Jean Narboni, Sylvie Pierre, Jacques Rivette
Translated by Tom Milne.

Circumstances: Aix-en-Provence, more specifically the Centre Dramatique du Sud-Est where, at the invitation of Antoine Bourseiller, four of us found ourselves along with fifteen films on February 7, 8 and 9. In the little time available between screenings, discussions on these films, loosely hinging on the notion of montage, took place between the audience and us. This text is derived from a consideration of those discussions. The form of the text: neither debate nor round table discussion, collection of articles or treatise by several, voices, but a "montage" of critical texts. Reading of the text: 1. Non-linear, without beginning or end, attempting to open a reading space where the blanks and deficiencies, omissions or redundancies, hopefully leave the reader free to interject his own opinions or his reservations. 2. Not circumscribed, since a network of notes challenges it, opens it out, defines it more precisely. The unsigned notes are by the person to whose contribution they relate. 3. Not concluded. Provisional: between open doors and yawning questions, an arena of probing draughts. (Further articles will attempt to offer remedies.)

JACQUES RIVETTE: What was the principle of the "journees" at Aix? To take the notion of montage as a connecting thread, a notion that today becomes central to the consideration of other matters than cinema (cf. the fact, for instance, that Sollers or Faye can cite Eisenstein on an equal footing with leading literary theorists and practitioners); and on this basis to view or re-view a certain number of films, regrouping, arranging, 'superimposing' them, and from this superimposition (as with patterns) to try to discover the common grounds and the differences.

In practice, therefore, alongside a characteristic example -- Eisenstein's The General Line [aka The Old and the New -- Ed]; and it was only for practical reasons that Griffith's Intolerance could not be screened -- a number of trail-blazing films from the last ten years were brought together: About Something Else (Chytilova), Made in USA (Godard), Mediterranee (Pollet), Machorka-Muff, Not Reconciled and The Bridegroom, The Comedienne and the Pimp (Straub), Marie pour memoire (Garrel), The Hour of the Furnaces (Solanas); facing these, following the hypothesis whereby the resurgence of montage over the last ten years began with the emergence of direct methods, two key stages in "direct cinema": Shadows (Cassavetes) and Pour la suite du monde (Perrault), Rouch being omitted only through unavoidable circumstances; and finally, to put our thesis to the test, if indeed there was such a thing as a thesis, the antithesis of two supreme achievements of the so-called "classical" cinema: Mizoguchi (The Empress Yang Kwei Fei) and Dreyer (Gertrud) -- though it might equally well have been Renoir, Ford or Rossellini.

The ambition behind this grouping was in effect to attempt, in a rather hazardous (indeed aleatory) manner, a "montage of films": to interrelate, by means of these examples, different approaches to methods of structuring film, and to see what these connections and continuities might produce.

SYLVIE PIERRE: Perhaps we started out from a slight misunderstanding in that we were trying to talk simultaneously about the problems of cinematographic montage, and about the problems posed for the cinema by a more abstract idea, "the idea of montage," which in fact derives from a sort of metaphorical extension of cinematographic montage within non-cinematographic areas. (1) So on the one hand we were examining montage as a technique of juxtaposing shots, which quite naturally led us to the consideration of films representing extreme cases of montage: whether "over"-edited (Eisenstein, Pollet) or "under"-edited (Dreyer, Mizoguchi). (2) Whereas montage as a metaphorical term, on the other hand -- in other words collage, broadly speaking -- led us into quite another area of speculation, for instance with Made in USA.

RIVETTE: I don't think there was any misunderstanding insofar as the choice of films was concerned: let's say that originally we stated or implied a sort of methodological a priori, distinguishing between all films having in common that they went through the editorial stage as a creative stage, and the rest; or to put it another way, between directors who "make" the film essentially during shooting (and in the pre-planning: for example Ford and Renoir), and those for whom this work of writing or strategy and the actual shooting is merely the accumulation of material (a working stock) which is subsequently subjected to scrutiny anew, and only acquires its order and meaning in the editing room (Rouch and Perrault as well as Godard and Eisenstein): two families which we weren't comparing in quality but opposing -- provisionally at least -- in an attempt to understand more clearly.

JEAN NARBONI: To my mind another misunderstanding might have arisen from the confusion which still persists between montage as the idea of montage and montage as effect (or effects, most often understood pejoratively). When we presented these films, the intention wasn't to focus them on the secondary sense alone -- montage asking, sovereign organizing principle of the film, manipulation in control -- but more generally on the primary meaning: montage (even where it doesn't manifest itself in obvious "effects") as essential productive work. Objection could be made that in this sense any film, insofar as it comprises a certain number of shots placed end to end and stuck one to another, is dependent on montage, and that therefore any film would qualify (or indeed any shot, since Eisenstein demonstrated so forcibly that montage could not be absent even within a shot). This was why I used the words work, productive, creative montage, montage texture, to make a clear distinction between the films that are dependent on montage; and those in which the business of arranging shots, of interrupting them at this point or that, is merely the continuation, the completion -- a nuance or improvement or two apart -- of a preconceived purpose, and contents itself with following an already determined meaning instead of making a new one emerge. Exemplary in respect to the first category, for instance, are two films as different as About Something Else and Pour la suite du monde where, without being manifest through "exterior signs," the montage plays an active role as mainspring, motor, propellant, mobile suspension between two shots, but also -- especially -- between the larger units of the discourse. (3)

With regard to the famous montage "effects," we ought to reconsider this term and its usage very carefully and precisely in order to avoid perpetuating the errors and vague notions still evoked by any "edited" film. If by "montage effects" we mean manufactured devices, extraneous ornamentation, rhetorical tricks, then the term should be applied strictly to films which use them as such, which reduce to pure formula and cliche what constituted the foundation stone and not just the trimmings of the great Soviet silents (Eisenstein, Vertov, Dovzhenko), who are unquestionably beyond reproach in this matter in that they considered montage as a dynamic creative process. By including all films where montage plays a primordial role in the same debatable category, one risks falling into the same error which led to the long-held view that a poetic discourse was merely a prose discourse with something extra -- the poetic flavor or effect, in fact. But if one examines a film like Gertrud, apparently remote from the preoccupations of montage, one realizes that montage is very much present, though as a screening effect, a mask, and that it can intervene in a film as a creative process equally well through its effacement as through its attested credentials (a process which has nothing to do with the "transparent montage" of the American cinema, for example). Jean-Luc Godard's comparison of montage to a heartbeat may be profitably recalled to pursue the analogy by saying that, just as the cardiac function alternates diastole and systole, silences -- large and small -- and sounds, so montage is as effective in its pervasions as in its voids, in its absences as in its traces.

RIVETTE: This is why it was essential to see Chytilova's film again right away: a film in which the role played by montage-manipulation is quite obvious, where one sees quite clearly how both detail (plastic and dynamic: the "trait") and the effect of each splice have been systematically rethought on the editola; but above all a film where the work of montage at this primary level (the level of micro-structures) has manifest repercussions on the "thought" of the film as a whole (what musicians call "the grand form") and vice versa: each incessantly gearing itself on the other. So the film functions as the irregular alternation, the "matching" of two autonomous films, or of two hypergeneric sequences, each one being governed at every level by its own formal laws, not only in so far as methods of "mise en scene," camera attitudes and directing the actors are concerned, but in its own internal montage.

This said, what makes About Something Else particularly interesting is to see how this principle in fact works not purely and simply as the alternation of two parallel actions, as merely the sum of the two, but as a multiplication of each "level" by the other: and this without any interference or reference from one to the other, but on the contrary through the independence that is affirmed at every moment, re-established, reconstituted, recreated by each of them; it is through an incessant process of rejection, much more than through "connections," that the micro-formal web organizes itself: the act whereby the montage effectively becomes a productive process is sustained here by a rigorous system of deception. (4)

At the same time, the interlacing of the plots is not so much what arouses the interest (the participation) as what blocks it, frustrates it (deprives it of its dividend by threatening its capital) by its displacement of the referent -- of the relation: of "the real" -- to their manipulation. Hence the full scope that subsequently arises for space-time distortions, without any possibility of verification from the narrative: one can equally well feel that Chytilova makes too much of these or too little. A fictitious (fictional) space-time, strictly non-psychological (nothing to do with the imaginary according to Robbe-Grillet), the continuum of the film is opening, like space (on the stage) and time (on magnetic tape) for Merce Cunningham, but which is not pre-existent here, which therefore doesn't have to be filled, which is nothing other than a void, and like the imprint of a fossil, is left gaping by the decay of the old narrative and representational space-time.

PIERRE: At the outset, at any rate, we envisaged the films within three characteristic types of situation with regard to montage:
1. Films dependent, as Jean Narboni was saying, on "montage texture": films based on montage as the instrument of a dialectic and of a discourse (Eisenstein, Solanas).

2. Films which do not seem to establish themselves in relation to montage as creative work, in which montage is not present as a dominant effect, but in which, as we have seen, the apparent absence of montage at the creative stage may conceal various montage manoeuvres: whether the employment, to maximum effect, of a small number of liaisons between lengthy shots, or whether the displacement of the work of montage by means of other hinges in the cinematic combinative than those of montage properly speaking (through the decoupage -- e.g. Straub -- and through articulation within the shot itself -- e.g. Mizoguchi or Renoir).

3. Films based, as in 1, on a creative work of montage, but which use it less for its power to carry meanings than, on the contrary, for its power to block them. The montage serving, in other words, a preoccupation with obscuring or even banishing the meaning (Pollet). In this last category, the montage principles of underground or "undergroundish" cinema should of course be considered. Here a veritable passion for montage seems to derive less from a concern (as with Pollet) to give the film a poetic structure, than from a terrorist desire for atom using, for exploding the very notion of an "oeuvre". Montage, rapid by preference, thus being used as a means (among others) to a "systematic derangement" of the discourse. (5) & (6)

RIVETTE: Perhaps we should now abandon this area of a priori classifications for the moment, and make the leap into trying to see what these classifications "mean," what they correspond to in the activity of the films themselves. One very soon realizes, in fact, that as soon as one wants to make a rather closer analysis of the "work" (7) of one of these films (work of the film-maker on the film, operation of the film on the "reader"), one has to begin by carefully examining the categories to which it is usually subordinated.

In the case of Made in USA, for instance, the now generally accepted notion of "collage" has first to be challenged: not to deny it, but in order to try to understand better how the collage worked in this case and what particular form of collage Godard's method derived from. For what distinguishes his films from those of Chytilova, Eisenstein or Pollet, is that with him one feels there was (or used to be) an earlier state of the film, an inference the others do not permit. In Made in USA Godard leaves the impression of an earlier film, rejected, contested, defaced, torn to shreds: destroyed as such, but still "subjacent." The film only functions in relation to simultaneous referents, more or less tacit but proliferating, encroaching on each other so that they themselves ravel up and weave the entire filmic texture, since ultimately one can feel that there is nothing, no phrase, shot or movement, that is not a more or less "pure" citation or referent: the important thing being, during the course of the film, not to try to identify all these referents, which would be both impossible and pointless, but to realize (to see within the perspective of the idea) that everything is referential; though the referents are set with traps and dissembled, deconsecrated, by an operation that is literally "terrorist".

The initial impulse of the film, what one can probably think of as the point of departure for Godard's activity, is in fact a montage idea: what happens if one "edits" together, if one combines some lousy serie noire novel with the Ben Barka affair: not of course the "reality" of the affair, which I don't know, which escapes me, but as I might have read about it in the papers, as I might reconstruct it, imagine it, from a collage of newspaper cuttings; hence, a montage of two "texts" (but also, shredding of the pretexts). A reading of the film proper, which offers itself as "completed," must in a sense retrace this movement by de-montage and, through a decipherment of both the tattered remains of the thriller plot and the echoes of political coordinates (a task itself embroiled, obstructed, frankly presented as unfeasible), finally attain -- later, and likely after re-viewing -- a level where the film becomes immediately legible as it unfolds on the screen.

Almost all of Godard's films function, as a matter of fact, through the embroilment of sub-texts. In Le Mepris. for instance, The Odyssey, Fritz Lang, Moravia's novel, Cinecitta... In Mediterranee, on the other hand, Pollet makes use of the fascination exerted by a more or less comparable ideological background: bullfighting, ruined Greek temples, Egyptian statues, the sea. But, he wants to use each of these elements as a word closed in on itself, loaded with the full charge of its potential meaning, (8) whereas with Godard now, most of the time, he baulks at any clear and distinct expression of the "word;" and increasingly, moreover, the reference for each citation is, if not false, at least falsified: whereas in the early films quotations still played their traditional role, presented openly with an indication of the source and its traditional connotations, nowadays the fragmentation of his referents constitutes the texture and the very matter of the film, and in a certain sense, its end.

PIERRE: It may be no accident that there are so many torn posters in Made in USA: defacements which epitomize the complementary action of the collage. For in order for collage to exist, each element must first have been torn from its context, and this preliminary operation involves a violence at least equal to that subsequently involved in the shock of producing the new combination. With Godard, what is all the more violent about this excision of the elements is their loss of identity (the fact that one cannot recognize the quotations, but also that everything becomes quotation, even what isn't) in favor of a sort of general super-identity: an overall hyper-Godardization. In Pollet's case, the relationship of the parts to the whole is of the same order. There is indeed loss of identity (of the proper name) among the parts in favor of the whole (in other words, the idea of Mediterranean as wholly mythical), but in point of fact, inasmuch as it refers back to such a prestigious whole, each element becomes fascinating again for its own sake, acquiring more weight than it has itself. It is then up to the commentary to take over the work of laceration, both of the parts (by disrupting the fascination exerted by each image) and of the whole (the commentary's task of demystification with respect to the Mediterranean myth). This commentary method obviously cannot result in a pulverisation of the elements as radical as Godard's. Pollet's commentary-images montage remains edifying, in the sense that the film aims to be a poem, an edifice.

NARBONI: Mediterranee is an exemplary case. Among the films we selected, in fact, it established precisely the question around which our choice was organized, since it is an interrogation of montage, a question endlessly put and put again to montage: when, how, why pass from one thing to another? A sentence in Sollers' commentary defines and underlines this interrogation: "And if at the same time somewhere in some unimaginable somewhere someone calmly began to replace you?" (to examine in their turn the "somewhere," the "unimaginable," the "calmly," could lead us too far, though still to the point of the film and the question it asks of montage). Similarly, "how to end?", "how to begin?", all those apparently self-evident questions that are never asked in films (but are beginning to be: cf. for instance, Le Gai savoir or Marie pour memoire), are also omnipresent throughout the film, and not merely at the beginning and end. It is the course of the film that poses them. (9) For those who see in Mediterranee only a series of shots stirred up more or less felicitously, accompanied by a "literary" commentary by Philippe Sollers, let us recall that similar preoccupations are the subject of a novel, Drame by Philippe Sollers in fact, concerning which Roland Barthes writes: "The narration is in fact merely the free figure in this question: what is a story? At what level of myself, of the world, shall I decide that something is happening to me? The earliest poets, the authors of those very old epic ballads preceding The Iliad, exorcised the terrifying arbitrariness of narrative (why begin here rather than there) by a preamble whose ritual meaning was this: the story is infinite, it began long ago (did it ever begin?): I take it up at this point, where I start."

How does the montage work in Mediterranee, what is its role, its function, its mechanism? Precisely, it seems to me, in the area of effacing the meanings and connotations with which the content of the shot is charged even before the film starts. Selecting a limited number of shots revolving around the Mediterranean, Pollet edits and organizes them, introducing and rearranging them in a process designed gradually to make them equivalent in value, to equalize them in their symbolic importance. In my opinion it isn't true to say that there is no text pre-existing Mediterranee, that the film establishes a first text. The graphic inscription (ecriture) of the film is of course creative, but against and in relation to another text, which is, extra-cinematographic -- yes -- but cultural and ideological. The film acquires meaning (of being, in fact, Mediterranee and nothing else) only by effacing all previous meanings. It is articulated around two kinds of shot, some of them very marked, heavily charged with cultural symbolisms and connotations (pyramid, temple, or places of modern myth such as factory and hospital), others more neutral; and their intermixture, their distribution and redistribution, their alternation and recurrence, initiate -- purely through the impetus of the film's course, its movement, through the analogy of situation that comes into play (all these shots are of the film) -- a process of levelling, of equalization. The antithetical coupling of ancient and modern, marked and unmarked, gives way to a locus where unequal valuations, hierarchical degrees and temporal differences no longer hold sway. This seems to me to be the reason why Pollet has tried to turn each of his shots into the equivalent of a word, or at least -- since such a thing is impossible -- the nearest thing to a word. We know, in fact, as Christian Metz has clearly demonstrated in five points, that a film shot, no matter how parsimonious its information, can be the equivalent only of at least a sentence, never of a word:

1. Shots are infinite, the words in a language finite in number.

2. Shots are inventions by the film-maker, words pre-exist in a dictionary.

3. The shot yields an indefinite amount of information.

4. The shot is an actualized unit of the discourse; the word, a unit in the dictionary, is purely virtual: an image of a bench signifying in effect "here is a bench" and not simply "bench".

5. A shot assumes its meaning only to a small degree through paradigmatic opposition to the other shots that might have appeared at the same point in the chain (since these are indefinite in number), whereas a word always forms part of one or several more or less organized semantic fields.

Pollet, one realizes however, has intuitively tried to reduce these distinctions by taking all these points into account:
1. he has limited the number of his shots, and played on their recurrence rather than their variety;

2. he has "invented" or "staged" as little as possible in them;

3. in attempting to reduce them each time to a single unit of content, he made a maximum reduction in the information they can give;

5. he has tried to establish a rich paradigmatic of the film, first by playing on the title, Mediterranee, as a "reserve" furnishing a limited number of shots, so that each of them is in effect buoyed by the ensemble of other shots that the theme of Mediterranee might have supplied (even if we do not represent them, we think them as the other, all-embracing and lacking from each shot); but also through the ordering of the shots themselves, whose periodical recurrence (even when filmed from a different angle) denotes each time that they have been chosen -- levied -- from a relatively restricted field;

4. and finally, by effecting a perversion (the commentary is crucially important here) of the actualization of the images and of their quality of assertiveness, so that the "here is a bench," for example, apprehended in a text (imaged and voiced) where distinctions between dream/sleep, imaginary/real and so forth all subside, is eventually called into question in its turn. The pivot, the axis constituted by the commentary, is the basic element around which the vacillation in the spectacle takes place, the inversion of spectator/performer, seer/seen ("And suppose one were being watched?", "A spectacle, however, which one knows very well will not come from without"). We are a long way here from the accusations of "poeticism" the film incurred.

PIERRE: This reflection of Pollet's on montage, which has been contrasted with Eisenstein's cinema of montage, in fact establishes itself with Eisenstein on the horizon, with a reflection on montage according to Eisenstein; the implication being that montage is the only way to create non-reactionary cinema as opposed to the cinema of beguilement, of representation. But here we must be extremely careful and very precise in the terms used, for as soon as political implications come into it one is all too inclined to be taken in by vague metaphors. Undoubtedly one can consider that a cinema which conceals its shot-to-shot liaisons from you (the "unobtrusive montage" discussed earlier), or which gives preference to long takes (the "attenuated montage" also discussed), maneuvers the spectator in a way one might describe as reactionary because what is then involved is either illusionism (concealment of cinematographic discontinuity) or beguilement by means of the shot. In each case it is the impossibility of escaping what is on the screen. So, where one can by contrast qualify Eisensteinian montage as "progressist" is, paradoxically, in its most dictatorial aspect: the passages from one shot to another deprive the spectator of any possibility of escaping from thought, from the need to maintain a reflective distance in relation to the shot. Hence there is no way one can give oneself up to the representation, and it is this impossibility that Pollet took up in his turn; though rejecting the dictatorship by discourse and meaning.

NARBONI: Eisenstein was of course trying to convey a meaning by and in his films which wasn't simply the meaning of the film as meaning, as self-designation, but the meaning of Communism itself. What places him incomparably higher than the other propagandist film-makers is that he himself set out in quest of this meaning -- which, as a Marxist film-maker, he controlled -- dismembering and reconstituting it, sweeping the spectator along with him, and thus verifying Marx's words (which he quotes in Notes of a Film Director) on the necessity for the investigation of truth itself to be true, on the means as part of the truth just as much as the result, on investigation as deployed and dispersed truth reuniting in the result.

Inasmuch as Pollet declares war on the meanings which weigh down the shots in Mediterranee with their whole cultural weight, he tends to adumbrate nothing but the film as meaning, to say nothing in the film but the work of the film.

RIVETTE: There is also the fact that the idea of meaning is "progressist" in the context in which Eisenstein worked, whereas it functions in a reactionary way -- as "truth" -- in Pollet's context (which is also ours): Eisenstein, and to a certain extent Solanas, (10) produce a film in a milieu where meaning is still relatively innocent (and they take this "relative" strictly into account), whereas in our case, whether we like it or not, it is invariably taken up by the commercial market, accessory to the ideology of exchange. (11)

And one mustn't forget how Eisenstein has very consciously implemented his own "text" through the perversion and transposition of an earlier text. Griffith was the first to draw the inferences from his historical situation, he made the first great synthesis (Birth of a Nation) of the implicit and random "discoveries" of his predecessors; but his masterstroke remains that, having just completed The Mother and the Law, a film reflecting an unconsciously reactionary liberal ideology, he immediately adopted it as the mother cell and motor element of his next film, Intolerance: the very fact of interlacing four "stories" into a single flow, of imposing the same law on four periods, of gradually substituting the single course of the film for the succession of stories (a "gesture" of revelation/annulment that absolutely dominates the whole final reel), literally turns the meanings of the Ur-film upside down. It is this "intuition" of Griffith's that Eisenstein, in full awareness, chose to adopt; perfecting in the light of Marxism what Griffith had only been able to portend within a bourgeois ideology, he undertook precisely the same operation with respect to this intuition as Marx did in relation to Hegel, and through a systematic refraction and inversion of its data, gave the post-Dickensian guilty liberal conscience its full meaning as a class struggle.

NARBONI: When Dreyer borrowed the thematic and construction of Intolerance to make Leaves from Satan's Book, he reconstituted a film with four different and quite distinct stories, chronologically told, failing to recognize the possibilities of reactivation, contamination and subversive interaction that the intermingling of the stories offered to Griffith. The latter is a typical example in film history of someone capable of producing a form or a concept without being able to formulate the theory of that concept correctly, and this is because the historical present in which he existed, the cultural and ideological text he inherited from his period, furnished him with neither the means nor the terrain, or even the need. (12)

RIVETTE: Yet as everyone knows, this "theory" of montage -- though its practice persists after a fashion here and there -- whether American (but Griffith remained isolated; Stroheim and Vidor were already playing the "stage" card -- sound cinema, in other words) or Soviet (Eisenstein, Vertov, Dovzhenko), was virtually wiped out by the arrival of sound, despite the celebrated Manifesto of 1928. Yet on a basis of speech (Resnais, Godard) (13) or of "direct" (Rouch, Leacock, Perrault and many others), the "resumption" of Griffith/Eisenstein has been gradually taking place over the last ten years: diffuse, often confused or barely conscious, but representing a collective will to reactivate the idea of montage on the basis of -- and in terms of -- the knowledge acquired over the thirty intervening years.

Implicitly with Griffith, then explicitly with Eisenstein and with all film-makers who endeavour to be even moderately lucid as to the meanings of their work, to think the montage is to think the criticism of a pre-existing text: of a "datum" which is itself -- and this is what the process of the textual operations reveals -- in fact only a fabrication. Hence this working hypothesis: if all coherent thought concerning montage is de facto critical thought, doesn't any form of rejection or disregard of montage imply a theological mentality, in other words acceptance of the world as it is, and if not resignation, at least passive contemplation of the being there purely as presence, involving neither History nor mediation, with all the concepts of permanence and fate bound up with this ideology?

Of course to say that montage and critical thought go together may simply seem tautological; but what must be stressed is that it is the material work, the concrete manipulation of montage (as soon as this goes beyond the level of continuity and ellipsis, the narrative level of "stylistic tricks") that "generates" this work of critical thought, and this at all levels of the film, including some the film-maker probably hasn't considered: any questioning of the superstructures reverberates a shock on the level of the infrastructures. Another consequence: this critical movement is not limited to the results of its functioning in the film, for the film preserves it intact through the course of its development: not as an imprint (a fixed "montage effect") but as a dynamic (montage as act) affecting the spectator as such; so it becomes impossible for him to abandon himself comfortably to the telling of a story, to the representation of a fable or pseudo-reality: he must, if he wants to read the film, assume responsibility in his turn for this critical work; if he wants to see the film, he must fulfill this responsibility.

NARBONI: But the practice of montage as absolute manipulation, as an all-powerful technique of all-purpose arrangement, has long been -- and continues to be -- held as authoritarian, manipulative of the spectator on whom it supposedly imposes a series of univocal and unquestionable meanings. Broadly speaking, this was the attitude that lay behind the theories of Andre Bazin, who was for instance more responsive to the cinema of deep focus or the sequence-shot as being to his way of thinking more respectful both of the freedom of the spectator -- whose eye and understanding are thereby not subjected to a strictly programmed course -- and of the "ambiguity of reality". We realize today that when he wrote "in analyzing reality, montage assumed, through its very nature, a unity of meaning in the dramatic event," Bazin was right in so far as the work of someone like Pudovkin is concerned -- where the fragmentation of scenes, the breaking down into shots, never had any purpose other than carrying analysis to its extremes, dislocating a situation in order to dramatize or magnify it -- but not Eisenstein, for whom it was a question each time of "involving the spectator in the course of a process productive of meaning". The integration of montage with visual space which Bazin recognized as the mark of modern cinema can be found in many Eisenstein scenes and many of his writings, just as Eisenstein's formula, "the number of intervals determining the tensional pressure," could perfectly well apply to the examples on which Bazin based his analyses (Wyler's films, the kitchen sequence in The Magnificent Ambersons, which were constructed on the principle of potential voltage difference and of the slow dramatic charge in the shots). The freedom allowed to the spectator in these films was never more than whatever freedom -- guided, oriented between certain poles and strong-points perfectly disposed at intervals within the shot -- the film-maker chose to grant him, before finally imposing a predetermined meaning which, because of his delay in conjuring it, might seem to have been discovered by the spectator himself. Here one finds the most extreme contradiction in Bazin's analyses, preoccupied on the one hand by a belief in the ambiguity of reality, and on the other by the conviction that an international language exists, a natural and hidden meaning to things which the cinema does not have to produce, whose advent it need only -- by virtue of its own perception and persistence -- apprehend.

RIVETTE: Historically, in fact, this notion of cinema as transparent, which can be resumed in the Renoir/Rossellini/Bazin trilogy, was itself established in reaction to a generalized "perversion" (perversion in the ordinary sense, bourgeois perversion) of Eisensteinian practice; for what was Pudovkin doing if not simply adopting the husk of Eisenstein's theoretical principles and placing them at the service of storytelling, in tow to narrative: the montage effect is no longer "utilized" except to lend greater effectiveness to a narrative subordinated to the development of character. By way of Pudovkin, this compromise and this caricature of the "art of montage" was taken over by whole areas of the commercial cinema. (One may note how at the same time and in the same way -- with the same finality orienting the same process -- Pabst was instrumental in effecting the liquidation of expressionism in favor of the aesthetic of "mise en scene" as a formal bluff which even today still governs the entire European and Hollywood cinema: Clement, Preminger, Chukrai, Rosi. This technique of manipulating "reality," where the director is the more or less invisible master, quickly ceased to be the art of montage to become the art of decoupage (and concomitantly, of "framing" and the "direction of actors".) It was in fact against dictatorship in this area that Renoir or Rossellini took a stance and not against montage, which with them is more of a censored area, a "blank": the fact that the film-maker no longer has any need to go to his cutting-room, no longer feels this need, leads them in practice -- and unconsciously, it would seem -- to reinvest a part of this montage thought at the construction level, and more particularly at the stage of actual shooting. (Cf. the role of the sequence-shot or the mobility of the camera with these film-makers or Welles, Hitchcock and Mizoguchi, in contrast to the more generalized analytic technique, and as a structuring of levels and formal conflicts).

So one might, very schematically, distinguish four moments: the invention of montage (Griffith, Eisenstein), its deviation (Pudovkin-Hollywood: elaboration of the techniques of propaganda cinema), the rejection of propaganda (a rejection loosely or closely allied to long takes, direct sound, amateur or auxiliary actors, non-linear narrative, heterogeneity of genres, elements or techniques, etc), and finally, what we have been observing over the last ten years, in other words the attempt to "salvage," to re-inject into contemporary methods the spirit and the theory of the first period, though without rejecting the contribution made by the third, but rather trying to cultivate one through the other, to dialectise them and, in a sense, to edit them.

NARBONI: Eisenstein, Pudovkin: today we must no longer think of the opposition between them in the generally agreed terms, categories and relationships -- intellectual montage/lyrical montage, cinema-cry/cinema-song, dominant creativity/dominant theory -- but according to their particular conceptions of the dynamics of cinema as revealed in their films and clarified in their writings. Here a decisive text must be quoted: "The basic element of Soviet cinema, its specific problem, is montage. Montage is neither a means of showing or narrating, fragment by fragment as a mason stacks up bricks (Kuleshov), nor a method for developing an idea through a succession of shots (Pudovkin's lyrical principle). The idea must result from the clash between two independent elements". From this we can see very clearly what differentiates Eisenstein's ecriture -- successive transformative effects whose motor elements are linked by dynamic signs of correlation and integration, where the operations actualized are multiplicatory and productive, where the collision between two elements engenders, through a crucial leap, a new concept -- from that of Pudovkin, a chain of shots each in turn carrying a single idea in a simple process of summation. What therefore distinguishes a multidimensional space and time, structured according to the principles of a complex polyphony, a signifying purpose, a volume in constant expansion, a scenography, from a spuriously dialectic linear time. I shall borrow a question and answer from L. Althusser: "How can a dialectic be late? Only on condition that it is the other name for a consciousness"... "there is -- in the strict sense -- no dialectic of consciousness opening, by virtue of its own contradictions, on to reality itself... For the consciousness attains reality not through its internal development but through a radical discovery of the other than self." It is this other of the consciousness that Pudovkin never attained. A film like Mother, for example, centred on a central character, a consciousness embodying within itself all the circumstances of the drama, quite unjustifiably assumes the mask of a dialectical and Marxist film inasmuch as Pudovkin's cinema was subject to a simple narrative logic which prevented it from bringing multiple, discontinuous temporalities -- merely time governed by a uniform successiveness -- into play.

NARBONI/RIVETTE: All of which leads us to re-examine this theme of the "awakening consciousness" and to expose its complicity with the method whereby Pudovkin "progresses" in his work only by following the thread of an idea which runs through the film like a watermark, and which is never produced by the shots, merely transmitted by them. If we compare Mother and The General Line, it can be seen that the former tells the story (is the narrative) of a character whose view of the world is gradually modified by accumulations from the various phases and circumstances of the plot (a story such as Ford, for example, could tell -- better -- in The Grapes of Wrath; but in counter-verification, Brecht's version of the same Gorki text); whereas the latter makes us witness, and collaborate in, a metamorphosis through a series of mutations of the "mediator-protagonist" who punctuates the course of the film -- and who is no longer a character but a node of forces and acts: actor (acted on/acting), and functions in the organization of the sequences like a word being transformed and exhausting all its possibilities one after the other: a consciousness no longer central, which never at any moment reflects or dominates the situation in its entirety, but is presented each time as an effect of the dynamic of the film. No "scene" shows or demonstrates a particular stage (conscious and considered) in the peasant woman's "long and hard road"; it is praxis alone that modifies her state; it is because the tractor breaks down that she makes (that she undergoes) the decisive qualitative leap: she is then at a stage Z corresponding to the final point in her evolution as a peasant (by jumps from the "alienated peasant" stage to that of "enlightened peasant"), the tractor stops, the mechanic rips up her skirt, stripping her of the rags of her present condition, to use the strips as rags (which thus have a part in cleaning the engine), and there is here a sudden, unexpected leap: she is a tractor-driver. (And the whole end of the film is simply montage of herself with herself: her successive aspects matching with the image of her "final" (within the term of the film) transformation. The character, far from subjecting the logic of the narrative to the laws of its thought processes, is produced by the transformational mechanism of the sequences. (14)

NARBONI: It is inconceivable that this discontinuity in the evolution shown in the peasant woman was a secondary discontinuity, achieved as an afterthought by eliminating intermediary stages and linking passages which had been filmed, that it was intended as something in the order of an ellipse or stylistic effect. It could happen with Eisenstein that during shooting, with a view to montage and with an idea in mind, he accumulated considerably more filmic material than he intended to retain; he frequently left possibilities open for unforeseen articulations, new concatenations, valencies to be saturated; while editing, he might breach and leave gaps in a continuity previously filmed, retaining only certain stages of a movement, moments of a trajectory, highlights of a situation: but it is inconceivable, in this particular case of the peasant woman in The Old and the New (The General Line), that he could have filmed it with the genesis of her evolution faithfully respected in its continuity. His strict application of the Marxist theories of the leap, of the sudden break as a revolutionary moment of total renewal, undoubtedly prohibited him from doing so.

RIVETTE: A detour, whereby we might perhaps come back to the problematic of the relationship between "direct" and montage; for a film like Pour la suite du monde shows very clearly how Perrault (like Rouch) was very soon able to go beyond the stage of montage as simply selecting and ordering material by definition overabundant in relation to the "projected" film, and how the film, over and above its value as a document, acquires a poetic quality only in so far as this material is reworked throughout in very precise formal patterns, while at the same time itself suggesting these patterns and informing them dialectically; this both on the level of shot-to-shot relationships, and in the structuring of the film in movements (musical) and chapters (fictional or thematic). Which is even more explicit in Le Regne du jour, just as the creative intervention of montage is more flagrant in La Chasse au lion a l'arc than in Jaguar or Moi un Noir: the latter are closer to the chronicle form, the former to the epic form.

Another point, arising from the preceding, another similarity: just as montage looms large on the horizon even at the pre-shooting stage for Eisenstein (all the more so during the shooting, if only in the sometimes systematic use of multiple cameras: Noel Burch's article, "Fonctions de lÕalea," Cahiers du Cinema, No.194), so the direct film-maker accumulates matter for the montage, with a view to the re-examination of this raw material and its destruction as such. This attitude plays the same motor role with Cassavetes, even though it is within a dramaturgical perspective (but a dramaturgy exposed, radically undermined by the use of such material: deflected and turned inside out) that it performs its task of intercepting the "text" (the pre-text at this stage, in its first state of "eruption," closely scrutinized by the meaning). (15)

Contrariwise, in the case of Not Reconciled, one can see that the montage is detailed with absolute precision, how Straub has tightened or loosened each liaison, played on variations of tempi and so on -- in other words, materialized the principle of the film on the editola -- but also how only what was strictly necessary in view of the "anticipated" film was shot, how the film therefore pre-existed its matter from the moment of its ecriture; but at the same time one must note how this work of condensation, choice and re-ordering was in fact effected on the basis of an extensive source material (i.e. Boll's text, Billiards at Half-Past Nine, which here undergoes an operation of reduction, dislocation and conversion which no longer has anything to do with what is normally called "adaptation"): here, therefore, the preparatory work of ecriture functions as montage... (16)

Moreover, Straub imposes on the spectator (the virgin spectator viewing the film for the first time, at any rate, but also in part at subsequent viewings) an obscurity in the language, which seems wilfully indirect, apparently unaware of him as the addressee (even if he nevertheless, though tacitly, fulfils his task), and which prevents him from direct attainment of the "knowledge" it seemed to be entrusted with bringing him (17) ; the film functions before him as a dream, one might say, as the product, of an unconscious (but whose unconscious? Does it belong to the literary text? To fifty years of German history? The Straubs? The "characters" in the film?), whose structure comprises only multiple re-crossings and literal echoes, the ultimate play on words and/or images, all the informational elements also being annexed to the puzzle, though dislocated, secreted, shuffled: for instance the central monologue by the mother (who is, not by chance, at the point where all the components of the fasces (18) converge and diverge), discourse of a space-time where all times and all spaces are collided and compounded (resorbed by a process of montage/mixing).

Now, it was a very similar problematic that faced us when we re-viewed Gertrud a few hours later: if Dreyer's film, more "logical," in any case more chronological, doesn't function formally as a dream, it nevertheless also prescribes (19) an "oneiric" vocabulary: at once the telling of a dream and a session of analysis (an analysis in which the roles are unceasingly changing; subjected to the flow, the regular tide of the long takes, the mesmeric passes of the incessant camera movements, the even monotone of the voices, the steadiness of the eyes -- always turned aside, often parallel, towards us: a little above us -- the strained immobility of the bodies, huddled in armchairs, on sofas behind which the other silently stands, fixed in ritual attitudes which make them no more than corridors for speech to pass through, gliding through a semi-obscurity arbitrarily punctuated with luminous zones into which the somnambulists emerge of their own accord...). (20) So, two films which impose, by converging routes, the same analogy between their functioning (their operation) and that of "all" that is implied by the word unconscious; but at the same time, two films where the basic work seems to have taken place at the level of the intention and the ecriture (with Straub, pulverization of the original text; with Dreyer, condensation and "concentration" of this text (21) ; but films, finally, where the moment of the montage "acts" as the fulfillment of this work, but also as an intervention by the arbitrary. Now, this "enigmatic" function of montage, constant with Dreyer, always operates in his work through the "imposition" of gaps (marks of censorship?): cf. how the beginnings and ends of each shot in Master of the House are systematically interrupted, chopped, cut off, in the movement (invariably lacunary in part), each articulation "false" by a few frames; cf. even more so, the whole of La Passion de Jeanne d'Arc and how, from Vampyr to Ordet, Dreyer arrests and cuts off almost all his camera movements en route; cf. finally, in Gertrud, the three or four cut-ellipses at the junctures of two long takes, tranquilly intervening within the supposed continuity of the scene: tantalizing cuts, deliberately disturbing, which mean that the spectator is made to wonder where Gertrud "went"; well, she went in the splice. And perhaps it is through this deliberate desire to introduce, at the montage stage (instead of limiting himself to having it recopy the pre-shooting text, or like Bresson limiting it to a role that is above all "musical"), into the ecriture, no matter how precise and closely controlled it may have been in the earlier stages, these cuts, these ruptures, these leaps: this irrational -- that the "passage" of the unconscious, trapped by the literal game, is effected.

NARBONI: The reference to music just made in connection with Bresson may also be applied to all of Straub's films, which are so rife with preoccupations tending in this direction, so essentially a search for possible homologies. One might cite, more or less at random, the distribution and proportionment of tempi, the alternation of zones of tension and release, of dense nuclei and silent expanses, the complex and variable interplay of autonomy and interdependence among the "cells", the composition in large blocks or pinpoint elements, the combination of solidly built structures with other "freer" ones, and finally, the application of the principle never belied by Stravinsky, the rejection of expressivity. Let us recall the terms, equally valid for all of Straub's films, in which Stockhausen wrote about Machorka-Muff: "What interested me above all in your film was the composition of the film-time -- it is closely related to music. You have achieved good durational proportions between the scenes where the events are almost without movement -- how astonishing that a film which is relatively taut and brief should have the courage of slow tempi, pauses, rests -- and those where they are extremely fast -- how dazzling to have chosen for these the newspaper cuttings displayed at all angles on the screen. What's more, the relative density of the changes of tempo is well done. You have let each element arrive at its own irreplaceable moment; and there is no ornamentation. 'Everything is essential,' as Webern said in similar circumstances (but everything in its time, one should add)... I like the sharpness of the film, the strangely flashing movement of the camera in the street scenes, and the empty walls of the hotel room on which the camera comes to rest for long periods, that bareness from which it cannot break away. I also like the 'unreal' condensation of time, and yet one never feels hurried. Progress is only possible on that ridge between truth, concentration and that sharpening which penetrates by burning into our perception of reality..."

Valid for the twenty-minute account of a day in the life of a West German officer -- a day particularly rich in incident -- these remarks could apply equally well to the treatment, to the transformation into fifty-five minutes of film of fifty no less busy years of German history (Not Reconciled), or of thirty years of intense musical creativity into an hour and a hairs flow of images and sounds (The Chronicle of Anna Magdalena Bach). One can simply try to discover the function assigned by Straub to this treatment of time: what is the rhyme or reason for this combination of overloaded signifying nodes, saturated with information (sometimes to the limits of our capacity for assimilation, our speed in deciphering them), with the pauses, the "sustained notes", the empty fringes, the blemishes and "unnecessary" temporal effusions, the vacant passages (which can come at the beginning or the end of a shot, and sometimes exercise a shot in its entirety, in the insistence on its progression). One seems to be able to divide this function into three registers, to link it to (at least) three orders of preoccupation:

1. structural, rhythmic, compositional: interplay of continuity/discontinuity, of retention/protension, conducted on the model of the "lacunary body" (22),

2. anti-expressive: referring therefore to Stravinsky's phrase suggesting that music is incapable of expressing anything whatsoever, with as corollary the "empty" shots (though not necessarily empty of characters), voided of everything that might involve ascendancy of a meaning, domination of a prior intent,

3. transformational: on the one hand, the time specific to the film effects the takeover and mutation to its own behalf of chronological, referential, "vital" time, but in order to validate this strictly filmic regime (to succeed, for example, in conveying a whole life in an hour and a half, to give the impression that it is unfolding and not merely captured at certain privileged and emblematic moments), it must also reinvest this time of life in shots whose rhythm and continuity seem to adhere to it. It must mark, alongside the gulfs and breaches into which ("vital") duration is swallowed up from sight, moments where one has the impression that it has time to pass into the film-time. An effect one might call "effect of temporal reality", engendering a very particular type of suspense, without finality, which acts on us as a power to recharge and reactivate, subjecting our attention to a beating, throbbing flow.

RIVETTE: And a purely formal suspense: what is the shot going to be? And not: what is going to be in the shot? At the same time, this desire to empty certain shots, to have a shot filled with information followed by one which seems to offer none, or, likewise, the proliferation of false information at certain points (false because non-referential in the context of the film, non-"informative": false trails where the reader's memory and powers of concentration lose their way -- the mass of proper names, the paprika...), all this seems to me to form part of what enables the film to function as an account of the unconscious. The film must be over before its reading (its re-reading) can be started; the telling of the dream must be finished so that the analysis, setting aside all non-literal matter, can discover the recurring, genuinely significant elements, together with the slips of the tongue, the masks, the metamorphoses, the censorships.

Translated by Tom Milne

  1. In the initial stages of our (attempt at) systematic reflection, our reading of the first issue in the "Change" series, entitled Le Montage, and even more particularly the very fact of its appearance, were extremely influential and, as it were, encouraging. For, keeping all due proportion in mind, what we were trying to do in examining montage more specifically in the cinema, was to establish ourselves within the same general problematic: that is, an inquiry into all notions of liaison, juxtaposition, combinative (and their corollaries: difference. rupture. analysis). An inquiry which we had hoped would implement itself both through our analysis of the films themselves, and through the play (itself combinative, in that it set up multiple transitions from one film to another) of this analysis.

  2. Here, it became evident after the discussions at Aix, one must consider two negative directions: montage attenuated by the length of the shots (Gertrud and Marie pour memoire) and montage made unobtrusive by a continuity -- narrative or musical -- which obliterates the passages from one shot to the next (The Empress Yang Kwei Fei and La Regle du jeu). In the first case (attenuated montage), the paucity of liaisons can of course be tactical: the montage is all the more concerted in that it is sparing of its means and effects. In the second (unobtrusive montage), the montage is concerned to make one forget its presence, to conceal its function of discontinuity. In both cases, therefore, the negative idea can be reversed: to attenuate the montage, or make it unobtrusive, is still on occasion a montage tactic.

  3. A misconception is still rife which one would have thought should have been cleared up long ago: the identification of montage (in its active effects) with rapid montage, thus assuming that the work of montage necessarily implies the proliferation and atomization of shots. In the silent Soviet cinema, the fragmentation of scenes might at a pinch mobilize the attention sufficiently to obscure the rigorous work of articulation between one sequence and the next (as we know, however, this construction was of prime importance for Eisenstein and Dovzhenko), but it is difficult to see why this unwarrantable identification of montage with short, sharp, staccato discontinuity should have continued in currency over the last ten years. For a sizeable portion of the modern cinema is concerned with the movement of compact blocks, the arrangement of long, continuous effluxions, the gradual and carefully controlled imbrication of homogeneous parts, of narrative elements that seem themselves to be seeking and indicating their most appropriate position within the overall system of the film. La Chinoise is a characteristic example of this sort of film, where no definite intention pre-existed the arrangement of its parts, where the logic of the narrative has imposed its authority rather than been itself imposed by the "author," creating its own connections, entailing the inclusion of one element or the rejection of another through the movement of its own genesis, tentative, hesitant, finally infallible, each block that is displaced retaining the trace of its passages and its redistributions, the imprint, the mark of the combinative. The montage, consequently, is not work on a pre-existing material, but work by that material, self-fashioning, self-productive, at once mould and matter, locus of a movement and sum of the constituent elements of this movement. Marie pour memoire, similar in certain ways to La Chinoise: homogeneous blocks, undivided, units displacing themselves as index throughout the still absent matrix of the montage (becoming, at their point of insertion, designated object instead of index). The montage thus cannot be said to be better or worse than it might have been (a normative illusion that refers a product back to a perfect and definitive model) since, being what it is, it also comprises its own attributes. What can one deduce from this? 1. All charges made against Garrel's work to the effect that it falls into the category of symbolism are suspect (it is, on the contrary, literal in the quality of its materiality: the body translated to the letter, the letter translated to the body). A radical differentiation must be made here between 'symbolism' as a fixed and rigid system, and 'the symbolic' as understood by Lacan: which is mobile, constantly shifting and substituting, leeway for signifiers. 2. This idea of montage as being what it is cannot result where the film-maker simply decides this is how it should be; it derives its necessity from the work, whose locus it has been, whose imprint it retains, as if a profound, muffled upheaval had taken place and persists in the final form of the film (each shot wins its place itself). Work, in Garrel's case, that is to be understood in its gynecological sense of labor: just as the woman in child-bed retains the perturbations of her pregnant state over long months, so with the film: mingled tranquillity and ferment. And birth, the fertilization of the shots, has occurred in every way, reciprocally and against the grain as well (thus, in the film, Marie is mother, wife and daughter to Jesus).

    In La Concentration another process took place. Shut away with his actors and crew in a tiny studio, Garrel filmed for three whole days without interruption. At one point, with tension, fatigue and other imposed conditions all playing their part, he began to fear that the sequence of shots (which were also to form the chronological sequence of the film, since he had for the first time pre-planned his montage) might be overloaded with too great a charge of intensity. So he changed the order of shooting. first filming the end, then the penultimate section. Is this montage generating the shooting and its methods?

  4. One can perhaps get an inkling from Chytilova's film of how the very principle of montage risks becoming a principle of rejection and suppression -- and not merely of elision, but quite literally of subtraction, erasure, or even impediment and 'persecution' (of the spectator-voyeur: thus Eisenstein refused to let him contemplate the trajectory of a gesture, and forced him to create an "idea" of the action while refusing him the pleasure and short-circuiting its conditions). Thus montage doesn't mean adding but withdrawing (and the withdrawal in action), not doing but un-doing: the negative at work. The film should be seen as a residuum, the network of traces left by the dual process of an action (the shooting, a process of accumulation) and its negation (the montage, a process of consumption): the latter thus functions by 'intaglio', not as the absence but the act of hollowing itself, the effacement, the movement of retreat, of the Other. At its extreme, film is the rejection of film, its contradiction (its 'anti-film'?): only the milestones remain, the tokens of its "passage," forever past/future; just as the film in the projector exists only through the effacement of one by the other, the incessant difference, the consumption-destruction of all its "images": a false presence, a deception constantly renewed, constantly deferred. The montage is the functioning of this deception.

  5. The extremism of this systematic could be quite well represented by a film like Taylor Mead's European Diary which, filmed image by image, can save itself the bother of an editing stage and be edited as it is shot. The extreme rapidity of the liaisons goes so far as to prohibit even perception of each shot The spectator, ruthlessly left all at sea, is gripped by a dizzying monotony. But then another discourse may perhaps establish itself (if the spectator brings a certain goodwill to it, or helps out with a certain protective conscience), not in the film but in the spectator himself, on the strength of infra-perceived fragments of the film. Like those complicated, overabundant dreams that are immediately forgotten on waking (a).

    a) Opposite perspective: Andy Warhol's (more) celebrated (than seen) Chelsea Girls. Absolute non-montage, since the film is merely the alternation or juxtaposition (ordered by chance) of reels exactly as they left the camera, uncut and including both unforeseen accidents and reel-ends; and yet the simple fact of projection, therefore of the successions and the simultaneity (through the coexistence of the two screens) of raw shots creates montage: different each time, but inescapable. As though one couldn't leave the circle, as if it were impossible to break the montage's seal. (J.R.)

  6. With Cassavetes the use of montage is very different: naturalistic. (As Jean Narboni rightly says, Cassavetes' work is a "natural expressionism.") What has to be expressed is edginess, doubts, hesitations, illuminations, fleeting and contradictory expressions, lassitudes, irritations, idle moments and bursts of activity succeeding one another as they do in life. The montage becomes the privileged means: the instrument of touch. And the phrase is used only barely metaphorically: in painting, a touch of green brings a realistic contradiction to red. One of those contradictions brought to life out of respect for life. It isn't a question of nuance (nothing is more assertive than nuance), but rather of a war waged, by tremors and hesitations, on meaning in its living inexactitude.

  7. To be understood, in the circumstances, in its gymnastic and sporting sense rather than the Marxist one, I presume (as we shall be using it later). (J.N.)

  8. Cf. Jean Narboni's criticism further on of this correlation of shot and word; but it is a fact that Pollet wanted to take each element in the film -- Venice, operating table, Greek temple -- to its most extreme point of "purity'"(and modesty): like the words in the Mallarme-ish poem (the reference to Un Coup de des is explicit throughout Sollers' text), polished, orbited, crystallised, as though cut off from any lexicographic impurities. Whereas Godard works to destroy this minimum element (this moneme): similarly Joyce "worked on" his words both from "within" and in the dictionary context: dismembered, dissected, collided, commingled.

  9. Even through its title, Chytilova's film also poses a question (questions). It is remarkable, actually, how almost all the titles of these films are "signifiers" of their functioning: About Something Else, of course. but also The Old and the New (which are at work, and in conflict, in each sequence, each cell, each frame), Not Reconciled (true of each shot, locked in on its own cognizance: deliberate banishment of compromise from the Adenauer world, tranquil rejection of a sham harmony). Intolerance, Made in USA, Pour la suite du monde, Mediterranee ("sea surrounded by land"): each of these titles is like a "directions for use" for the film. Whereas Gertrud, The Empress Yang Kwei Fei, are merely labels (though it would certainly be easy to find Renoir or Rossellini, but also Ford or Dreyer films with ambivalent titles like this, more or less clearly indicating the same awareness of form as being the "content of form"...). (J.R.)

  10. And it is precisely in this distance maintained by Pollet with regard to Eisenstein (his rejection of the dictatorship of meaning) that the case of Solanas should be considered. In The Hour of the Furnaces, there is an extremely violent dictatorship of the discourse (implemented chiefly by the montage of sound and image). Those of us accustomed, in a French context, to the idea that an obligatory meaning is a reactionary principle, are therefore suddenly required to go an extremely long way in adjusting the relativity of our reasoning. As far, in fact, as bringing it directly into touch with the current situation in Argentina. Is this meaning, then (that Peronism, as an already existing force for regrouping the popular masses, should be the starting-point for propaganda and revolutionary action) to be imposed on the Argentinians? I confess I have no idea. But in any case doesn't the very notion of political violence (a revolution, for instance, but it could equally well be to do with reaction of fascism) itself impose, as a corollary, the notion of violence in meaning? So the production of an obligatory meaning would no longer have to be considered reactionary, but one would have to consider nothing but this meaning. Are we forced into this regression by politics themselves? Is the notion of an open work, in other words, one of the last manifestations of Western liberalism? Or must we examine the contexts of the work yet again with greater stringency? (S.P.)

  11. Cf. the authoritative articles by Jean-Joseph Goux: "Marx et lÕInscription du travail," "Numismatiques" (Tel Quel, Nos. 33, 35,36).

  12. Conversely, we know the way in which Eisenstein took over Lang's Dr Mabuse, to re-edit and correct it. We also know how, in order to give it its full political meaning, he betrayed Dreiser's novel An American Tragedy: "Undoubtedly a first-class novel -- although not, from our point of view, a class novel" -- stripping it of all the vaguely "progressivist" ideology that encumbered it, instead of giving his producers "an uncomplicated whodunit with a good murder and a nice love story" (of course they turned it down). (Notes of a Film Director)

  13. Note how it is the same desire to annex to the contemporary "vocabulary" of film, in one case the texts of Duras, Robbe-Grillet, Cayrol, but for the other the speech that is most threatened (everyday, contingent, trivial, transient) -- which seems to constrain them to rediscover the techniques of discontinuity.

  14. Neither the Russian release title (The Old and the New) nor the title Eisenstein wanted to use (The General Line) give a correct account of the real dynamic and overall system of the film, in that both titles still belong to the category of linear and continuous time, progressively generated, the time of historical succession (a succession broken in the release title by the sudden shift to the New, a movement of dogged progress towards Communism in Eisenstein's title). The film, on the contrary, functioning through blocks and ensembles, through discontinuous series, is never bisected once and for all by the miraculous line which supposedly marks the definitive passage from Old to New (a type of "progress" characteristic of "liberal" American films, complying with the ideology of an unbroken history guided by some starry horizon of enlightenment); each scene, moreover, is itself traversed by this Old/New line, the movement is one of more and more radical leaps from scene to scene, each one embodying all its predecessors before being absorbed in its turn. This movement might be resumed diagrammatically:

    S (1) S (2)
    Old/New Old/New
    Old New
    Thus one sees that the film, itself, in its entirety, and only at its end, can be described as "the New," with the important reservation that no sooner has the last shot faded than the film is surrendered in its turn to the Old, demanding that its revolutionary movement be in its turn taken up, extended, pursued by each spectator, and this time in life.

    The impossibility of representing the movement of Eisenstein's film in our unidimensional Western language clearly demonstrates the extent to which the film effected a breach in History (history of the cinema, its storytelling) through the establishment of a volumetric space, a plural time, a complex topology. (J.N.)

  15. Of what is direct (sound/cinema) the proof? Cf. in particular "Le detour par le direct'"(Jean-Louis Comolli, Cahiers du Cinema, No. 209): Direct sound = sign of the eruption of a fragment of "reality," evidence of the operation of interception at a certain precise moment in History (hic et nunc, but also past/elsewhere) of some fortuity: the take is a product by machine of the "event," a primitive inscription by encounter: hence the rushes, the nascent stage of the film. And montage = tactic for encounters between successive sound-images, but at the same time between soundtrack and visual track, towards "the" film in its posthumous state. Dual dual-process, and productivity through bringing into contact (into conflict) these two re-fissured 'blocks': film 1 and film 2 (in front/behind the camera//co-existence/succession of images-sounds). A dual intersection spatializing the process of the film (a dynamic "cube'" on all spatio-temporal vectors.

  16. It should be noted, moreover, that it was through the action of the oral discourse which governs the film -- or rather the "series" of fragmentary discourses -- that Straub found himself forced, as it were (after a first chronological "treatment"), into his definitive construction.

  17. Not through any desire for obscurity, but on the contrary because it pushes to their point of fusion, simultaneously, all the functions (rigor of the liaisons, autonomy of the elements) which are more commonly used only in succession and in a looser manner, and reduces -- too much, all things considered -- the area of imprecision to which one is accustomed; likewise there is a Mallarme-ish side to the text which is '"obscure" only by sheer force of speed and logic, by force of clarity.

  18. A form which "explains itself," since it is Fascism that is in question here. In the same spirit, I must confess I cannot resist the temptation to write: in so far as the film is structured as a language, it acts as though (it mimes the action of an) unconscious.

  19. This is unequivocally signalled both by the nightmare with the dogs (and its recurrence in the tapestry) and by the very specific allusions to Charcot's group and methods of hypnotism.

  20. What is suggested here about Dreyer could probably be applied equally well (with all the evident changes and "corrections" made) to Mizoguchi: but the film we should have seen again is Ugetsu Monogatari or The Life of Oharu, rather than The Empress Yang Kwei Fei. Let us simply recall here the phenomenon of gliding between multiple levels, abetted by the indecisiveness, the instability of the "signs" marking each of these levels, with which the movement of the film and most of its elements are informed...

  21. A question must then arise (a question that remains open here): can films where the formal work intervenes only at the montage stage, without previous work on the ecriture, relate back so directly to the workings of the unconscious? Mediterranee certainly "works" on the unconscious (the reader's), but does it function as such? Or again: if there is retrenchment of the pre-text, can there be a "return" of the repressed?

  22. One can compare the 'holes' productive of "silences" in Webern, of which Boulez said that they did not act simply as elements of rhythm, but modify the neighboring sounds, acting on the morphology of pitch. "Silence" in Straub's films has a similar operative function, being not merely pause, scansion, but acting on the "frequency," the vibration of the preceding shot (or beginning of the shot of which it is the end) and the one following (or continuation of the following shot where it is the beginning).

Originally appeared in Cahiers du Cinema No. 210, March 1969. Translated by Tom Milne. Published in this form in Jacques Rivette: Texts and Interviews (British Film Institute, 1977).