Un Chien Andalou: A Surrealist Film by Luis Bunuel and Salvador Dali
Max Morise commented in ‘Les Yeux Enchantes’ that, ‘the succession of images, the passage of ideas are a fundamental condition of any Surrealist manifestation.’ Does the Surrealist film Un Chien Andalou satisfy this assertion?
At first glance it may seem that Max Morise’s statement could be applied to any film, since, technically, all productions of the cinematic medium are based upon the succession of images. The presence of static frames, animated by rapid succession, is cinema’s fundamental condition. When considering his specification about the passage of ideas, however, it becomes clear that his statement refers, not to the literal succession of images, but to the logic inferred by it to a greater or lesser extent. With this understanding in mind a profound disparity does indeed become apparent between Un chien andalou and the so-called ‘classical narrative cinema’. This disparity exists between cinema based, on the one hand upon unrelated or tenuously linked images and illogical ideas; on the other, upon an intelligible narrative (generally understood as being the gradual revelation of a single idea or coherent series of ideas). The overwhelming tendency has been to see Un chien andalou as being aligned with the former mode of filmmaking. The aim of this essay is to assess the extent to which this assumption is valid. Implicit in Morise’s statement is the belief that cinema is the ideal Surrealist medium. This also warrants discussion as it allows the film to be located within the whole arena of Surrealist activity and theory.
To challenge the idea of Un chien andalou as the juxtaposition of images it is necessary to highlight those aspects of the film pointing to narrative continuity. These appear in surprising quantity and seem to indicate the presence of a coherent, underlying narrative, which the more obvious juxtaposition disguises but does not eradicate. Essential to this is the fact that we seek to understand the film by reference to our knowledge of genre conventions (which depends on our having viewed previous films). Thus can we make our way through the succession of disorientating images and events. For example, despite the strangeness of their actions, characters are for the most part motivated by recognisable human concerns. Clearly, the cyclist is consumed by sexual desire and his behaviour, though bizarre, is founded upon attempts at seduction. Similarly, the woman refuses his advances and her actions convey her disdain.
It is through our familiarity with the narratives of cinematic love-stories that we interpret the characters as suitor, love-interest, and, in the case of the man waiting on the beach, the woman’s preferred lover. Arguably, the filmmakers have let this thread of continuity run through their work so that the images, though shocking, nevertheless exhibit some form of logical progression. It is like a map provided to help us find our way in a bewildering landscape. Furthermore, out dependence on genre conventions as the only means of understanding the film makes us more aware of them, which in turn makes us conscious of the psychic processes involved in cinema viewing. Such an investigation of the cinematic medium had been central to Surrealist practice since the movement’s inception.
Searching for an underlying narrative encourages us to interpret all aspects of the film towards this end. This has been attempted by Raymond Durgnat, who is therefore a notable exponent of the idea that Un chien andalou is a narrative flow, rather than a series of disjointed images and ideas. With characteristic vehemence he reduces every event in the film to a single, definite meaning, which thus cohere into an indisputable narrative. For example, the two pianos laden with dead mules and attached to a pair of priests represents the ‘dead weight of [the cyclist’s] education’1: this and nothing else. Of the infamous prologue he writes: ‘The razor and the eye are fairly evident symbols for the male and female organ, and the cutting for sexuality viewed as a destructive activity.’2 Tempting as this reading is, his argument is not entirely convincing, chiefly because it seems to work too well. His conviction that each event has an immutable meaning ignores the possibility of variant meanings, and thus limits the film’s potential to affect its audience. For example, the prologue could also be seen as symbolising rape, castration, and an attack on vision, either to ‘shock the bourgeoisie’ or to announce the prevalence of dream-logic over verisimilitude in the ensuing film. None of these meanings are exclusive and I would argue that each is equally valid, whether or not it was intended by Buñuel and Dalí. This is reinforced by the fact that the Surrealists sought to create ‘multi-layered realities’3 in their works. The critical articles they wrote on cinema often took the form of ‘parallel readings of individual films, clearly reflecting [their] belief in the essentially creative role of the spectator in negotiating the film’s meanings.’4
This is reminiscent of André Breton’s accounts of his behaviour while at the ‘cinema age’. The so-called ‘Pope of Surrealism’ wrote how he would arrive at the cinema without having consulted the amusement pages and would watch whatever was playing. He would leave at the first hint of boredom and behave in the same way at another cinema. Linda Williams has interpreted this practice as an attempt to ‘defeat the passivity inherent in the filmgoing experience.’5 This has the effect of showing the fictional world of the film to be an illusion, thus forbidding the audience to believe in it.
