This appeared originally in Film Comment, July-August 1978, and was reprinted by Saul Symonds in March 2005, with separate new prefaces by myself (reproduced below) and David Ehrenstein, in the online Light Sleeper (which is no longer up, alas). It’s also reprinted in my recent book Cinematic Encounters: Interviews and Dialogues (2018). — J.R.
Obscure Objects of Desire: A Jam Session on Non-Narrative
By Raymond Durgnat, David Ehrenstein and Jonathan Rosenbaum
Preface by Jonathan Rosenbaum, March 2005:
When this piece was written, or more precisely assembled, over 30 years ago, Ray Durgnat and I were sharing a house in Del Mar, California with experimental filmmaker Louis Hock. Ray and Louis were teaching film in the Visual Arts Department at the University of California, San Diego, where I had taught the previous year — having been coaxed by Manny Farber into leaving my job as assistant editor of Monthly Film Bulletin and staff writer of Sight and Sound in London, at the British Film Institute, and returning to the U.S. after almost eight years of living in Europe. Ray, already a friend, also came over from London to take my position when I wasn’t rehired, and I was starting to work on a book that eventually became Moving Places: A Life at the Movies (1980). (For further details, see my autobiographical “They Drive by Night: The Criticism of Manny Farber” in my 1995 collection Placing Movies: The Practice of Film Criticism and also on this website.)
I was the instigator of this critical round-robin, and my original hope was to find a way of collaborating on a writing project with Manny Farber and Patricia Patterson as well as Ray — all three of whom, I thought, had an especially acute feeling for nonnarrative aspects of movies. When Manny and Patricia reported back that they weren’t interested, or didn’t have the time, I enlisted first Ray and then David Ehrenstein — an old friend living in Los Angeles who at this point had never met Ray, and didn’t in fact meet him until well after the piece was finished and published. The first paragraph was written by me; after that, if memory serves, Ray’s and my interventions were tape recorded, transcribed, and then rewritten before the results were mailed to David for interjections …. One reason why it seems worth reviving the piece now is that the subject has rarely been addressed in these terms in relation to cinema as a whole, especially in playful rather than academic terms.
Why talk about non-narrative?
JONATHAN ROSENBAUM: To broach a subject and isolate a problem that most film criticism represses, stumbles over, or refuses to acknowledge, preferring to stick to the authoritarian guidelines of the synopsis and the plot summary. “Telling a story” — the task that for many critics is the only game in town twenty- four hours a day, 365 days a year — becomes a singular grid through which all the diverse structures and operations of movies can theoretically be apprehended, codified, and converted into meanings. The implicit suggestion that nothing important can elude this structural model of plot — acting, editing, direction, theme, social relevance — becomes a self-serving prophecy that freezes film analysis into a monotonous treadmill of tautologies. The question is, does this correspond invariably to the way that nonspecialized viewers look at movies, or is it a model that exists in order to facilitate the critic’s work? The synopsis that is handed out at press screenings only to be regurgitated or adapted into reviews (and related marketing devices) clearly bypasses a complex set of experiences that every spectator has, but few are able to articulate in critical frameworks.
RAYMOND DURGNAT: If you ask most people to synopsize a film after they’ve seen it, they’re completely lost. It’s a skill that critics learn, often badly, over the years, and I very much doubt whether the ordinary member of the audience remembers the plot structure as a whole in the way in which criticism implies that it does. Half the people in the audience have as much a pictorial eye as a narrative eye.
JR: Peter Gidal addresses a related question when he complains about several critics interpreting even Michael Snow films in relation to narrative models. The problem is, Gidal’s definitions and descriptions of non-narrative structures are nearly all negative indications. What we need are some positive ones.
RD: If we made a checklist on non-narrative dimensions of film, most of them would be familiar. We know about visual composition, plastic values, editing structures, intellectual structures, music, and so on. The only disadvantage these non-narrative elements have is the difficulty of briefly verbalizing them. One can summarize a plot in one sentence, whereas it’s fairly difficult to summarize one frame.
DAVID EHRENSTEIN: That depends. The screen can be filled with enormous amounts of detail, as in Visconti, and about all it amounts to is “Street outside the house with people passing.” On the other hand, shots that contain only a single object can be impossible to summarize easily. As Christian Metz says, a shot of a gun can never be read by the spectator simply as “gun”; at the very least, the shot constitutes “Here is a gun.” And the reason for that is context. The shot comes from somewhere (even if it’s the first shot of the film), and is going somewhere else. Of course most of the time a shot of a gun — which is to say in most cases a hand holding a gun — is merely functional in the scene.
Sometimes shots like this can work in another manner. The little tin of something-or-other held up by a hand at the beginning of Rameau’s Nephew is impossible to pin down for a particular reading, because it appears only to vanish into nothing — it never returns again or is alluded to in any other context in the rest of the film. In other cases, such as Robbe-Grillet’s films, the backgrounds in which the objects are placed keep shifting and affecting meaning as a result. In Glissements Progressifs du Plaisir, a blue shoe and a broken wine glass move around so much that all potential meaning for them is abolished. Lubitsch does something similar in his films. Consider the purse in Trouble in Paradise: as it changes hands its meaning in the film changes. But the effect reached is toward a gradual accumulation of meaning with no end in sight.
JR: Obviously, part of the problem in summarizing individual frames is terminology, which feeds back into the question of context. How does one distinguish between action and event, for instance? Is it a matter of defining units?
RD: No, it’s a question of defining relations. Semiologists often assume that you start from units and then define the relationships between them by certain syntactical procedures. But the alternative approach, taken from structuralism in the life sciences, or Gestalt psychology, or many different positions in philosophy, is that units are only phases in structures. You don’t start from the unit, you start from the structure … You have to distinguish among movement, action, event, and narrative. They’re four completely different things, or rather the last three shade into each other. But time and time again people imply that movement entails narrative, making film a mainly narrative act. Yet movement often isn’t even action. All the leaves moving around on a tree don’t constitute as many narratives as there are leaves.
DE: It was really nice to read Gilbert Adair’s “Paris Journal” in the March-April Film Comment, where he mentions in passing his fascination with the color of Cary Grant’s socks in the crop-dusting sequence of North by Northwest. For what it proves isn’t that Adair is nodding at the switch — not paying attention during one of the most famous set pieces in movie history — but rather that he’s really on the ball. For what’s involved is his awareness of other elements within the film, and his willingness to deal with them, regardless of the fact that such dealing goes against the grain of the film’s affectivity.
The North by Northwest example is especially important, as it’s a moment of prime audience “involvement” when the economy of narrative articulation is functioning at its most ruthless pace. If films were only their narratives, only their stories, such things wouldn’t be possible. And the fact that they are possible isn’t something to be swept under the rug or ignored as irrelevant, but rather investigated in as thoroughgoing a manner as possible. For what’s at stake is the relative autonomy of the spectator on the one hand, and the actual process of his intellection of elements on the other — something that diegesis- bound semiotic studies (such as Stephen Heath’s on Touch of Evil) haven’t really been willing to come to grips with. Heath shows how everything can be returned or translated into the story. He notes a few loose ends or gray areas, but that’s all.
JR: At least he begins to acknowledge part of the problem when he quotes diverse synopses of Touch of Evil and demonstrates the inadequacy of each. From this point of view, my two and a half years on the staff of Monthly Film Bulletin, where every review is preceded by a detailed synopsis, was very instructive. Readers of the magazine who believe that it should stick only to credits and synopses, and omit the reviews, always seem to assume that the former represent data that are objective, while the latter are subjective. Yet from a material point of view, all three categories are subjective selections of data, funneled through specific channels that are anything but neutral. The point at which writing a synopsis became interesting was when I was reviewing Wavelength and used the form as a means of posing the question “What are the contents of this movie?” — a crucial question that has to be asked on some level before it can be reviewed coherently.
When the synopsis straightjacket became impossible, an actual form of censorship, was when I had to impose it on Dreyer’s Vampyr, a movie that can be read as a paraphrasable story only if one ignores all the syntactical contradictions in its construction. Thanks to my own subliterary training, this is an aspect of the film that I was literally unable to see (or hear) until I started examining the film closely. A plot summary in this case represents a refusal to look at the film closely, a smoothing over and flattening out of all the discontinuities that confuse the spectator’s unconscious bridgework with the actual substance of Vampyr.
There are many other narrative films in this category, films that can be synopsized only if one omits precisely what makes them interesting, reading a new text through the tinted lenses of older texts: Gertrud, Mizoguchi’s The Loyal 47 Ronin of the Genroku Era, Persona, Playtime, Une Simple Histoire. Doing a synopsis of Je t’aime, je t’aime means concentrating on the clumsy s-f framing device and trying to reconstruct an inner plot out of the scrambled and fragmented chronology — none of which comes within miles of dealing with what the film’s about, or what its deepest sources of pleasure are.
RD: I don’t think anybody would pretend that dealing with Dreyer in terms of narrative was adequate. Everybody has a sense that here we are with these visual chunks, the way chunks strike chunks or the texture of muscles in a face, or the way one face jabs at another — it doesn’t matter whether it’s narrative or not, what matters is description: face as landscape. Dreyer said it. If you approach film from painting, it’s very obvious that narrative may be only a thread in the tapestry. In his “Trilogy of Life”, Pasolini uses it like that.
DE: In The Arabian Nights, the Ninetto sequence is protracted for an awfully long time because Pasolini wants you to look at the actors’ faces. They’re beautiful and should be lingered over, and the only way to do this is to bring the narrative to a halt. Suddenly the film begins to “give” and you can begin to get into it in any number of ways.
This is part of the reason why much of the hysteria over point-of-view constructs (I’m thinking especially of Heath’s recent piece on Oshima’s In the Realm of the Senses) is misplaced — a busy surface texture automatically destroys any fixed narrative line. I’m also thinking in particular of two films I’ve seen recently: Wenders’ The American Friend and Chytilova’s The Apple Game. Both of them have simple “linear” stories, but are so rich in surface texture in regard to objects, shapes, colors, etc., that only an idiot could regard them as “fixed” or “absolute.” They’re incredibly volatile films, their simple narrative bases making them that much more subversive.
