cling to anything but everything clings to me. They want me to look at them. To look at the hopelessness of things.
Jacques Rancière rightly detects the intimation of the director’s filmmaking method behind Karrer’s words:
Karrer’s intimation is at the same time a declaration of cinematographic method. Béla Tarr films the way things grab onto individuals. Things, such as the tireless cable cars in front of the window, the decrepit walls of buildings, the stacks of glasses on the bar counter, the sound of billiard balls or the tempting neon of the italic letters at Bar Titanik. This is the meaning of the opening sequence: it is not the individuals who inhabit spaces and make use of things. It is rather things themselves that go to them, that surround them, penetrate them or reject them.95

The agency that Rancière attributes to objects makes the concomitant passivity of the human subject’s gaze even more poignant. The things, ‘they’, want Karrer to look at them, and he obliges, with an obsession that seems to lack any ideological anchoring — ‘in vain’. Tarr’s camera doubles this operation of looking that is effectuated by the protagonist. The duration of these shots, accompanied by a lack of action or narrative progression and the slowness of the camera’s pace, imbues them with a sense of the persistence of something non-actualized or non-actualizable that one must endure. Karrer must look ‘for years and years’. As he endures this activity imposed on him by ‘things’, through a mise en abyme of the gaze, first the camera, then the spectator will reduplicate it.Beginning of page[p. 139]
I want to turn to a film that stands out in relation to this foregrounding of vision: in Werckmeister harmóniák, the activity of vision does not simply lead to a climactic inertia, as in Kárhozat or, most poignantly, A torinói ló, but ends up destroying the seer. As opposed to its usual polyphonic point of view, in Werckmeister, Tarr’s camera is attracted to a single character, János Valuska.96 With the camera on his leash, it is his itinerary we follow from one Beginning of page[p. 140] end of the ordinary small town to the other, as it is gradually overtaken by the fear and unrest caused by a travelling circus and its main attractions, the enormous corpse of a whale, and the Prince, a mysterious diminutive man. János, the local paperboy and holy fool, is wide-eyed like a child, but has the protracted chin and compressed jaw of an older man. He is in many ways the quintessential Deleuzian post-war powerless onlooker of a situation that is beyond his control. His positive demeanour and innocence set him up as Deleuze’s child seer, who, like Edmund in De Sica’s Germania anno zero (Germany Year Zero) (1948) or Truffaut’s Antoine Doinel in Les quatre cent coups (The 400 Blows) (1959), ‘in the adult world, is affected by a certain motor helplessness, but one which makes him all the more capable of seeing and hearing’.97 Rancière equally sees János as a seer, a voyant. His adventures are primarily visions, he says, and it is visions that affect him: the black mass of the immense container driving slowly past him in the night, the ominous crowds waiting in the town square, the body of the oceanic monster.98
But János’s innocence and his consequent demise do not presage a simple battle of good and evil at the former’s expense. The trope that most forcefully inhabits the film is rather that of order and disorder. What they represent, however, is not a binary of good and evil, with the order of the cosmos on one side and the brutality and chaos of mob violence on the other. The two categories are imbricated, constantly and seamlessly and on all levels, from the microscopic to the cosmic. The opening image announces the amalgam of forces in the everyday object of a coal stove: a contained fire. The same opening sequence demonstrates Beginning of page[p. 141] that the cosmos in Werckmeister is not symbolic of a majestic neatness, but the awkward, drunken dance routine of a group of middle-aged Hungarians, orchestrated by János, the town fool, in a drab local pub (Fig. 2.7). Conversely, the chaos that erupts towards the end of the film, causing the local men to enact senseless acts of violence, is in fact kindled by the forces of order and fascist rule. The shadowy minuscule figure of the Prince, whom we never see and whose mere presence, rumour has it, is the cause of unrest in all of the towns visited by the circus, provides only a mysterious excuse for the breakout of civil disorder. It is Mrs Eszter who, having formed a committee of propriety for the establishment of ‘order’ in the town, and paired up with her lover, the chief of police, ultimately instigates the madness and profits from it to establish police rule and take over the town. The unnervingly ordered and silent march of the mob on their way to the hospital to wreak havoc confirms this visually. No brows furrowed in anger, no words of rage or protest, but an utterly muted constellation of melancholic expressions, with the thumping sound of footsteps and the puffs of smoke from frozen breaths giving the impression of a not-so-well-oiled machine.

