A Deleuzian Analysis of Tarkovsky’s Theory of Time-Pressure, Part 1


by David George Menard Volume 7, Issue 8 / August 2003 28 minutes (6965 words)

The purpose of this essay is to offer a Deleuzian time-image analysis of Tarkovsky’s montage theory of “time-pressure,” foregrounded against the historical backdrop of Eisenstein’s montage of attractions. Several films from Tarkovsky’s later work will be examined for montage elements that support or contravene these theories.

The history of the post-Revolution USSR can be broken into three major periods: 1) Stalinist period (1927-1953) 2) post-Stalinist period (1953-1986) and 3) Gorbachev (and post) period (1986-). During the Stalin period the Soviet film industry was under the control of the communist regime.1

The Russian new wave began during the Post-Stalinist period, at about the same time as it did in the rest of Europe. It is characterized by films about national history and WWII, social dramas, and poetic films. The Russian new wave filmmakers are many, among them, Andrei Tarkovsky, Sergei Paradjanov, Alexei German, Tengiz Abuladze, Otar Iosseliani, and Nikita Mikhalkov. To speak of Soviet-Russian cinema is to discuss the achievements of the old and new school of film making, and one can easily begin by comparing the works and theories of Sergei Eisenstein and Andrei Tarkovsky.

Soviet montage developed after the 1917 Russian revolution. Sergei Eisenstein is considered, if not the father of Soviet montage, its most articulate spokesperson. Soviet montage views movement and space as the distinctive characteristics of cinema as opposed to theater. Montage defines the way in which images are cut and assembled together. The director composes separated filmed fragments into a whole and juxtaposes these fragments into an integral structure to achieve a rhythmical effect. Eisenstein considered montage as the basis of art cinema (film art). The “montage of attraction” puts objects, ideas, and symbols in collision to produce an intellectual and critical response from the viewer.2

Andrei Tarkovsky is the most celebrated filmmaker of the 1960’s, 70’s and 80’s (until his death in Paris on Dec. 29, 1986). He is the patriarch of the contemporary Soviet “poetic film.” Tarkovsky strongly opposed montage and believed that the basis of art cinema (film art) is the internal rhythm of the shot. Tarkovsky’s idea of “Sculpting in Time” proposes cinema as the representation of distinctive currents or waves of time, conveyed in the shot by its internal rhythm.

Tarkovsky believed the film image not to be a composite of different shots arranged in a structure within a specific sequence progressing in time. He reasoned that if the film image is not a composite then the dominant factor of the film must be its rhythm. Rhythm is at the core of the “poetic film.” But Tarkovsky’s idea of rhythm is not that of Eisenstein, instead he envisioned cinematic rhythm as some kind of movement within the frame, and not as a sequence of shots in time. Hence, the main characteristic of poetic film is the process of Sculpting in Time as opposed to Eisenstein montage of attractions. While Eisenstein’s process of editing is guided by intellectual and conceptual juxtaposition of images, Tarkovsky’s time sculpting involves editing techniques which allow spontaneous unification of the shot as a self-organizing structure. Instead of the interplay of concepts (Eisensteinian montage), Tarkovsky creates the film image as an expression of the matter world, or simply the world. For Eisenstein, the concept dictated the cut; but for Tarkovsky, it is time that rules, dictating the editing techniques. Therefore, time within the frame expresses something significant and truthful that goes beyond the events on the screen and those in the frame; and so, the direct perception of time is like a pointer to infinity (this approach is quite different to the montage of attraction between shots where elements in the shot juxtapose concepts, making the viewer produce some intellectual link). While the montage of attraction produces a burst of meaning, arousing the viewer with the purpose to suggest specific ideas and concepts, Tarkovskian time rhythms illustrate a way of seeing life in its essence, life’s movements. Moreover, this poetic expression of the material world may go beyond the artist’s intention and be received differently by each viewer. In the Tarkovskian School of film poetics, the filmmaker expresses his philosophy of life as opposed to creating a new perception of a social reality.

