Shortly into the second part of Chris Marker’s epic 1977 compilation film Le Fond de l’air est rouge, one of the great acts of montage in the history of the cinema takes place. After presenting us with a prolonged sequence shifting from newsreel footage of Soviet tanks freeing Prague from Nazi occupation in 1945 to the same tanks returning to the city 24 years later, this time not to liberate the city, but to crush the political aspirations unleashed by the Prague Spring, Marker homes in on an emblematic scene from this event. Czech citizens and news crews huddle around a callow Red Army soldier as he emerges from a tank, and implore him not to acquiesce in the invasion. Invoking an innate sense of solidarity from the youth, stupefied by the attention he has suddenly been given, the Czechs attempt to pierce through the shroud of incomprehension by calling out to him in the few words of Russian they know – from within the crowd, cries of “Comrade! Brother!” can be heard.

At once, Marker enacts a cut which transcends nations, historical eras and even cinematic genres, by transporting the spectator from the streets of Prague to a close-up of the sailor Vakulinchuk, hero of Eisenstein’s Bronenosets Potyomkin (Battleship Potemkin, 1925) making an impassioned exhortation direct to camera. The content of Vakulinchuk’s utterance is quickly given, as the sequence cuts now to one of the most famous intertitles of the silent cinema, taken from Potemkin: the word “Bratya!” (“Brothers!”) in block Cyrillic lettering. Immediately upon this plea appearing on screen, Marker cuts back to the original scene: the Russian soldier, Vakulinchuk’s distant descendent, stares towards the newsreel camera trained on his face, in utter perplexity at the historical stage onto which he has just stridden. (Figures 1-4)

Figure 1


Figure 2

Figure 3

Figure 4

The resonance Marker achieves between these historical moments – between the Soviet Union as the vanguard of political and aesthetic revolution, and the same country, four decades later, as oppressive superpower shifting its weight against a smaller neighbour – occurs in a flash: each of the two shots from Potemkin inserted into the sequence last barely a second, with the montage-act functioning close to the threshold of perception. No sooner does the spectator register the effect than the micro-sequence is over, and we are returned to the original scene as if nothing had happened.

Irresistibly, Walter Benjamin’s concept of the “dialectical image” imposes itself as a conceptual framework for understanding the effect of this sequence. In several passages from Convolute N (titled: “On the Theory of Knowledge, Theory of Progress”) of The Arcades Project, the German critical theorist introduces this notion, and it is worth quoting from the first of these passages in full (from section N2a,3), as it gives a succinct overview of how Benjamin conceives of the dialectical image:

“It’s not that what is past casts its light on what is present, or what is present its light on what is past; rather, image is that wherein what has been comes together in a flash with the now to form a constellation. In other words, image is dialectics at a standstill. For while the relation of the present to the past is a purely temporal, continuous one, the relation of what-has-been to the now is dialectical: is not progression but image, suddenly emergent. – Only dialectical images are genuine images (that is. not archaic); and the place where one encounters them is language.” (1)

The dialectical image, then, is typified by a “flash-like” coming together of past and present, which forms a constellation of ephemeral, fleeting duration, whose potential meaning or affect is elusive, shifting, polysemic. The notion, while remaining within the general orbit of dialectical theory, represents a marked divergence from the tradition of Hegel and Marx, as evinced by the aporetic definition of the image as “dialectics at a standstill.” Later in the Arcades Project, Benjamin will extend the notion, and apply it to a dialectical approach to history, in the following extract:

“Where thinking comes to a standstill in a constellation saturated with tensions – there the dialectical image appears. […] It is to be found, in a word, where the tension between dialectical opposites is greatest. Hence, the object constructed in the materialist presentation of history is itself the dialectical image. The latter is identical with the historical object; it justifies its violent expulsion from the continuum of historical process.” (2)