János’s belief in the beauty of the cosmos is mirrored in Mr Eszter, his reclusive friend and Mrs Eszter’s ex-husband, a man well respected for his social status and musical scholarship. Mr Eszter piously believes in an order of his own: he has discovered that the well-tempered harmonies invented by Andreas Werckmeister in the late seventeenth century are of a disingenuous artificiality that goes against the purity of natural pitch. Driven by this belief, he discords his piano and engages in the frustrating mission of playing the masterworks of Western music on the twelve natural half notes, with cacophonous results. The question arises of what counts as order: the false system of division of the Beginning of page[p. 142] twelve half notes into twelve equidistant parts, accepted by Werckmeister’s contemporaries, and according to which the greatest of Western music was composed; or the natural harmonies of absolute pitch, akin to the harmonies of the cosmos, but which, when used on all those ‘most wonderful harmonies and most sublime mutual vibrations’,99 throw them out of kilter? After the violence has ended, and order has been established — the police order of the chief and Mrs Eszter, who also establish themselves in Mr Eszter’s house, forcing him to sleep in his kitchen — the Beginning of page[p. 143] old man reassures the irreparably traumatized János that he can come live with him when he leaves the psychiatric hospital, and that he has retuned his piano to Werckmeister’s false harmonies. At stake in the film is precisely this uncomfortable proximity between what Rancière calls ‘the normal disorder of the order of things “dis-illusioned”, and the extremes of destruction and madness’.100
Neither his guilelessness nor his capacity to see, crystallized in his illuminated eyes, make János a morally privileged figure. Despite their best intentions, János and Mr Eszter, whose social standing and erudition will not be enough to salvage him either, will become accomplices in the machinations of Mrs Eszter. The seer is thus, unbeknownst to him, recuperated by the powers that be. Or, to say it conversely, neither the ‘innocent fool’, nor the gentlemanly savant will ultimately be able to see enough or well enough, their vision revealed to be partial and inadequate.101
János’s status as the innocent onlooker who is partly blamed and eventually also pays for the catastrophe that engulfs the town is echoed in the body of the giant dead whale. When János is addressing the enormous eye of the circus attraction, in one of the most haunting moments of the film, he says: ‘see how much trouble you’ve caused, even though you haven’t been able to harm anyone for a long time’ (Fig. 2.8). The whale becomes a magnet for civil unrest: like another sun, the massive body of the whale attracts the smaller bodies of the men, who gather around the town square in waiting. As Rancière points out, despite the various rumours circulating about the malicious intent Beginning of page[p. 144] of the crowd, its threat lies primarily in its being there without a reason, with nothing to do but wait.102 This other cosmos, fashioned through the civil frustration of a life that promises nothing and leaves no room for action, bursts into a violent darkness, like the total eclipse happily staged by János at the beginning of the film. Madness reigns for a while — ‘will Heaven fall upon us?’ — but ultimately, as in János’s lesson in cosmology, order is reinstated.

The camera does not rise beyond the chaos any more than the rest of the characters do. Six out of the thirty-seven sequence shots of the film are almost entirely devoted to Beginning of page[p. 145] János walking, wandering the streets of the small Hungarian town. The role of both János and the camera is that of flâneur and observer. Most often, the camera is an observer of János — he appears in all but two sequences in the film. Right before János visits the whale for the first time, we see him in the main square. As usual, he appears in the shot from behind a fixed camera, jolting it into movement. With the camera in close pursuit, János, big eyes and mouth half open, wanders about in the square, observing the men he encounters with a mix of fascination and fear. At first, the camera stays focused on him, gradually becoming caught up in his activity, mirroring his moves in what ends up resembling a dance duet, akin to the cosmic dance that opens the film. The two of them envelop the men in their entranced double gaze — a gaze that, as we have seen, proves to be a futile, and ultimately self-destructive exercise. The guileless János is implicated, blamed, persecuted, and eventually sectioned. The implication seems to be that there is no redemption for the passive observer. The other onlooker, the camera, is inevitably equally implicated: if cinema is an observer of reality, Werckmeister appears to be saying, it does not follow that this observation is unfiltered or innocent.