More typical than Paradjanov’s treatment as a film director was the plight of Andrei Tarkovsky [1932 – 1986], the second major figure from the postwar Soviet cinema during the sixties.3 Tarkovsky was probably influenced in his decision to study at the VGIK (Moscow film school) under Mikhail Romm by his father, the Russian poet Arseni Tarkovsky.4

In 1974 Tarkovsky produced an autobiographical film a la Fellini’s Amarcord (1973), entitled Zerkalo (Mirror), which was criticized as being labyrinth in form and parabolic in nature. In 1976, he directed an acclaimed stage version of Hamlet in Moscow. Late in the decade, Tarkovsky made Stalker (1979), an ambiguous allegory of decay shot in Estonia which was read by some as an indictment of the Soviet government’s repression of intellectual freedom. It is an eerily dismal film about a writer and a scientist who are guided by a man called the Stalker, on a journey through a mysterious wasteland referred to as the “Zone.” Their goal being to travel to a place called the “Room” where all wishes may be granted, but they fail in their quest through lack of will power. In 1982, he began shooting Nostalghia (1982) in Italy for Gaumont and RAI, with Soviet cooperation.5 In 1985, he completed his final film, the international co-production The Sacrifice (1986) in Sweden.6

Tarkovsky’s Mirror is a philosophically personal and autobiographical film dealing with memory and temporality. The concepts of time and remembrance have been the tropes of investigations by other authors; for instance, Alain Resnais’ Muriel (1963) is a personal film that explores such issues through the socially changing framework of French history (circa 1960). Tarkovsky’s Mirror, alternatively titled Mirror or A White, White Day, delves into the personal histories of its characters through a kaleidoscopic mesh of interwoven time periods. In Mirror, history becomes an enigmatic Mirror that reflects three different levels of temporality in which time expresses a sense of universal oneness. There are three distinctive periods that differentiate time from an otherwise unified portrayal of the personal history of its principal protagonist, Alexei, the narrator (voice-over), who is glimpsed only the end of the film:

1) the ‘present’ circa 1975,

2) the ‘past’ of post-WWII (mid-1940s),

3) and another ‘past’ of pre-WWII (1930s).

In Mirror Tarkovsky differentiates history into temporal categories and integrates within them the personal histories of the characters to expose the unifying aspects of time. He captures a temporal oneness through a sensibly oneiric cinematography and carefully structured mise-en-scéne, especially in its color tones, surface textures, and sound designs. Moreover, he matches stock footage from these periods with the time sensibility of his own shots, that is, he cuts documentary footage with the time-image_s emerging out from his _time-pressure editing. This type of time-thrust cutting, which he used to make his thesis film The Steamroller and the Violin (1960) is in contrast to the Soviet montage style developed by Sergei Eisenstein. This new way of conceptualizing montage involves the matching of the internal rhythms within the shots with each other, and it is this new brand of temporal linkage that dictates the cut (not the concept which dictates the cut in Eisensteinian montage). These inner rhythms are related to the flow of time, the direct perception of time that exists and emanates from the shots; and as with any dynamic continuum, the flow of time carries a temporal mass or momentum, definable by a so-called “time-pressure,” a hypothetical concept in Tarkovsky’s montage theory of Sculpting in Time.7

Tarkovsky explains that editing cannot be the dominant structural element of a film, as the protagonists of Soviet montage cinema (Kuleshov and Eisenstein) maintained in the 1920’s. The film image comes into being during shooting, and exists within the frame. As Donato Totaro explains, editing brings together shots which are already filled with time (1992, 24). The function of editing is to organize the time-image_s into a wave structure inherent to film, that is, the _time-pressure wave. Tarkovsky’s concept of time-pressure is like a meteorological time-front that propagates from shot-to-shot and throughout the film, or a cardiopulmonary time-pulse that thrust against the arterial walls of the scenes, bringing temporal oxygenation to the shots and overall meaning to the film-form.