This effect was achieved by Un chien andalou. Films usually endeavour to create verisimilitude through ‘transparent’ editing. Cuts between action aim to be seamless so that events are read as occurring within the same space and time. Essential to this is continuity, which, if disrupted, may reveal scenes to be filmed out of order, or spatial proximity, or both. Un chien andalou deliberately subverts these conventions, making it impossible for the audience to believe in the fiction. For example, the prologue features Luís Buñuel sharpening a razor. He is wearing a striped shirt and a watch. The following shot is a close-up of a woman’s face. A man stands behind her, his face unseen. Due to the usual cinematic principle of continuity we see him as the same man, since he is wearing the same shirt. But he is now wearing a striped tie and no watch. The presence of the shirt and razor in both shots encourage us to read the action as continuous, but the other elements prove it cannot be.6
Similarly, a later scene shows the same woman (eye unharmed) leaving her Paris apartment. Stepping through the door she finds herself on a beach. These and numerous other examples banish continuity and verisimilitude from the film, making the audience aware of the medium’s illusionary nature. It becomes impossible to be a passive receptor in the manner of mainstream cinema. Un chien andalou therefore shares the Surrealist demand that audiences have an active role in the viewing of the film. Meaning is not obviously present to be received: it has to be negotiated, for which the audience must indeed take a ‘creative role’. This is verified by Buñuel’s hope that the film would ‘arouse the deepest impulses of the viewer’7 and be a ‘passionate call to murder.’8
Salvador Dali (right) as an unusual priest
Diametrically opposed to Durgnat’s belief in a coherent narrative is that of Un chien andalou as purely a succession of disjointed images. This is best represented by the film critics Bordwell and Thompson. ‘Surrealist cinema is overtly anti-narrative,’9 they argue. They go further to say that ‘many Surrealist films tease us to find a narrative logic that is simply absent . . . events [are] juxtaposed for their disturbing effect.’10
Un chien andalou contains many aspects that seem to verify these assertions. The film opens with a title card reading ‘Once upon a time’. This suggests that a narrative will follow, and even goes so far as to specify the type of narrative – that of the fairytale. Soon afterwards, however, we are presented with ‘Eight years later’ and ‘Sixteen years before’ – purely arbitrary notices since no corresponding changes in the characters are visible. The eye-cutting scene is usually interpreted as a prologue to the film – summarising its main themes but not strictly forming part of the narrative. This is reinforced by the fact that it is Buñuel himself who performs the act, thus making the scene appear to be a director’s comments on his work. But if this is so then the title ‘Once upon a time,’ which immediately precedes the scene, conflicts with its prologue-status by encouraging us to include it in the narrative. The scene is therefore not a prologue but part of the narrative. Since it contradicts the rest of the narrative however (the woman later reappears unharmed) what the film presents cannot be a ‘narrative’ at all. The title cards are therefore used ironically: purporting to facilitate the telling of a narrative, they actually emphasise its absence.
Clearly the film relies heavily upon dream logic. The full extent of this will be discussed later, but for now the oneiric quality is relevant for the unusual way in which it is presented. Dream sequences are common in films (Dalí himself created a memorable example for Hitchcock’s Spellbound). Typically these are signified by the use of a standard notation.11 A variety of methods may be employed to distort the image and thereby delineate the dream from the rest of the narrative. The Spellbound sequence, for example, uses Surrealistic settings, while Nosferatu uses negative photography. In signifying a digression from the narrative these techniques actually reinforce its logic. In these examples too the sequence is always motivated as a character’s dream or hallucination. Un chien andalou, however, uses no such techniques. It is never indicated that a character is hallucinating or that any event depicts subjective experience. It is as if the film itself is hallucinating or is an hallucination. But this quality resides in the bizarre succession of images rather than in individual ones. No single image is distorted: the oneiric quality is only evident sequentially.