This brings us to the question of the spectator. Having traced a range of properties — the imbrication of film time and pro-filmic time, the centrality of vision, the insistence of observation, endurance — that are predicated on long duration, I want to bring this framework to bear upon the different modes of spectatorship it implies. In Slow cinema, the viewer is prompted to mirror the intensity of the protagonists’ insistent observation. But what are the terms of Beginning of page[p. 146] this contract between film and spectator? Without a doubt, they are partly conditioned by discourse, situating Slow cinema between the two poles of the ‘boring art film’ and the privileged art form that places the burden of misapprehension on the shoulders of a spectator lacking aesthetic sensibilities. The socio-political connotations of spectatorial reception will be considered in the fourth chapter, but for now, I would like to focus on the philosophical issues that arise from the relationship between a Slow film and its viewer.
With Deleuze, the question of spectatorship will always be vexing, since Deleuze does not provide anything resembling a systematic account of spectatorship in his work on cinema. This problem has been superseded in different ways, for instance, by muting certain aspects of his theory in favour of an emphasis on his affinities with phenomenological thought, as Laura Marks has done.103 This is also questionable, since disputing the subject–object relationship is one of the fundamental operations in Deleuze’s thought. The closest Deleuze comes to discussing the spectator is through reference to ‘psychomechanics’ and the ‘spiritual automaton’,104 which he subsequently redescribes as ‘pre-individual singularities’ or ‘non-personal individuations’.105 Such definitions have been criticized for implying an essentialist view of the spectator as a hovering substance without a social, cultural, or political background that they bring with them to the activity of Beginning of page[p. 147] viewing.106 This is indeed a risk. At the same time, Deleuze leaves open a different possibility. The relationship between spectator and film can be formed in the process of viewing through the formation of linkages. There is no guarantee this will happen: we have seen Costa bemoaning the situation in which a spectator ‘only sees himself’ in films in which ‘he recognises everything’,107 for example, implying a relationship in which both spectator and film are ‘facialized’, determined subjects. Instead, if we posit a spectatorial constituency that has the ability to draw ‘lines of flight’ from its determined identity, and films that are equally capable of ‘losing face’ in the ways we explored above, we allow for more malleable encounters and, ideally, for something resembling a radical effect.
Another way to describe this relationship between film and spectator is by way of Rancière’s politics of aesthetics. From this point of view, we can see the effect of films on the spectatorial synthesis as proceeding from the operation of making visible what is normally invisible in the current distribution of the sensible. Rancière defines the distribution of the sensible as ‘the distribution of spaces, times, and forms of activity that determines the very manner in which something in common lends itself to participation and in what way various individuals have a part in this distribution’.108 So, making the invisible visible can mean three things, as we will see in more detail in the following chapters: to bring into view constituencies and modes of Beginning of page[p. 148] living that are usually partitioned or systematically muted, like Tarr’s ‘ugly’ men and women,109 Costa’s immigrants and sub-proletarians, or Frammartino’s animal or mineral life; to make visible the hegemonic distribution of the sensible so as to depose it from its claim to a ‘natural’ order of things; or to do both of these things in formats and modes of presentation that redistribute the normalized order of presentation itself.