In Tarkovsky’s view, the Eisensteinian theoretical structure of the montage of attraction, as a form of editing that brings together two concepts and generates a new, third one, cannot fully explain the nature of cinema. This view is akin to the inconsistencies that exist in Newton’s law of motions (analogous to Eisenstein’s laws of montage) in explaining why light bends around the sun or in predicting the perihelion shift of the orbit of Mercury. Even though Tarkovsky has great admiration for Eisenstein’s pioneering efforts, the traditional Soviet montage cinema can only be a subset of a more encompassing grand unified theory of film. Tarkovsky’s theory derives from the incomplete understanding of time that exists in Eisenstein’s theory (whose temporal concepts are very similar to the indirect perception and treatment of time in Newtonian physics). Tarkovsky hypothesizes that time can be directly perceived in film. He theorizes that the time-thrust (temporal force or energy) is equivalent to the cinematic material (filmic matter projected on the screen at the speed of light – analogous to E = MC2) and therefore it is inherent to every shot. Therefore, just as Albert Einstein’s theory of general relativity reduces to Newton’s gravitational law of force, as the material velocity becomes very much smaller than the speed of light; so too does Tarkovsky’s theory of time-rhythm montage (or Sculpting in Time) reduces to Eisenstein’s theory of shot-concept montage, as the division of the sensory-motor link becomes much more rational than irrational. Tarkovsky believed that film’s ultimate goal cannot be the interplay of concepts because the film-image, specifically the time-image, is tied to the concreteness of time and the temporality of matter, reaching out along mysterious paths to regions beyond infinity. For this reason, the poetics of cinema, a mixture of base, everyday material substances, is very much resistant to symbolism (in the Tarkovskian sense).

Even though Tarkovsky and Eisenstein appear to have opposite views on the idea of montage, Sculpting in Time incorporates some of Eisenstein’s montage categories (rhythmic, tonal and overtonal) into its own system of cutting. Eisenstein’s film structure can be described as an organic form with artificial content while Tarkovsky’s approach is organic in body, exhibiting the characteristics of a living organism, and thus, it is essentially natural. Tarkovskian time-rhythm montage is no more than the natural time variant of the unification of the shots that exists in the material of the film. It allows the separate scenes and shots to come together spontaneously, joining up according to their own intrinsic pattern of relationships and articulations. Retroactively, a self-generating structure forms during editing because of the inherent material temporality that is caught during shooting. The grouping of the shots creates the structure of a film, but not necessarily its rhythm. In a brilliant occasion of praxis aligning with theory, Tarkovsky writes: “The distinctive time running through the shots makes the rhythm…rhythm is not determined by the length of the edited pieces, but by the pressure of the time that runs through them (1986, 117).

The flow of time running through the shots creates the rhythm of the film. Moreover, rhythm is determined not by the length of the shots, but by the pressure of the time that runs through them. Therefore, montage can only be a feature of style because it cannot create this time-rhythm, the truly dominant element of our modern cinema.

Flowing systems permeate all aspects of life, matter, energy and time; but they manifest their myriad rhythms in a contradictory duality of nature, exhibiting particle-wave characteristics. On the microscopic level, nature behaves as a matter-wave whose motion is described by a matrix energy operator, H, operating on a particle-wave function, within the dynamics of the Hamiltonian representation where time is effectively stationary at the moment of the measurement, or within the Schrodinger representation where time evolves along with the moving matter, through the operations of a 2nd order space differential operator. A finely exquisite understanding of elementary quantum physics offers a penetrative insight in the Tarkovskian theory of time-pressure. The material duality of time [in Einstein’s general theory of relativity, time (t) is an equivalent coordinate to those of space (x, y, z), where all four quantities are expressible as a single 4-vector in space-time (x, y, z, t)]. The implication is tremendous because time behaves much like a light-wave that never stops moving, but also as a particle-wave that can be relatively arrested (an implication that allows for the direct perception of time in Tarkovsky’s cinema). It is not surprising that a cinematic physics exists which leads us to ask questions such as: “what is a time-thrust and/or time-pressure?” In response to this question, Totaro replies, simply, that it is the use of nature in cinema that guides and gauges the degrees of temporality of the audiovisual presentations and since nature exhibits an immense repertoire of all kinds of activities [water, fire, rain, mud, snow, wind, and even milk (just to name a few)], they become propagators of time-image_s in which the flow of time is perceived directly through a _time-thrust or time-pressure.8 The use of nature in film is purely organic and has a sense of circularity, akin to certain Eastern philosophies which, like Buddhism, are characterized by non-linear forms of thinking; for example parallelistic logic, where A is equal to not A, is in marked opposition to Aristotelian logic where A can never equal not A.