It is difficult to date these passages precisely, in a work which preoccupied Benjamin from 1927 until his premature death in 1940. Nonetheless, the concept of the “dialectical image” critically informs Benjamin’s historical method in “Theses on the Philosophy of History.” Here he develops an idiosyncratic version of historical materialism which, in contrast to the “homogenous, empty time” of Social-Democratic, progressist historicism, views “the present as the ‘time of the now’ which is shot through with chips of Messianic time.” (3) In Messianic time, “the true picture of the past flits by. The past can be seized only as an image which flashes up at the instant when it can be recognised and is never seen again.” (4)

The Arcades Project does not only contain a theoretical exposition of the dialectical image; its entire formal structure is permeated by the notion. A gargantuan collection of brief extracts, whether by Benjamin’s own hand or quoted from other sources, is assembled into constellations of text given an elaborate alphanumeric code. With the connections between passages at once tenuous and forceful, Benjamin freely avowed that the method of his project was one of “literary montage,” (5) and, given the discussion of the work of Dziga Vertov in “The Work of Art in the Age of Mechanical Reproduction,” it is probable that the Soviet filmmaker’s montage practice was a major influence on Benjamin’s use of the method as a literary technique. (6) This would lead us to dissent with Benjamin’s judgement that language is the only place where one encounters dialectical images. In the cinema, too (or even above all), the dialectical image is to be found.

Moreso even than in Vertov, a montage practice based on the dialectical image can be found in the multifaceted œuvre of Chris Marker. The extent to which Benjamin’s concept is a conscious reference point for Marker can be debated – but it is notable that, interviewed in 1999 by Libération, in the guise of his cat Guillaume, Marker would claim of his CD-ROM Immemory that:

“It is somewhat like the project that Walter Benjamin had in The Arcades Project: to attach oneself to details, to the tiniest of things which historians and sociologists hold in disdain, and to arrive through their intermeshing at the portrait of an era, with the symbolism of the passageway, as we find it in Aragon or Cortázar – but I shouldn’t say anymore about this, or else I’ll get caught out.” (7)

Indeed, our sequence from Le Fond is a remarkable example of the diverse ways in which Marker has been able to create Benjaminian “constellations” (or “passageways”) between past and present in his films and multimedia works. Not only are resonances created between his own film and Eisenstein’s Potemkin – that is, between the revolutionary 1920s and that decade’s echoes in the 1960s and 1970s – but they also occur between different sections of Marker’s film itself. In particular, the Prague Spring episode harks back to the Le Fond’s opening sequence, a frenetically-edited tour de force of montage.

In this sequence, as an introduction to the film’s broader project of crafting a retrospective look at the decade of political struggle between 1967 and 1977, Marker spliced together contemporary 16mm footage of protests from around the world with a number of key moments from Potemkin. In spite of the major aesthetic differences between his work and that of Eisenstein, it is clear that Potemkin, in particular, occupies a privileged iconographic position for Marker.

Marker’s sequence begins with an off-screen voiceover by Simone Signoret. Admitting that she did not watch Potemkin when it was first released (being too young), Signoret nonetheless recalls, from the film, “those shots of the meat with the worms,” the “little tent where the dead man lay” and “a tall sailor with a moustache who shouted a word which was written up on the screen: ‘Brothers!’” Images from the film also appear at this point, but they invariably are at a remove from Signoret’s descriptions: we see sepia-toned images of an irate officer, the tarpaulin spread over the sailors, the firing squad’s rifles being raised and Vakulinchuk’s steely gaze among others. It is only when Signoret utters the word “Frères!” that synchronisation between sound and image kicks in, as the “Bratya!”-intertitle appears on screen, presaging its use in the later Prague Spring sequence, and forging a forceful effect of resonance between two segments separated by close to two hours of filmic time.