This idea of a cinema that can ‘think’ has been identified by Linda Williams: ‘Instead of showing what a character thinks, the Surrealist tendency in film was to show how images themselves can ‘think’.’12 Perhaps the clearest example that can be used to verify this point is the brief sequence following the episode of the ants in the cyclist’s hand. The image dissolves to a shot of the underarm hair of a female bather, a sea urchin, and finally, the androgyne’s head. The sequence is a cycle that seems to take the action on a brief digression: the shots of the cyclist’s hand and of the androgyne occur within and just outside the woman’s apartment. The two shots are linked with a conventional eye-line match (the cyclist looks down into the street), indicating that they occur in the same space and time. But the other two shots both occur on a beach – a space external to the main front of action – and seem to bear no relation to the other shots beyond physical resemblance. It is as if the film observes the ants in the cyclist’s hand and begins listing similar shapes, with no regard to narrative continuity. Consequently, this episode is strong evidence for the idea that Un chien andalou is purely a succession of images. But the same sequence can be used to put forward the opposite viewpoint. The underarm hair reoccurs in a scene between the woman and the cyclist. Similarly, the sea urchin evokes the last scene on the beach. Both events foreshadow later ones, implying that the film possesses its own logic. Since the sequence can be applied to either argument its real significance lies in the fact that it re-affirms Williams’s point: the images can think, and do so independently of any narrative or character. The sequence, and the film’s succession of images in general, is not a juxtaposition of unrelated images but an independent ‘stream of consciousness’ or passage of ideas, which goes some way to verifying Morise’s statement.
Un chien andalou’s succession of images is intended to replicate, in Artaud’s phrase, ‘the mechanics of the dream.’13 The Surrealists loved cinema for its oneiric quality and its ability to undermine the distinction between dream and life. No art form had previously represented reality so convincingly, and it was this aspect that intrigued them, precisely because it allowed dreams and the unconscious to be portrayed with such tangibility. The notion that this realm of the psyche constituted a heightened or surreality was of course central to Surrealism.
The first indication of a ‘trompe l’oeil fixation on dreams’14 is given by the eye-cutting scene. This has been interpreted as a notification that verisimilitude has been abandoned in favour of the oneiric. What ensues are not visual images but those of a dream. It is the inner eye of the unconscious that is functioning.
What follows in the film is clearly influenced by Freud’s understanding of the unconscious. We are presented with images that function symbolically in the way that Freud outlined in The Interpretation of Dreams. Thus the severed hand in the street evokes a fear of castration, while the second male character, who berates the cyclist, is interpreted as the super-ego exerting self-censorship. As well as dream symbols, Un chien andalou uses Freudian mechanisms of concealment. Displacement, whereby obsessions are transferred from their original object to one that disguises their nature, was at work in this episode of the severed hand. Condensation involves the combination of several ideas into one figure. The woman may be cited as an example of this mechanism. Her first reaction to the cyclist is irritation and reluctant fascination. This occurs when she hears him passing below her window. She descends into the street and finds him lying unconscious. She then begins to kiss and nurture him, behaving in a maternal way as if he was a small child. Subsequently, she becomes the object of his desire. Likewise, the cyclist is present in several manifestations. Initially he is dressed in feminine frills and represents an emasculated identity. Alternatively he loses his frills and becomes a representation of aggressive masculinity. These examples demonstrate the sophistication of Buñuel and Dalí’s understanding of Freud beyond that of directors who simply appropriated certain dream symbols in their films
This identification and interpretation of dream symbols was the basis of Durgnat’s reading, which has already been shown to be limited. A reason for this is that he depends upon a misunderstanding of Freud. As Williams has written, ‘a common mistake in dream analysis [is] to assume that the unconscious is a pre-existing storehouse of meaning which the dream symbols simply represent.’15 Durgnat ignores the fact that neither cinematic nor dream images have codified meanings. His imposition of a narrative parallels this misreading of Freud that sees the dream as ‘the fixed symbolisation of certain unconscious thoughts rather than as the production and creation of meaning.’16 Dreams may have latent meanings, but none is a ‘pre-existing entity that can be reached through mechanistic decoding.’17 In the same way, Un chien andalou does not feature a latent narrative that can be interpreted immediately with the ease demonstrated by Durgnat, but a complex system of meanings, as already discussed. Therefore, Un chien andalou is closer to the actual ‘mechanics of the dream’ than Durgnat and others have estimated.
