‘Room, door… room’: Oneiric Space in L’Avventura and Last Year in Marienbad


Wilson, E. P. (2018, May 21). “Room, door . . . room”: Oneiric Space in L’Avventura and Last Year in Marienbad. Dreaming. Advance online publication. http://dx.doi.org/10.1037/drm0000075.

Around two years ago I watched Michelangelo Antonioni’s film L’Avventura (1960) followed by Alain Resnais’ Last Year In Marienbad (1961). Both are highly experimental and significant entries in film history, and both, I noted, illuminate the differences between lucid (Marienbad) and non-lucid (L’Avventura) dream states. This essay, which I have revisited over these two years, seeks to marry current dream theory to avant-garde film theory, critically unearthing these films that seem almost impenetrable to criticism. My research primarily builds on J. Allan Hobson’s research on REM (Rapid Eye Movement) sleep cycles, and its ongoing conflict with Freudian and Jungian psychoanalysis. I propose the thesis that the treatment of composition and space onscreen and off-screen, essentially what we see and what we do not see, interacts with the subconscious of the characters in the films, and through that, the spectator’s subconscious mind.

L’Avventura, as a non-lucid dream, demonstrates a drifting through setting and space – a lack of attention and a forgetting – correlating to successions of meaningless thoughts and actions in its characters. Marienbad, however, displays an awareness of its own vivid absurdity as it locks and focuses into space, cultivated in manipulations of time, setting and light. As Andrei Tarkovsky (1986) suggests in his book Sculpting In Time, the aim is to observe a dream or film “as one admires a landscape”. To bring a personal reflection on watching the film to a mode of film criticism, and in doing so, illuminate a way of watching avant-garde or experimental cinema.




Taking Tarkovsky’s comment literally, then, physical and abstract elements of a ‘landscape’ or spectacle seen in waking life have the potential to be translated to spectator’s dream memory. As Hobson (1980, p. 14) theorises in Film and the Physiology of Dreaming Sleep, when one dreams the brain behaves “like a projector” where “stored images are pulled out of memory and assembled into a synthetic perceptual whole”. It is undoubted that images seen in reality reappear in dreams, but this comparison seems limiting. The projector displays an image that is cropped on four sides; the pro-filmic or ‘screen space’, as Noël Burch (1981, p. 106) names it in Theory of Film Practice. Hobson’s theory rightly points out that, for example, a chair can be projected from the memory of an image seen in waking life to an object in the dream. But a projector limits the field of view; you do not see what is around it or outside it. If the brain behaved like a projector, it would adhere to the constriction of space, to the four sides of the projected image. The dream, on the other hand, creates the illusion of space circulating the dreamer, which is why it seems so real. This is not to say that there are no projections of stored images in dreams, that is, for example, a chair or a human, but the dream usually assumes the role of the eye rather than the camera, and as F. E. Sparshott (1971, p. 13) points out, “an eye is not a camera”. 

I believe Noël Burch’s categorisation of ‘off-screen’ space illuminates Hobson’s statement. Burch (1981) explains that this includes the space outside, behind and in front of the frame, and subsequently the action, or rather the feeling of action, occurring within these spaces. Crucially, the suspicion that something is taking place just outside of one’s visual capacity is a central, common element of dreaming. So, to link this with Hobson’s statement, stored images are indeed pulled out of memory and assembled into a perceptual whole, but so are stored images, feelings and sensations that we cannot see. Because of this, we cannot be completely certain of what is being projected from our brain when we dream, whereas the image from a projector reveals its image onscreen. 




Inge Strauch and Barbara Meier (1996, p. 138) report in their dream research that a N-REM (Non-Rapid Eye Movement) dream, which would not be lucid, “offers greater uncertainty in its dream course”, the dreamer acting more as an “observer or commentator”. This is crucial to understanding the link between film and dream as the dreamer, even when lucid, never has total clarity over their dream domain, just as neither projector nor spectator have complete views of the world portrayed in film. Translating real spectacles into dream projections subsequently gives rise to the potential of uncertainty, allowing the spectator to question the reality of what is seen, and unseen, as I will now elaborate. 




David Forgacs (2000, p. 108) indicates that L’Avventura’s island sequence on Lisca Bianca lingers on this uncertainty, emphasising the “possibility of someone or something hovering just outside” the frame. It is true that in adopting an oneiric sense of space, this particular sequence withholds images from the viewer. It becomes a mode of storytelling that subverts its visual expectations, reminding us of what we cannot see. It elicits a dreamlike uncertainty that deliberately confuses the spectator. In one shot, entirely dominated by the landscape, Sandro is seen entering from the bottom left corner of the frame, making his way to the centre and then stopping to call out for Anna, who has disappeared. The camera then tracks his movement as he almost exits the frame and is ‘replaced’, as Forgacs (2000) would say, by the landscape. He walks further away from the camera, stops abruptly, and turns his head to the right, as if something off-screen and on the ground had caught his attention. In the next shot the sea crashes violently against the rocks, and it is clear that this is not Sandro’s point of view as he enters from the bottom right of this shot in a similar fashion. While small, this is a detail left unknown to the spectator, and creates the same feeling of uncertainty referred to by Strauch and Meier in their dream research. 




But as Forgacs (2000) also comments on the types of pressure Antonioni creates within the frame, it is necessary to point out that these ‘hidden’ dream-images also have the potential to both represent themselves within the pro-filmic space and produce neurotic thoughts in the film viewer. That is, the onscreen manipulation of images, such as when Claudia mistakes the distant, obscured figure of Giulia to be Anna, and the paranoia that there is constantly some unknown detail to be looking out for, as experienced in the long, unnerving shots of ‘people-less’ landscape (Forgacs, 2000, p. 101). Freud (1899/2010, p. 73) notes in The Interpretation of Dreams that “we very often believe that we remember [the dream] incompletely”, and therefore spend time trying to remember it in waking life, similar in a way to Leslie Halpern’s (2003, p. 29) description of the ‘neurotic thought loops’ experienced in N-REM dreams ‘(Did I lock the door? I know I locked the door. Did I lock the door?)’ Because of this, there is an endless cycle of uncertainty in both the film and the dream, extending beyond the island sequence as any presence of Anna slowly fades from the film altogether, just as a dream “fades in the course of the day, leaving only a few trifling remnants” (Freud, 1899/2010, 73). Some spectators may ponder the disappearance of Anna (what happened to Anna?), trying to remember if he or she missed a vital detail. Others may forget about her entirely, as the other characters do, only to ask the question once the credits roll. The fallacy of human visual memory appears to be the sole idea almost agreed upon by both Freud and Hobson, the latter reporting that “practically all dream experience is forgotten” after waking (Hobson, 1980, p. 15). While this is certainly debatable in the greater context of dreams, in the context of L’Avventura and its bleak, distracted alienation it highlights the idea that characters are trapped in a perpetually half-forgotten dream, some of which the spectator experiences.\




To say that ‘all dream experience is forgotten’ in L’Avventura would perhaps be too extreme, though the aimless, repetitious drifting through different environments, never once revisiting a setting already seen, mirrors the progression of a non-lucid dream. In light of this, the film breaks down into sequences: the beginning, the island, the train, the empty town, the crowds, the tower and the party. There are, of course, more scenes than this, but the point is to envision a formulaic structure to portray aimless, existential wandering paired with bored dialogue. Claudia, for example, asks Sandro to “tell me you love me”, to which he replies, “I love you”. She then asks him to “tell me you don’t love me”, to which Sandro replies “I don’t love you”. Sandro is turned away from the camera, and Claudia’s face is blank before a weak smile. According to Roger Ebert’s (1997) review, love in this film is ‘an attempt to pass the time’. Clearly, there is nothing felt. Not only this, but it feels like one of the characters’ many attempts to control their surroundings – Anna making up the shark, Claudia ringing the bells, Sandro deliberately knocking over the ink on the architect’s drawing – equivocal scenes, but all sharing a feeling that characters are trying to satisfy themselves by controlling minute actions in their otherwise meaningless existence. A scene in the script was supposedly cut where one of the characters on the island suggests Anna “simply drowned” (Antonioni, 1996, p. 273). That death is simple reveals the reality of the film, a reality that never changes, and the dawn that ends the film ironically marks nothing but emptiness in the characters. There is, as Hobson (1980, p. 15) says of dreams, ‘a delusional acceptance’ of their existence, therefore trapping them in a dream that is not only half-forgotten, but never-ending. 

I have established here that there is a consistency between L’Avventura and the more existential aspects of dreaming, namely the drifting and the forgetting. It would be misleading to suggest that any film can represent the entirety of the dream world, however, or that the dream in its most conceptual sense is translated correctly to the screen. Rather, there are spatial dream elements that permeate parts of the film, gathered by Hobson’s ‘projector’ and reproduced as a ‘synthetic perceptual whole’ (Hobson, 1980).




In realising this, it is necessary to pause and consider how a film can directly influence the dream state, and vice versa. Freud’s thesis that every dream symbolises a wish fulfillment is questionable, as Hobson (2005, p. 46), who dismisses it as ‘the mystique of fortune-cookie dream interpretation’, emphasises his own thesis that the brain is a projector in dreams. Carl Jung (1959/1981), on the other hand, theorises that there are dream ‘figures’ – an old wise man, a mother, a trickster – common to all dreams, which is understandable in the sense that there are subconscious projections in dreams, yet just as questionable as Freud’s thesis as both generalise dreams to a stereotypical few. Indeed, they all acknowledge that dreams reveal aspects of an individual. As far as non-lucid dreams go, L’Avventura shares elements of the dream space, but as Hobson (1980, p. 20) points out by analysing the beginning sequence of Ingmar Bergman’s Persona (1966), dreams have the potential of holding up a basic story or simply ‘a sequence of visual images linked by a strong emotion.’ 




Critics like Sparshott (1971) rightly argue that there are differences between the film and the dream, emphasising that the film watcher is awake. It also depends on the type and genre of film being watched. A modern Hollywood blockbuster, for example, would perhaps be less situated to dream theory as an avant-garde or experimental film made during a time when the narrative film was undergoing revisions. A classical design, at its most basic, would follow a clear protagonist struggling against forces of antagonism, towards a definitive, closed ending. What you see onscreen is all you really need to see. 




Instead, I propose the idea that both films represent the dreams of their characters. ‘Oneiric Space’, then, refers to characters that are experiencing space in front of and beyond the camera as their own dream, lucid in Marienbad, non-lucid in L’Avventura. It is difficult to make sense of Marienbad, in fact I do not think a spectator of the film should try to make ‘sense’ of it. Rather, I believe the images should speak for themselves, and be completed with the spectator’s own thoughts in the moment of viewing. I could go into depth about the context in which the film was made, but I do not think it relevant to this article. Even Resnais and Robbe-Grillet, the film’s screenwriter, appeared to contradict each other about what happened in the film (Brunius, 1962). Robbe-Grillet wrote a screenplay, Resnais directed. Nothing else needs to be known to view it.

In one scene, I note, the film watcher witnesses the lucid dream of ‘X – the man with the Italian accent’, who, following his arrival, has the power to create and edit scenarios between himself and ‘A – the brunette woman’. In one shot ‘X’ commands ‘A’ to “Remember” as he sets the scene in “the gardens of Frederiksbad”, cutting to a slow tracking shot of the gardens from a balcony. As soon as he tells her that she was alone, the shot reveals the lone figure of ’A’ standing on the balcony. ‘X’ tells her “you were leaning on a stone balustrade, resting your hand on it, your arm outstretched”. What follows is ‘A’ mirroring his words in her actions, as if X is in total control. He says he told her how real she seemed, as if he was not present in reality or a real world, further suggesting his dreamer position. In another scene, ‘X’ and ‘A’ are facing the camera, and in mentioning ‘A’ as “a shadowy figure” that “slowly advanced”, ‘M – the other man with the thin face’, ‘A’’s husband, appears behind them as a shadowy figure and slowly approaches. Like ‘M’ and ‘A’, then, Claudia and Sandro are trapped at the end of L’Avventura in an existential void, Sandro, crying after having slept with a prostitute, and Claudia looking on disconnectedly. In the context of these two films, Andre Bazin’s (1971/2005, p. 171) statement that “it is a mistake to equate the word dream with some anarchic freedom of the imagination”, that “nothing is more predetermined and censored than dreams”, rings true for L’Avventura, but for Marienbad, at least for ‘X’, there is a sort of anarchic freedom, though one that for some reason is constantly grounded in the same setting. The question this raises is one of whether or not the lucid dream is more connected than a non-lucid dream, and I will examine this next. 




In order to do so I would like to revisit Hobson’s earlier statement that dreams are often “a sequence of visual images linked by a strong emotion” to examine the connectivity of events in Marienbad as a lucid dream against L’Avventura as a non-lucid dream. I would also like to revisit the initial quote from Hobson (1980) in this essay; that stored images combine to create a ’synthetic perceptual whole’. There is a suggestion here, in order to prevent Hobson taking steps into Freudian territory, that this ‘whole’ is random. Leslie Halpern (2003, p. 29) would perhaps agree with this assumption, as according to her the REM sleep cycle “initiates random electrical signals to the higher mental centres of the forebrain”, and is therefore responsible for vivid yet disconnected, meaningless images”. However, she states that this is because of the “lack of external stimuli” (Halpern, 2003, p. 29). Likewise, Hobson (1980, p. 21) points out that “images from the outside world are not available”. Neither seem to take into account the power of sound. In L’Avventura, there is silence, penetrated at times by a haunting score. In Marienbad, there is an unnervingly hypnotic musical assault that is penetrated at times with total silence. There is a scene towards the end where the screen is filled with so much light that it is difficult to distinguish any image, which then cuts to darkness. One becomes aware that the final corridor shot of the film is in fact not a corridor but the illusion of a corridor (stairs), demonstrating the deconstruction of this setting as ‘X’ leaves his dream. In a sense, then, these provide sequences of visual images with a single strong emotion running throughout, but it is crucially a sensory experience. Strauch and Meier (1996) report that the two most dominant perceptions in dream are the visual and auditory, but Hobson and Halpern do not address this. The plots of both films are seemingly secondary to the experience. 




There is one final question to consider here, pointed out by Christian Metz (1982, p. 121) in The Imaginary Signifier. To achieve ‘true absurdity’ in film, as Metz calls it, is impossible, and subsequently any ‘dream sequence’ in narrative films are ‘nearly always unbelievable’. True, the purpose of this essay is not to argue that dreams can be perfectly represented in films. What dreams do for films, however, is provide a different style of criticism. As I stated initially, the screen space complements the dream space. In a way, this ‘dream language’ in film has been present since the beginnings of cinema – the Lumière Brothers’ Arrival of a Train at La Ciotat (1896) for example – tricking spectators into seeing and feeling an illusion. 




My paper set out to explore the marriage of film theory to dream theory through the analysis of two landmark films. In doing so, I pointed out how the treatment of space onscreen and off-screen interacts with spectator and character. The spectator, in this sense, is not dissimilar to the characters being spectated. Characters drift through scenes, as in a dream, but the spectators are watching this dream. The language used to convey this further allows the spectator to identify and create personal associations with the content of the film, similar in a way to the personal content of their own dreams. As a result, no clear plot or narrative is needed; the language speaks without words and is unique to the viewer. And I am not just ratifying the value of film experience, but pointing out that this particular mode of spectatorship has always been present in cinema, in different forms, and has the danger of being lost. The reason being that one has to look as far back as the 1960s to find landmark avant-garde or experimental films to combine what one sees with what one feels and imagines. This is, I believe, what matters about the argument presented, to unearth these films and point out a different, while perhaps not new, direction under which to theorise them. The films transcend their mechanical format, and the effect is mixed, neutral, but deeply human in its feeling.