What do I mean by “resonance” in this context? The figure, perhaps, to have best theorised the phenomenon is the Armenian filmmaker Artavazd Pelechian, who in a text discussing his own work developed the notion of “distance montage” (or, alternatively, “contrapuntal montage”). Here, Pelechian specifically opposes his use of montage to what he termed the “classical montage” of Vertov and Eisenstein. Whereas his forebears focus attention on the “reciprocal relationship of juxtaposed scenes” in order to generate “meaning, appreciation, conclusion” in a sequence, in Pelechian’s films: “the very essence and the principal accent of the montage resided […] less in the assembly of scenes than in the possibility of disconnecting them, not in their juxtaposition but in their separation.” Given two montage elements which, due to what Pelechian calls their “visual resonance [obraznoe zvoutchanie]”, would have been brought together in Eisenstein’s or Vertov’s aesthetic systems, Pelechian seeks instead to “separate them by inserting between them a third, fifth or tenth element.” It is thus through the “interaction of numerous links” between two shots situated remotely from each other that Pelechian is capable of expressing the meaning of his sequence in a manner “far stronger and more profound than through direct collage.” The film’s expressivity, according to the Armenian, thus “becomes more intense and its informative capacity takes on colossal proportions.” (8) The parallels with Marker’s work are striking, and yet the two developed their aesthetic systems quite independently of each other – Pelechian was not “discovered” in France until 1983, (9) and Marker, in spite of his profound familiarity with Soviet cinema, makes no mention of him in any of his films or public statements. The deep affinity between their work was, however, recognised by Marker – if belatedly – by his inclusion of Pelechian’s films in a carte blanche programme held in the Cinémathèque française in 1998. (10)

But let us return to Le Fond. As its opening passage continues, Marker cuts to a collection of hands making the peace sign against the blue background of the sky – now, however, the image is in colour, and the scene is evidently much closer to Marker’s present-day. Meanwhile, Signoret’s recollections have been replaced, on the soundtrack, by a recording of Luciano Berio’s composition “Musica notturna nelle strade di Madrid.” From this point, a 3-minute sequence regularly cuts between scenes of protest or political foment from Eisenstein’s films and comparable scenes taken from contemporary documentary footage taken in Cuba, Mexico, France, Italy, Japan and elsewhere. Resonances between the two types of images abound: mourners taking part in public funeral processions wipe tears from their faces, speakers exhort the surrounding crowds, women harangue police officers, protestors clench their fists with resolve or shake them in defiance. The affinity is visually heightened by the fact that much of the contemporary 16mm footage is processed with tinted monochromatic filmstock, giving it a striking resemblance to the print from which the Eisenstein footage is taken.

The highpoint of this breathtaking sequence – which, with 112 shots (including title cards) in 3min:50 of film time, contains an average shot length of 2.05 seconds – comes when Marker introduces resonances between footage from the Tsarist repression of demonstrators on the Odessa steps and the unerringly similar acts of repression carried out in the present day: police belt unarmed protestors with batons, threaten them with rifles, push and grab at them and, most startlingly, leave them with bloodied faces or even lying dead on the street, while others flee in terror. Abruptly, however, the sequence comes to a close, and another resonance between historical footage and the present day (of the late 1970s) is established, as the spectator is transported to the Odessa steps in its present guise as a tourist attraction (one of the few such sites in the world to have gained its renown from a single film). At intermittent intervals, however, shots from Potemkin fleetingly appear when a resonance is noticed with the contemporary footage which comprises the rest of the film – a technique which is most palpably used in the Prague Spring sequence discussed above.

It is worth asking why Eisenstein’s film should have such a privileged role in this regard. Potemkin makes a re-appearance, in Marker’s work, in the 1993 essay-film Le Tombeau d’Alexandre, with a key scene returning once more to the Odessa steps and commenting on Eisenstein’s ability to dilate a 30-second movement down the steps into a seven-minute film sequence. Godard, too, makes ample use of Potemkin in his own efforts at historical montage – often using precisely the same shots as Marker. (11) It is unlikely, however, that Marker shares Godard’s view, aired in numerous interviews surrounding the release of Histoire(s) du cinéma, that the great directors in cinema history were gifted with the ability to predict future events subconsciously, through the images they produce. Instead, one feels, while Marker was working his way through the stockpile of archival footage, he saw visual resonances with Eisenstein’s work again and again, for the simple reason that the Soviet filmmaker’s talent enabled him to tap into a certain eternal iconography of political protest. In a visual equivalent to Alain Badiou’s notion of communist invariants, (12) Marker’s recourse to Eisenstein shows, above all, that the same gestures, the same movements, the same motifs, repeat themselves throughout the long history of class struggle.