This third premise is enacted in Slow cinema primarily through the treatment of time. We have established that directors of slowness are especially conscious of duration within shots and sequences, and in the relationship between shots and units of meaning. This attitude can be traced back, as we have seen, to filmmakers such as Antonioni and Akerman, whose lessons Slow directors appear to have pushed to their extreme. What András Bálint Kovács calls ‘inverted dramatic construction’ in Antonioni, a procedure of transference of dramatic peaks from the end to the beginning of the film,110 becomes an elimination in Slow cinema. In Antonioni, says Kovács, plot is still related to a conflict,
but what follows the exposure of the conflict is not the solution of it, but rather the ‘eternalizing’ of it by emptying out the initial situation of its dramatic tension. […] The principle meaning of the situation the characters find themselves in is thus continuity, or eternity of the situation’s existence.111Beginning of page[p. 149]
The elimination of dramatic peaks, what Kovács calls the ‘radicalized minimalist continuity style’, is effected through this altered pacing, structured around the passing of time and the observation of characters.
Nevertheless, we need to keep in mind that spectatorship is formed through film history, and what is eventless for one film audience constituency is flamboyantly eventful for another. Vittorio De Sica’s Umberto D (1952), to use a frequently cited example, or Rossellini’s Viaggio in Italia (Journey to Italy) (1954), films that were scandalizing in their uneventfulness to contemporary spectators, may seem to have a rather conventional flow today. Or, as examples from the other end of the spectrum, films like The Bourne Identity (2002), The Bourne Ultimatum (2007), and Transformers (2007), that were initially criticized for their fast cuts and ‘queasicam’ style verging on incomprehensibility, have now set the pace for Hollywood film speed, at an average shot length of under two seconds.112
One way of addressing this difference is through what Deleuze calls a ‘threshold of perception’ that allows events to emerge from imperceptibility to the perceptible. Translating this into Bergsonian terminology, there are different ‘layers or strata of processes’ that we need to leap into in order to see them and to recognize the events occurring within them. In this view, since beings need to make their lives eventful, they effect a ‘de-temporalization’ or condensation or suppression of ‘other times and the times of others’, choosing, as it were, the time stratum in which their lives can become meaningful. ‘Only unexpected forces’, says John Mullarkey, ‘can extend [this stratum] to offer Beginning of page[p. 150] a glimpse of what lies beyond.’113 Like Deleuze, Mullarkey illustrates his argument of perception as suppression through narrative metaphor: the Scalosians in the Star Trek episode ‘Wink of an Eye’ (1969) live at a different speed, which is imperceptible to the crew of the Enterprise; the catatonic patients in Awakenings (1990) seem immobile because they move at a speed imperceptible to the doctors.
As Rancière says of Deleuze, Mullarkey treats ‘these fictional situations […] as simple allegories’.114 But it might be more interesting to look at the filmic image itself for evidence of this argument. What are the ‘unexpected forces’ that have the potential to dislodge our perception and launch it into different temporalities in the example with which Mullarkey begins his discussion of the threshold, Tarr’s Sátántangó? Can we posit that this film, along with Tarr and Hranitzky’s entire oeuvre and, perhaps, Slow cinema tout court, is an apt candidate for the transposition of the threshold of perception if thought of in terms of its difference in relation to mainstream narrative cinema? In Tarr and Hranitzky (or Serra, Monteiro, Hou Hsiao-hsien, or Weerasethakul), the long takes and fixed cameras bring to an abrupt halt the eye trained in speed, but the halt does not amount to an eventless stasis. This movement from acceleration to arrest is in itself a (cultural) event, a Badiouian interruption, an unexpected force that shakes our perception into a revelation of what ‘lies beyond’, the discovery of a different plane of time. On the one hand, this is indeed a question of the particular cultural moment — the binary discourse according to which Hollywood is becoming faster and ‘queasier’, so for art cinema to be radical Beginning of page[p. 151] it must be shockingly slow. On the other hand, it is a perceptual experience: we suddenly become aware of duration as it rises to the status of protagonist. In Sátántangó, says Yvette Bíro, ‘man is surrounded by a dead past and a dead future. There is only the present stretching to infinity […]. Intention is replaced by distention: the prolongation and inexorable extension of time.’115 Instead of a time of linear development — the time of cause–effect links, of an action-image of the before and after — we experience something more akin to a disorienting eventlessness that vexes the limits of our ‘thresholds of perception’.