Time dictates the particular cutting strategies because it is imprinted in the frame. The shots that won’t edit or properly join, are pieces that record different kinds of time; implying that actual time cannot be joined with conceptual time. The temporal consistency that propagates through the shot is defined as its rhythmic intensity (or “sloppiness”). It follows from cinematic physics that editing is the assembly of the shots which results from the time-impedance matching (analogous to the impedance matching characteristics of filters used in electronic circuitry) of their inherent time-pressure_s. In short, to make a Tarkovskian film is to maintain the operative _time-pressure (or “thrust”) that unifies the impact of radically different shots.

How does time-pressure makes itself felt in a shot? To paraphrase Tarkovsky with a Deleuzian twist, time materializes when there is a feeling of something significant and truthful that goes beyond the optical and sound situations on the screen. The audiovisual events depicted on the screen are merely material indicators of something stretching out beyond the infinity of the image (in electromagnetic field theory, the light wave’s potential becomes zero only at infinity) – what Tarkovsky calls “pointers to life.” Thus, a truly real film stretches beyond the boundaries of its sound-images, creating more thoughts, ideas, than consciously put there by the filmmaker. It does so by recording on film the time-waves which flow beyond the edges of the frame; and like time’s dual nature (particle-wave), the dominant factor in Tarkovskian cinema is a dual or two-way process in which a real film lives within time only if time lives within it. A ‘real’ film is like a living organism because it grows in form and meaning after leaving the editing bench, detaching itself from authorial intent and allowing itself to be experienced and interpreted in individually personalized ways – just as those unique and precious moments in real life.

This is a radical movement in modern cinema because it liberates film from the constraints of the author who creates it, allowing the film to live in time on its own.9 Ian Christie discusses the issues of formalism and neo-formalism in the modern cinema:

Formalism, they [David Bordwell & Kristin Thompson] believe, unlike some structuralist and psychoanalytical methodologies, crucially implies an active spectator … Bordwell proposes a ‘constructivist’ theory which links perception [Tarkovsky] and cognition [Deleuze] … Both Thompson and Bordwell make use of the term ‘parametric cinema’, adapted from Burch (1973) to take their neo-formalist analyses into challenging terrain … defined as the foregrounding of an artistic motivation in a systematic, structuring fashion … it was not until the 1980s that … the long-neglected work of Mikhail Bakhtin …Bakhtin’s most influential concept is probably ‘dialogism’, which emerged particularly from his study of Dostoevsky’s novels … involves distinguishing between an author’s direct speech and that of his characters, which can approach the relationship between two sides in a dialogue … two of Bakhtin’s other contributions seem even more pertinent to cinema …Bakhtin showed how these (i.e. ‘speech genres’) interact with literary genres to define a ‘genre memory’ [Tarkovsky] which sets limits to each genre … This term (i.e. ‘chronotype’), taken from mathematics, is used by Bakhtin (1981) to refer to the specific interrelationship of time and space in different forms of narrative … Maya Turovskaya (1989) has used the concept of the chronotope to illuminate Andrei Tarkovsky’s idea of cinema as ‘imprinted time’.10

Tarkovsky strictly applies time-thrust editing theory to the minute details of his films and achieves a separation between authorial intent and spectator participation. Films such as Solaris and Mirror tap into the time-memory elements of the viewers’ personal histories, allowing each individual to develop his or her variant forms of understanding to what he or she perceives.

Tarkovsky’s theoria appears to be a complicated cinematic construct but it is not.11 What is difficult about Tarkovsky’s cinema is its praxis. It is not easy to identify the proper time-image match which is necessary to maintain the desired stability of the time-pressure level between shots. In short, it is imperative that the time-wave propagates freely from shot-to-shot, otherwise, a form of resistance develops in the “gap” and the purity of the optical and sound situations which is the basis for the direct perception of time is lost.

Gilles Deleuze discusses the significance of this “breach” in the sensory-motor linkage:

bq.… from its first appearance, something different happens in what is called modern cinema … the sensory-motor schema is no longer in operation, but at the same time it is not overtaken or overcome. It is shattered from the inside … perceptions and actions ceased to be linked together, and spaces are now neither coordinated nor filled [ i.e. “gaps”] … It is here [in the “gaps”] that the reversal [where sign and image transposed their relation] is produced: movement is no longer simply aberrant, aberration is now valid in itself and designates time as its cause …. It is no longer time that depends on movement; it is aberrant movement that depends on time.12
The stationary relation of the sensor-motor link and the indirect image of time is transformed to a delocalized relation of the pure optical and sound situations and the direct image of time.13

Deleuze writes about Tarkovsky’s text of the ‘cinematographic figure’ as follows:

…Tarkovsky says that what is essential is the way time flows in the shot, its tension [i.e. time-pressure] or rarefaction, ‘the pressure of time in the shot’. He appears to subscribe to the classical alternative, shot or montage, and to opt strongly for the shot (the ‘cinematographic figure’ only exists inside the shot). But this is only a superficial appearance, because the force or pressure of time goes outside the limits of the shot, and montage itself works and lives in time …. Tarkovsky calls his text ‘On the cinematographic figure’, because he calls figure that which expresses the ‘typical’, but expresses it in a pure singularity, something unique. This is the sign, it is the very function of the sign … It is only when the sign opens directly on to time, when time provides the signaletic material itself, that the type [cinematographic figure => ‘typical’], which has become temporal, coincides with the feature of singularity separated from its motor associations. It is here that Tarkovsky’s wish [a reference to Stalker] comes true: that ‘ the cinematographer succeeds in fixing time in its indices [in its signs] perceptible by the senses [my italics].’”14

In short, modern film is not a language operating with predefined cinematic units (unit-shot = montage-cell) and montage is not a super-unitary system that organizes sub-unit shots.15

The time-thrust can be easily overlooked because they are often unperceivable optical and sound situations, with no commensurable links to each other and no easily inferable connections to conventional referents.16 For example, Mirror is structured by the interposition of personal memories within a timeline of significant historical events and socio-cultural situations. Tarkovsky’s style of cutting tends to keep the historical temporalities separate, but now and then, he allows them to co-exist in the same diegetic space and sometimes in the same shot. In Mirror, the act of remembering alternates between two worlds, one actual and the other virtual, and sometimes, memory exist simultaneously in both worlds. The memory-scape bifurcates into actual and virtual situations that parallel each other within a temporal quandary.17

Mirror extracts images from thought-memories and surrounds them in a world of time. Material objects are reflected in a Mirror-image (time-image) as a double movement of liberation and capture, where the virtual object Mirror_s the real; as if, momentarily, the image in a _Mirror separates from its surface and crystallizes into physicality, only to reabsorb again and become mentality. In short, the time-image has an image-structure, a coalescence of the actual and the virtual. Donato Totaro explains that the physicality relates to matter as an extension of space and to the movement-image, while the mentality is tied into memory as a duration of thought and the time-image. The movement-image is a spatialized cinema, as seen in Hollywood genre films, where time is measured by movement and determined by action. The time-image is a temporalized cinema, as in the European art films, where the temporal links between shots are non-rational and incommensurable, resulting in the emergence of empty, disconnected spaces; what Deleuze calls “any-space-whatevers.”18 Totaro correlates Bergson’s views on memory19 with another form of the time-image concept:

The crystal-image, which forms the cornerstone of Deleuze’s time-image, is a shot that fuses the pastness of the recorded event with the presentness of its viewing. The crystal-image is the indivisible unity of the virtual image and the actual image … The crystal-image shapes time as a constant two-way Mirror [like in Mirror] that splits the present into two heterogeneous directions, “one of which is launched towards the future while the other falls into the past. Time consists of this split, and it is … time, that we see in the crystal [the crystal of Mirror]” (Deleuze’s Cinema 2, pp. 81).20

The significance of the crystal-image is of immense importance in Tarkovsky’s theory of time-pressure and in the time-rhythm montage of Mirror because it is the key that unlocks the door to the compositional domains of the opsigns and sonsigns which are the correlates of the time-images. Totaro writes that:

Deleuze uses the crystal-image as an aesthetic rather than purely theoretical tool by ascribing stylistic qualities to it [film style => philosophy] … Deleuze does not subscribe, as does Tarkovsky, to the notion that the long take, or time registered in the shot, is of a different value or type than time registered through montage (Eisenstein) … this is a superficial distinction because, “the force or pressure of time goes outside the limits of the shot, and montage itself works and lives in time” (Deleuze’s Cinema 2, 42).21

If one compares film with music, cinema stands out as giving time visible, real form. A piece of music can be played in different ways where musical time is a condition of certain causes and effects set out in a given order, carrying abstract and philosophical sensibilities – music records time inwardly. But film is able to record time outwardly with visible signs, recognizable to the senses – so time becomes the very foundation of cinema, as sound is in music, color in painting, character in drama. Rhythm is not the metrical sequences of the shots but the time-thrust within the frames. What is different about Tarkovsky’s editing is that it brings together time, imprinted in the segments of film. Cutting does not engender, or recreate, a new quality; but it brings out a quality already inherent in the frame that it joins. Editing has to do with temporal extensions and the degree of intensity with which these _time-thrust_s possess (_time-pressure_s). Editing represents intervals of time dealing with the diversity of life perceived. Rhythm exists in the life of the object visibly recorded in the frame while the temporal movement is conveyed by the flow of the life-process in the shot. It is through this time-rhythm that the director reveals his individuality and stylistic marks. Time-rhythms come into being spontaneously during shooting, with the filmmaker’s sensibility for nature’s rhythms in his search for these elusive _time-image_s. Tarkovsky’s editing style disturbs the passage of time by introducing time-flow interruptions (the act of cutting) which create temporal distortions. It is the distortion of time that gives it rhythmical expression.

This is the basis for Tarkovsky’s theory of Sculpting in Time. It exposes the direct figure of time by the deliberate and careful joining of shots of uneven time-pressure. The cutting has to come from the inner necessity (within the shot) and the organic process going on in the material as a whole (akin to Eisenstein’s overtonal montage). The process of joining segments of unequal time-value breaks the time-rhythm. However, if the temporal disjunctions are correlated by the _time-thrust_s (forces or pressures) within the assembled frames, then the desired rhythmic design can be achieved. As Totaro notes:

Matching shots of differing rhythms can be done without destroying this organicprocess if it grows out of an inner necessity. An example is the car journey sequence in Solaris. Through camera movement, sound and consistent forward direction the shots in this sequence share the same rhythm. The montage heightens to a frenzied single-frame fusion of overlapping highways, lights, skyscrapers and cars. This technological symphony is abruptly followed by a cut to astronaut Kelvin’s childhood dacha. The image is quiet, peaceful and serene. The time-pressure in this shot is opposite from that in the previous shots. …This is one of the few examples of Tarkovsky using expressionistic editing and deliberately matching shots of differing time-pressure_s. But the cut is theoretically justified because the stark contrast between the chaotic _time-pressure in the technological montage and the tranquil rhythm in the shot of the natural landscape reflects one of the film’s thematic conflicts of technology/nature, space/earth (1992, 24-25).
Tarkovsky’s sense of time is related to his innate perception of nature. His editing style is dictated by the rhythmic pressures in the segments of film. His authorial signature comes from his editing style and is the mark of his attitudes to the conception of cinema and philosophy of life. His art films are formed by organic processes and are living organisms with their own circulatory system (time flow) which must not be brought to stasis. Tarkovsky’s time-pressure montage represents the sensibility of an auteur, his film style, and personal philosophy. Tarkovsky’s films form a cinema of thought-images.

Read Part 2 Here.

A Deleuzian Analysis of Tarkovsky’s Theory of Time-Pressure, Part 1

David George Menard is a Polymath Physicist and Filmmaker, a Physics MSc graduate from the University of Tennessee Space Institute, Tullahoma, TN. David went to work for Martin Marietta Missiles Systems, the Electro-Optics Division, in Orlando Florida. Unfortunately, circa 1990, the Soviet Union collapsed and many scientists lost their jobs. So he began a new career in filmmaking, attending the Mel Hoppenheim School of Cinema in Montreal, graduating with a MFA in film production in 2010. After which, he moved to Los Angeles and began writing screenplays, continuing to do so while promoting “Termite Cat Productions, Ltd.”

Volume 7, Issue 8 / August 2003 Essays andrei tarkovskyfilm theorygilles deleuzehenri bergsonmontagepeople_deleuzetemporality