The remarkably polyphonic structure of Le Fond is acknowledged by Marker in a valuable, but rare, document, his preface to the published version of the film’s “script,” put out by Éditions Maspero in 1978. Marker begins the text by noting that the two types of contemporary footage present in the film –outtakes from militant films and television footage garnered through a co-production agreement – consists of two types of “the repressed.” The first comprises “our own repressed in the form of images”: whether shots which “were left at the bottom of our canisters after each film was completed,” or entire sequences which “at a certain point disappeared from the final edit.” (13) In the case of the second type of image, meanwhile, although the footage was “perfectly utilised, edited, broadcast,” it was nonetheless repressed by the fact that, shown on network television, it was “immediately absorbed by the shifting sands on which these empires are built: the sweeping away of one event by another, the substitution of dreams by perception, and the final fall into collective immemory.” The task, then, as Marker saw it, was to create “working hypotheses,” capable of “making these two series of the repressed act upon each other.” (14)

Marker presents the pessimistic view that the defeat of the revolutionary wave of the 1960s and 1970s was sealed as early as 1967 – with the stifling of the mass energies of the early stages of the Cultural Revolution and the failure of the revolutionary left in Venezuela – and in an arresting analogy he compares the events of 1968 (in Paris, Prague, Mexico and elsewhere) to Boris Karloff’s bowling ball in Scarface, which managed to knock down pins even after the man who launched it down the alley had been gunned to death. Nonetheless, the filmmaker sees the value of the radical movements of this period in their capacity for dialogue: “And then, above all, there is this dialogue, possible in the end, between all these voices, which had only been able to encounter each other for a brief moment due to the lyrical illusion of ‘68.” (15) From the perspective of 1978, however, a “triumphalist monophony” had returned; for Marker, therefore, the task of montage was to “restore […] polyphony to history,” and to fashion an “imaginary dialogue,” capable of creating “a third voice produced by the encounter between two earlier voices, and distinct from them.” Modestly, when answering the question posed to himself as to whether this method is dialectical, Marker remarks, “I do not boast of having succeeded in making a dialectical film.” (16) And yet, with Le Fond de l’air est rouge – which may be considered as a multimedia project consisting of both film and book – he has just offered both a monumental practical example of the dialectical image, through the use of resonance-montage, and a valuable theoretical exposition of this approach.


In creating a conceptual armature for the montage-structures of Marker’s densely-textured œuvre, I have mined the work of Walter Benjamin and Artavazd Pelechian. But perhaps an alternative source would have yielded just as rich a theoretical seam: the writings of André Bazin. In by far the most well-known response to Marker’s 1958 Lettre de Sibérie (Letter from Siberia), Bazin notes the radical reversal in aesthetic hierarchy at play in Marker’s development of the essay-film. In the case of other documentary filmmakers, even politically committed ones, the image is privileged as “primary matter of the film,” in which “the orientation is given by […] the montage, with the text capable of organising the meaning thus conferred on the document.” (17) With Marker’s film, by contrast, “the primary matter is intelligence, its immediate expression is speech, and the image only intervenes in third position with regards to this verbal intelligence.” Bazin proceeds to label Marker’s “absolutely new” approach “horizontal montage,” in opposition to “traditional montage, which functions in the sense of the length of the film through the relationship between shots.” In horizontal montage, by contrast, “the image does not refer back to what precedes or what follows it, but rather it refers laterally, in some way, to what is said about it.” Referring to the renowned sequence showcasing a street scene in Irkutsk, with the same image repeated three times and accompanied by politically contrasting commentaries (with two politically opposed viewpoints mediated by a supposedly objective synthesis), Bazin feels that the “operation which we have attended is, therefore, precisely dialectical,” given that it “consists of providing the same images with three different intellectual dossiers and receiving the echo from them.” (18)