What happens then in a film? Is the filmic event something tangible that a spectator identifies (or misses)? Is it, following Deleuze, that which is ‘always and at the same time something which has just happened and something about to happen; never something which is happening’?116 If so, then the perceived eventlessness of film, a prolongation of time where ‘nothing happens’, is only natural because nothing ever happens: the event is either about to happen or has just happened or both at once. Tarr emphatically invokes this approach to (filmic) time:
I despise stories, as they mislead people into believing that something has happened. In fact, nothing really happens as we flee from one condition to another. Because today there are only states of being — all stories have become obsolete and clichéd, and have resolved themselves. All that remains is time.117Beginning of page[p. 152]
But what if the spectator perceives the imposition of a different pace as uninteresting, of duration as boring? Mullarkey asks:
Why do some people tire more easily than others before some films […] and not others? Length does not seem to be important […]. It is the units of meaning and so the breadth of perspective that surely counts here. Mere quantitative changes in the parts […] can be seen by those who endure them as qualitative changes in the whole […]. But one must be patient to see the parts accumulate. This is not to say that there is one ideal, and very patient, spectator for every film, for it is always possible to watch for longer, watch more often, or watch more closely. Rather, there are different films that emerge with different types of spectatorship, and different types of spectator that emerge with different films. And this all happens in the event(s) of viewing.118
According to one line of thought, then, cinema, as a temporal medium, imposes its own duration on the spectator.119 As new media technology has accelerated the perforation of the boundary between the cinema and the gallery, however, the contract between the spectator and the visual image has altered. Already with the introduction of VHS in the seventies, and increasingly so since, the temporal length of a film has become unhinged from singular duration, as we have learned from Mulvey. A film Beginning of page[p. 153] can be paused, fast-forwarded, slowed down, watched over the course of a few days. Psycho does not have to be 109 minutes long, but can last, as in Douglas Gordon’s 24 Hour Psycho (1993), 1440 minutes. (Even more radically, Gordon’s 5 Year Drive-By (1995) turns John Ford’s 113-minute-long The Searchers into a five-year-long screening, to match the timeframe of the film’s story.) Gordon’s Psycho, an installation consisting of Hitchcock’s film projected at a slowed down rate of roughly two frames per second, marks the shifting relationship between spectator and moving image with the advent of the digital. No longer is the spectator expected to equate their experience of the visual with the temporal limits of the artwork on display, but they can walk in and out of the exhibition room at any time, structuring their own experience of duration.120 In many cases, this experience of spectatorship is predicated on decisions about boredom.
In his lengthy discussion of the notion of boredom in The Fundamental Concepts of Metaphysics (1983), Martin Heidegger posits that boredom is a temporal term intricately related to the ‘problem of time’,121 in the sense that ‘it is precisely in passing the time that we first gain the correct orientation in which we can encounter boredom undisguised’.122 Boredom is predicated on the passing of time, that is, and in fact on an ‘imposed experience’ of the passing of time.123 One can abandon the video installation Beginning of page[p. 154] hypothesized above as soon as interest wanes. In the (traditional) experience of film watching, however, there is an implied contract: one is ‘firmly stuck’ in the cinema theatre.124 As Richard Misek points out, this makes cinema ‘a privileged site of boredom’.125 We can widen this claim to account for the expansion of exhibition formats. Cinema is not a privileged site of boredom because of the specificity of the cinema theatre; in many ways, the implicit temporal contract is retained in relation to film viewing in different formats. In other words, as opposed to a looped installation, for example, a film does not negotiate its own temporal limits. We understand that we are expected to watch a film from beginning to end, whether at a theatre, on a laptop, or on a mobile phone. Even if we decide not to engage fully with that expectation, we are aware of the consequent distortion of our experience of film viewing.