It now remains to examine the extent to which cinema is the ideal Surrealist medium with reference to Un chien andalou. Today, Surrealism is mainly seen as a phenomenon of the plastic arts. Surprisingly, however, it began as a literary movement (the first essay to mention painting was Breton’s ‘Surrealism and painting’). Firstly in the journal La Littérature and later in other journals and several manifestos, Breton and his associates developed a Surrealistic approach to writing based on automatism and notions of the subconscious. Increasingly, however, their poems came to resemble film scenarios and used a mode of expression that was primarily cinematic (eg. Soupault’s L’Indifférence). Some, including Artaud, suggested that writing was of limited use in Surrealism because language, with its grammatical codification, was fundamentally restrictive, and therefore ran contrary to Surrealism’s spirit of revolution and liberation. Cinema was free of these conventions, since it had not existed long enough for a cinematic language to be codified.
Painting was also faulted, as its medium is space and it could only represent static images and single instants.18 Cinema, however, possesses a time dimension, allowing it to present a succession of images that evoke the oneiric quality so central to Surrealism.
As Un chien andalou testifies, cinema had an unprecedented facility for incorporating each of Surrealism’s central aspects. The juxtaposition of opposites admired by Breton and influenced by Lautréamont’s famous image (‘beautiful as the fortuitous encounter upon a dissecting table of an umbrella and a sewing machine’)19 could be achieved in cinema just as well as in any other medium. But here, perhaps, it could possess a greater credibility and impact, due to cinema’s closeness to reality. Cinema was condemned by the establishment as immoral, and this would no doubt have emphasised its potential to perform a revolutionary role.20
In conclusion, I would argue that the film occupies a space between the pure juxtaposition of disjointed images and narrative coherence. Although by a certain reading a narrative can be extracted (or imposed) and it is tempting to do so, this is not the ideal way of understanding the film. It is composed of a succession of images, each of which has several meanings, and therefore no single narrative can be satisfactorily confirmed. Yet the images are not purely disjointed. Each has no meaning in itself (i.e. no single, definite meaning), the film therefore is not a juxtaposition of separate, independent images appearing in succession. Rather, it is a logical progression, since each image can only be understood when viewed in relation to all the others. This relationship amounts, not to a narrative, but to a multi-layered work open to a variety of interpretations. For this reason Un chien andalou repays repeated viewings.
1. Raymond Durgnat, Luís Buñuel p30
2. Ibid, p24
3. Wendy Everett, ‘Screen as Threshold’ p142
4. Ibid, p142
5. Linda Williams, Figures of Desire p46
6. Ibid. p68
7. David Bordwell / Kirstin Thompson, Film Art p456
8. Ian Walker, ‘Once upon a time’ p4
9. David Bordwell / Kirstin Thompson, Film Art p455
10. Ibid. p456
11. Linda Williams, Figures of Desire p43 / 48
12. Ibid. p51
13. Antonin Artaud, quoted in Linda Williams, Figures of Desire p54
14. André Breton quoted in Silvano Levy’s introduction by to Surrealism and Surrealist Visuality p7
15. Linda Williams, Figures of Desire p75
16. Ibid. p75
17. Ibid. p75
18. Summary of ideas expressed by Max Morise in ‘Les yeux enchantes’
19. Lautréamont, quoted in Linda Williams, Figures of Desire p74
20. Wendy Everett, ‘Screen as Threshold’ p141
Bordwell, David / Thompson, Kristin Film Art: An Introduction (MacGraw-Hill, fifth edition,1997)
Durgnat, Raymond – Luís Buñuel (Movie Magasine Limited, London, 1968)
Everett, Wendy – ‘Screen as Threshold’ in Screen volume 39, number 2, Summer 1998
Morise, Max – ‘Les yeux enchantes’ in the hand-out to Roger Scott’s lecture (28th November)
Short, Robert – ‘Magritte and the cinema’ in Levy, Silvano (ed.) – Surrealism and Surrealist Visuality (Keele University Press, Edinburgh, 1997) pp95-109
Walker, Ian – ‘Once upon a time’ in Monthly Film Bulletin, vol. 35, no. 419, Dec 1968, p 205
Williams, Linda – Figures of Desire: A Theory and Analysis of Surrealist Film (University of Illinois, Chicago, 1981)
Please see my related articles: