'Marienbad Year Zero' ('Marienbad annee zero')

Alain Resnais 1961

Alain Resnais 1961

There will be no lack of people ready to stress Marienbad's exceptional nature and I have no intention of competing with them. I would rather take a step in the opposite direction. Alain Resnais's most recent film is a film with a historical place. It was made in particular conditions and was undoubtedly based on accumulated experience. It fits of necessity into the final phase of an evolutionary development in cinema, or an idea of that development. For that reason I will begin by saying that Marienbad is the last of the great neo-realist films.

You will recall Bazin's analysis of narrative art in the Italian neo-realist cinema, Paisà in particular.[1] That was fifteen years ago. Neo-realism was as yet the future of cinema. It was to bring about the first significant revolution in film art, without which the works we love today would not be quite what they are.

Essentially, neo-realism replaced the classic scenario and its theatrically-based arrangement of scenes by an open-ended scenario, which was consequently closer to the experience we have of reality. The neo-realist film presents a sequence of fragments bound by no apparent logic and separated from each other by gaps, the gaps and fragments representing upstrokes and downstrokes on a canvas which bears no relation to the close-woven fabric from which the cinema had hitherto drawn its sharpest effects.

The notable result of all this is that the new conception of cinema entailed a new way of looking at film. The passive spectator was succeeded by an active spectator who converted the discontinuous narrative threads into a coherent continuity, just as he would do in everyday life. From that point the film no longer functioned without him.

This revolution in narrative method - as Bazin also showed[2] - joined the revolution represented by Citizen Kane and The Magnificent Ambersons, although by a different route. Kane, like Paisà, also turned its back on the finished view of the scenario and mise en scene. They no longer offered predigested material ready for absorption. On the contrary, they present the spectator with raw material (even if it has in fact been heavily worked on) from which he may extract his own film. In short, the meaning of the film is no longer imposed on the spectator but has to be constructed by him from the basis provided by the elements of the film. As in everyday life again, the meaning of events is only ever a hypothesis, and if images constitute a language[3] it is a language which lacks signification.

In this perspective it at once becomes clear how Resnais's films, and Marienbad especially, fit naturally into the groove dug by neo-realism. The same gaps in the script, the same ambiguity in the events, the same effort demanded of the spectator. Some will object, with apparent justification, that at least it is possible to make out scenes between the gaps in Paisà. But can we be sure of this? Are we not prey to self-deception? Is it not rather the case that the spectacle of Paisà presents us with fragments of reality we have already appropriated to ourselves? If 'to represent' means slicing reality off from its past and its future in order to turn it into pure event, then even at the level of news film the project is an equivocal one. The proof is in the wide range of meanings instantly attached to an incident whether read about in the columns of the evening paper or seen in a newsreel. This shows how far we are incapable of looking at (I won't even say understanding) an incident without interpreting it and without our look adding to the event to form an amalgam, a mixture which by nature belongs as much to the documentary image as it does to the fiction with which we envelop it. All this boils down to saying that you cannot read reality as you might a novel, you read into it.

Traditional cinema had managed to do away with any possibility of ambiguity by building into every scene and shot what the spectator was meant to think of it: i.e. its meaning. Taken to its extreme, this kind of cinema did not need the spectator since he was already included in the film. The novelty of the films of Welles and the great neo-realists was to demand expressly the participation of the spectator. It is in this sense that we need to talk of phenomenology here: the spectator's look is as much creator of the film as is the intention of its makers.

By comparison with Welles or Rossellini, the originality of the films of Resnais and Robbe-Grillet is to have systematized this discovery. In Kane and Paisà the gaps in the narrative were in some sense tolerated; they were necessary because they were inevitable. But Resnais and Robbe-Grillet have turned those gaps into the actual theme of Marienbad. To such an extent that the film seems like a surface from which rise enigmatic images whose only palpable certainty is that they are part of the same film. In this sense Marienbad is a documentary, but a documentary of a special kind - the spectator does not know what is being documented. In short, Resnais and Robbe-Grillet have put together some documentary records, pilot images so to speak, leaving it to the spectator to convert them into a fiction which gives them a meaning. The first man to confront the nature of the world must have had the same difficulty in understanding it - ordering it. Faced with Marienbad, one is tempted to cry, like Valery: 'It's as if the world were scarcely older than the art of creating the world!'

It is curious that no one has remarked on the fact that none of the craftsmen of neo-realism - not Visconti, Rossellini, De Sica, or even Fellini - resorted to the device of recalling the past. Welles himself abandoned it after his first film. It is as if the whole great effort towards more realism in general, with which the history of neo-realism merges, inevitably culminated in the elimination of anything which might disrupt the chronology of events. Since in the last analysis the justification of cinema was history, historical chronology had to be sustained. In that perspective the flashback of course seemed like the ultimate kind of fraud, an inadmissible recourse.

Once it is allowed that the end result of the neo-realist aesthetic could be a film like the one Zavattini dreamed about, a precise and exhaustive record of ninety minutes in a man's life,[4] the dishonesty of the flashback can no longer be in doubt.

Things are not so simple in reality. As we have seen, in traditional cinema any gap or lack in the narrative was considered a fault, and the absolute aim of an auteur, scriptwriter as well as director, was to eliminate the gaps and present the film as a solid, unflawed block. The flashback therefore answered that same concern. Its only purpose was to fill in the remaining gaps, to weld the story to its immediate antecedents.

But what about Kane, you will say? And Lola Montès? And Hiroshima? There precisely lies the genius of Welles, Ophuls and Resnais. They used the classic device for ends diametrically opposed to those it had hitherto served: in Kane, Lola Montes and Hiroshima the function of the flashback is no longer to efface the discontinuity of the narrative. I would even go so far as to say the opposite. Of course the old function still persisted under the new: the flashback still served the story. It was still the sign of a desire to inject meaning since it was linked to a chronology. In Marienbad that sign disappears in turn. From this point the chronology of a story can only be what the most immediate 'essentialist' assumption makes it; to say that the film ends after three or four hundred shots is to give a meaning to a film which could flow on like a river into the sea from something that might be the beginning to something that could be the end. In Marienbad there is no end to anything. 'The End' corning up on the screen does not in itself bring to a close a story which might have worked itself out in time. In short, time (chronology) does not exist outside the way things are looked at, which is why Marienbad has a double need of the spectator if it is to constitute itself as a story. If this film exists, it is as an object - like the blots of the Rorschach test.

Put briefly, any scene in Marienbad, at whatever level of reality the consciousness of the spectator situates it, shares the same realism as those around it. If I may once again take up the parallel with the mode of the traditional narrative, I would say that usually the spectator is asked to adjust to the various shots in a film (just as the eye adjusts to moving from one object to another further away in the field of vision). Marienbad, however, presents itself to the spectator as a two-dimensional object whose parts are all situated on the same level of realism. There is no objective difference between a shot of the past and a shot of the present. It is the spectator who structures the film and establishes the differences of reality which give the object (the film) its perspective, in three, four or five dimensions, say.

To sum up, Resnais and Robbe-Grillet are doing in cinema what certain abstract artists have long been doing: they are offering not a story, but a sequence of images belonging to the same level of realism which is the film, and it is the spectator who introduces the depth. For the true successor of the figurative painter is not the abstract artist, but the spectator looking at the abstract painting. There is therefore less difference between Delacroix and Nicolas de Stael. Painting has changed its function. The task of the painter is no longer to paint a subject, but to make a canvas. So it is with cinema. The work of the film-maker is no longer to tell a story, but simply to make a film in which the spectator will discover a story. The true successor of the traditional film-maker is not Resnais or Robbe-Grillet, but the spectator of Marienbad.

Andre S. Labarthe
Translated by Diana Matias


1   Andre Bazin, 'Le realisme cinematographique et l'ecole italienne de la Liberation', published in Esprit, January 1948, translated as 'An Aesthetic of Reality: Cinematic Realism and the Italian School of the Liberation' in Bazin, What is Cinema? Vol. 2.

2   Andre Bazin's theses about deep focus, or composition in depth, and realism were developed particularly in his essays 'William Wyler, ou Ie janseniste de la mise en scene', originally published in La Revue du Cinema II, March 1948, translated as 'William Wyler, or the Jansenist of mise en scene' in Williams, Realism and the Cinema, pp. 36-52, and 'L'Evolution du langage cinematographique', first published as a single essay in Bazin, Qu'est-ce que le cinema? tome 1: Ontologie et langage (Paris, Editions du Cerf, 1958), translated as 'The Evolution of the Language of Cinema' in Bazin, What is Cinema? Vol. 1 (reprinted in Gerald Mast and Marshall Cohen, Film Theory and Criticism, New York, Oxford University Press, 1974) and as 'The Evolution of Film Language' in Graham, The New Wave.

3   'But it is beginning to be clear that nothing could be more certain. Does anyone still believe in a "cinematic grammar"?' (Author’s note). The 1960s were, of course, an intense period for work on 'film language': Christian Metz's 'Le Cinema: langue ou langage?' (translated as 'The Cinema: Language or Language System?') was published in 1964; his 'A propos de l'impression de la realite au cinema' (translated as 'On the Impression of Reality in the Cinema') was published in Cahiers 166-7, May-June 1965; these essays, and later ones on semiology and narrative in cinema, were collected in Metz's Essais sur la signification au cinema, Paris, Editions Klincksieck, 1971, translated as Film Language: A Semiotics of the Cinema, New York, Oxford University Press, 1974.

4   Cesare Zavattini's ideas are usefully represented in his article 'Some Ideas on the Cinema', translated in Sight and Sound, October 1953, reprinted in Richard Dyer MacCann, Film: A Montage of Theories, New York, E. P. Dutton, 1966; a different version has been translated as 'A Thesis on Neo-Realism' in David Overbey, Springtime in Italy: A Reader On Neo-Realism, London, Talisman Books, 1978.