Two points present themselves in response to Bazin’s article. Firstly, it is curious that the French theorist, unwittingly perhaps, reproduces the Eisensteinian opposition between “vertical” and “horizontal” montage. But here the roles are reversed. For Eisenstein, horizontal montage consisted of the sequential ordering of images, and the montage-relationships produced by such temporal juxtapositions, while vertical montage referred to temporally simultaneous montage effects, such as the “montage within the frame” engendered by the co-presence of different planes of action, or the montage effects brought about by a contrapuntal use of sound in relation to the image. For Bazin, by contrast, it is the sequential articulation of shots which is conceived of as being “vertical,” while the relationships between the different components of an audiovisual element are seen as “horizontal.” This reversal is curious, to say the least, and – although this is pure speculation – can perhaps be attributed to the different physical relationships Eisenstein and Bazin had with the filmstrip. As a filmmaker, Eisenstein would largely experience the filmstrip in a horizontal fashion, when working with it on the editing table; by contrast, Bazin, as a film viewer and critic, would experience it vertically (albeit in a much less hands-on fashion) as it passed through the projector during a screening.

Secondly, it is noteworthy that Bazin evokes an “echo” produced by montage in Letter from Siberia. Although it is not further developed in this article (or in any of Bazin’s other writings), the use of the term strikingly presages the notion of “visual resonance” developed in Pelechian’s theory of distance-montage, as well bearing strong associations with Benjamin’s concept of the dialectical image (of which Bazin would not have been aware). Resonances occur, dialectical images take shape, one can conclude, not only within and between films, but also within and between film theories.


  1. Walter Benjamin, The Arcades Project, tr. Howard Eiland and Kevin Mcaughlin (Cambridge, MA: Harvard University Press, 2002 [1982]), p. 462.
  2. Ibid., p. 475.
  3. Benjamin. “Theses on the Philosophy of History,” tr. Harry Zohn. In: Idem. (ed. Hannah Arendt) Illuminations (New York: Schocken, 1968 [1955]), pp. 253-264, p. 263.
  4. Ibid., 255.
  5. Benjamin, Arcades Project, p. 460.
  6. See: Benjamin, “The Work of Art in the Age of Mechanical Reproduction”, tr. Harry Zohn. In: Illuminations, pp. 217-251.
  7. Interview with: Guillaume le chat (alias for Chris Marker). “Tranche d’immémoire.” Libération, January 8, 1999.
  8. Artavazd Pelechian, “Le montage à contrepoint, ou la théorie de la distance” (1971-1972), trans. from Russian into French by Barbara Balmer-Stultz, in: Trafic #2, Spring 1992, pp. 90-105, p. 97. For more on this, see: Daniel Fairfax, “Great Directors: Artavazd Pelechian.” In: Senses of Cinema #62, 2012.
  9. Notably with an article by Serge Daney in Libération. See: Serge Daney, “À la recherche d’Arthur Pelechian”, Libération, August 11, 1983. Repr. in: Idem., La Maison cinéma et le monde v. II, ed. by Patrice Rollet, Jean-Claude Biette and Christophe Manon (Paris: Éditions P.O.L., 2002), pp. 410-413.
  10. See: Jean-Michel Frodon, “L’ésprit de la chouette dans’ l’ombres des Grands Boulevards”, Le Monde, January 8, 1998.
  11. See, in particular, the films Les Enfants jouent à la Russie (1993), Histoire(s) du cinéma (1988-1998), Notre Musique (2004) and Film socialisme (2010).
  12. See: Alain Badiou, L’Hypothèse communiste (Paris: Lignes, 2009).
  13. Chris Marker, Le Fond de l’air est rouge (Paris: Maspero, 1978), p. 5.
  14. Idem.
  15. Ibid., p. 6.
  16. Ibid., pp. 6-7.
  17. André Bazin, “Lettre de Sibérie.” In: Idem. Le Cinéma français de la Libération a la nouvelle vague (Paris: Cahiers du cinéma, 1983 [1958]), pp. 179-181, p. 180.
  18. Ibid. p. 181.