Sometimes we may find ourselves ‘stuck’ in a temporally delimited situation such as watching a film, as Misek suggests, or waiting at a railway platform, as in Heidegger’s example. In those cases, looking at our watch might be a sign that ‘our passing the time is not really succeeding’, and we wish to know ‘whether we must continue to struggle against the emergent boredom by this unsuccessful killing of time, strangely lacking in any goal’.126 How is it that we resort to standard, measured time to achieve some kind of insight into our subjective experience of the unsuccessful passing of time? We wish to measure our experience in relation to what, at that moment, registers as a different kind of time, time that passes ‘more quickly’.127 By extension,Beginning of page[p. 155] Heidegger concludes that time, in that particular instance, is going slowly. ‘Does being bored then mean grasping the fact that time is going slowly?’ he asks; ‘and where does this slowness stem from? In what does this slowness consist? Is it because time is too long?’128
According to Dunbar, Yossarian’s friend in Joseph Heller’s Catch-22 (1961), the answer is clearly yes. Boredom slows time down, which has the useful consequence of prolonging life:
Dunbar was lying motionless on his back again with his eyes staring up at the ceiling like a doll’s. He was working hard at increasing his life span. He did it by cultivating boredom. Dunbar was working so hard at increasing his life span that Yossarian thought he was dead.129
And again:
Dunbar loved shooting skeet because he hated every minute of it and the time passed so slowly. He had figured out that a single hour on the skeet-shooting range with people like Havermeyer and Appleby could be worth as much as eleven-times-seventeen years.130
For Heidegger, slowness that leads to boredom is understood to mean a slowness in which time is dragging. Becoming bored, by this token, is ‘a peculiar being affected in a paralysing way by time as it drags’.131 As Misek points out, in Heidegger, becoming bored by something is a subject–object relationship. Everything has ‘its time,’ Heidegger Beginning of page[p. 156] says, and that is what makes boredom possible. Time that is too slow, or time that drags, is thus, from a phenomenological point of view, a matter of the mismatch between the time of an object and our own time. In a film, Misek concludes, boredom stems from the ‘mismatch between the “ideal” speed at which we want it to move and the actual speed at which it moves’.132
As opposed to boredom, Heidegger tells us, killing time or making time pass means to seek ‘not to be occupied with time, not to dwell upon time or to ponder it’.133 Passing the time, then, paradoxically, means striving to stay unaware of the passing of time. As opposed to the long wait at a railway station, when we engage in an activity in order to kill time ‘we do not pay attention to time at all’.134 This, for Heidegger, should not be a desirable effect. Time, on these occasions, does not merely speed up, but becomes, as the expression implies, annihilated, cancelled out.
As Misek points out, ‘though railway stations are indifferent to the effect that imposed duration has on us, filmmakers are not’.135 Thus, following Heidegger, if certain films help us to pass the time without noticing it, they are responsible for our killing of time. Chantal Akerman addresses this as a problem of the ethics of cinematic time. She sees certain forms of narrative cinema as being made to ‘escape time’. People might say, ‘“Oh, I had a good evening. I didn’t see the time passing by.” Well,’ she says, ‘they were robbed of two hours of their life.’136 The slow, uninterrupted observation of Jeanne Dielman’s household chores Beginning of page[p. 157] in Akerman’s eponymous film thus becomes a question of ethics. That one might become bored in watching Jeanne cook a meal of veal cutlets is not a problem at all. Instead, it is ‘profoundly ethical’.137
The inability to see or feel time pass is an absence that implies another. For Siegfried Kracauer, it is only in being bored that one can feel one’s own presence:
Boredom becomes the only proper occupation, since it provides a kind of guarantee that one is, so to speak, still in control of one’s own existence. If one were never bored, one would presumably not really be present at all.138
This transforms boredom into a potentiality that Heidegger also detects in ‘attunement’, which is achieved through a sense of profound boredom.139 For Kracauer, the revolutionary potential of boredom (which his friend, Walter Benjamin, also wrote about in the Arcades Project),140 ‘the unfulfillment from which a fullness could sprout’, is undermined in time-passing activities such as movie-going. When watching a film, he says, one’s spirit ‘transforms itself into a trained dog that performs ludicrously clever tricks to please a film diva, gathers up into a storm amid towering mountain peaks, and turns into both a circus artist and a lion at the same time’.141 Instead, Kracauer implores his readers to ‘hang about in the train station’, like Heidegger’s bored subject, or stay at home and surrender Beginning of page[p. 158] themselves to boredom. If one is patient enough to endure the restlessness of boredom, according to Kracauer, one can experience ‘a kind of bliss that is almost unearthly’.142 Kracauer’s panegyric to boredom describes a sense of felt presence similar to Heidegger’s ‘attunement’, a less ecstatic way to describe this sense of heightened consciousness instigated through the state of being bored.
In this sense, Akerman’s aforementioned proclamation echoes in reverse a different complaint, wherein the spectator describes experiencing a boring film as a ‘waste of time’.143 As opposed to Akerman’s, this latter lament is not phrased in ethical but in economic terms. Time wasted equals time invested unsuccessfully, the lack of success being measured in returns. As we will see in detail in the fourth chapter, these returns are in turn linked to ideas about narrative efficiency and productivity. In fact, it is these capitalist formulations that are eluded through boredom, according to Kracauer’s definition. Boredom appears to function as an improbable wake-up call against a kind of ‘boundless imperialism’:144
People whose duties occasionally make them yawn may be less boring than those who do their business by inclination. The latter, unhappy types, are pushed deeper and deeper into the hustle and bustle until eventually they no longer know where their head is, and the extraordinary, radical boredom that might be able to reunite them with their heads remains eternally distant for them.145
For Kracauer, boredom thus becomes a necessary, cathartic state of being that salvages time and consciousness from Beginning of page[p. 159] annihilation. It is held up as the true sign of a subject that has not been taken over by what Henri Lefebvre called the ‘colonization of everyday life’, the usurpation of the subject by a modern culture of commodities and the increasing regulation of time. In other words, for Kracauer, mass culture designates the rejection of unstructured, empty time, making boredom a radical gesture against the saturation of images and events.146 Being bored during the viewing of a Slow film, by this token, may not only be a valid response, but an encouraged one.
The possibilities Slow cinema presents for an expansion of perception by refocusing on the mundane and undramatic are not necessarily precluded by boredom. Boredom can undoubtedly make for a negative viewing experience. Slow films are often ‘wasting time’ themselves, however, and in fact flaunt their capacity to restage the uneventful as a meaningful experience. We have seen that, along with the spectator, characters also have to look endlessly and endure the dead times of mundanity. By this token, one can be said to be bored with a Slow film, rather than by it.
The critique of mass culture by Heidegger and Kracauer persists in contemporary thought. Paul Virilio bemoans the collective transmutation of seeing effected by technological means of representation, which set up a ‘false day’ of electronic illuminations against the day of human experience.147 Bernard Stiegler similarly relates the way film, television, and more generally, digital technologies interrupt our sense of time and impose a synchronization of temporal experience, making it Beginning of page[p. 160] possible for a hyper-industrial consumer society to advance a complete ‘industrialisation of memory’.148 With capitalism’s constant technological advances, according to these thinkers, for the first time in history, a ‘systematic manufacture of machine-based speed’ was produced, which ‘began to assert itself — and insert itself — into the everyday lives of people in culture, economy and society’.149 According to this logic, humans have to negotiate two velocities at once: one reality based on the diurnal and nocturnal activities of everyday life; and another based on ‘the imperative of speed’.150 Mainstream US cinema, a little crudely, is presented by Virilio as one more instance of this usurpation of time, a ‘cinema of acceleration’ that dispenses with the dead spots of everyday life.151 But what happens when the timescape of the everyday is reinserted into film and transformed into the locus of attention? This will be the focus of the next chapter.
© by the author(s)
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.© 2025 ICI Berlin Press