JR: Perhaps the narrative bias partially comes from the way everyone’s been trained by television, where pictorial elements are reduced to an iconography that’s mainly designed to further the story line.
RD: Certainly in U.S. TV. It’s much more verbal — like radio with pictures — where English TV is far more pictorial. It perpetuates the difference between the British documentarist and The March of Time. March of Time would write the commentary first and cut the pictures to the sentences. They even disallowed moving camera, because of the difficulties in cutting into a moving shot.
There’s really a traditional Hollywood style of filmmaking that’s based on the story, and a European style which is more discursive and atmospheric; you can see it in Renoir and Hitchcock. Even in apparently narrative film, the entire narrative exists largely to maneuver two or three scenes into position to maximum effect. The narrative really has only a framing function. It’s a static construction, essentially. The particular scenes which are being nurtured are actually functioning in a lyrical way. They correspond to a lyrical poem rather than a narrative poem. It’s true that action may go on in them, but nonetheless they’re a lyrical description of an overall — and in that sense, static — situation. Hitchcock said of his English films that the central idea was just to present a series of strong scenes, and never mind how the people got from A to B. When he got to America, he found everybody worrying whether the plot was plausible or not, so he had to change his method of constructing a movie.
JR: A good example of Hitchcock’s first manner would be the whole first part of Number Seventeen — even the opening shot, where the camera tracks in the same direction as the wind past a windswept tree to follow a rolling hat, which is pursued in turn by a man who catches it in front of an abandoned house, where the story begins. You get this kind of delirium of continuity throughout, and one made all the more interesting by the fact that Hitchcock was literally trying to put one over on the critics by running through a nonsensical plot as rapidly as possible — a joke that worked, by the way, as far as contemporary reviewers were concerned. A women who’s introduced in the film as deaf and dumb suddenly starts to speak, with no explanation offered; a jewel thief sneezes at a particularly tense juncture for no apparent reason.
This kind of Lewis Carroll logic gives the film as much coherence in a way as something like Marnie, where all the motives get spelled out. It’s like an intensification of the European atmospheric approach that Ray refers to above, pushed to the point where the putative story line start to crumble, and other kinds of continuity come in to replace the gaps so that it’s hardly missed. A similar process is at work in Renoir’s La Nuit du Carrefour and Hawks’ The Big Sleep, and Rivette is clearly playing with the same principles even more systematically in Out 1: Spectre and Celine and Julie Go Boating.
DE: At this point, enter stage right the surrealists, whose love of cinema was based on the notion of narrative aphasia. For example, the intertitle in Nosferatu — “And when he crossed the bridge the shadows came to meet him” — cited by André Breton as having extraordinary poetic power is simply functional on a narrative level. But for Breton’s purpose it’s become something else. It’s as if he simply walked in the theater at the moment when the title flashed on the screen, read it, and left. Which is supposedly what many of the surrealists did in regard to films — regarding the part at the expense of the whole, finding it more interesting.
Some of this sort of practice can be seen in the films of Joseph Cornell, and the films and writings of Jack Smith. In Cornell’s Rose Hobart, a perfectly ordinary Thirties adventure melodrama called East of Borneo is taken and cut up into tiny bit and pieces in sequence, all of Cornell’s own devising. So we see the actress entering scenes, sitting down, turning to look at something that we don’t see, playing with a monkey in a garden, etc. The narrative is shattered and we end up in a sort of labyrinth of narrative possibilities or poses, similar to Marienbad or India Song of so many years later. Interestingly, in contrast to Rose Hobart, Bruce Conner’s A Movie — also made from other films — has a strict sense of narration and order (an evocation of dread and loathing in relation to the spectacularization of death). Even though it’s made of disparate pieces it forms a whole, whereas Rose Hobart, forged from a single unit, breaks into bits and pieces.
In Smith’s films, Flaming Creatures, Normal Love, and No President, there’s nothing but visual spectacle, completely at the expense of narration. He was very insistent on the spectator’s awareness of non-narrative elements, to the point of championing the Maria Montez films’ childish, simple-minded stories which made the visual-spectacle aspects all the more important — the aspects that audiences at the time were probably dealing with the most if they went to see these films, whether they were aware of it or not.
RD: Or let’s consider Vigo, times three. I don’t think one could describe A Propos de Nice as narrative. In Zero de Conduite, there is a narrative, and it does make sense, but nonetheless it’s just a broad structure, or rather just a drift, and the meat is the heavy atmospherics wound around and around the plot, at right angles to it. L’Atalante is the most conventional, but its atmospheres, not its story, are what make it Vigo. In the end one can’t define auteurs by their plot patterns, but predominately, or only, by atmospheres generated by style. That’s what Cahierism was supposed to be about, until the “literary” Anglo-Saxon auteurists and the verbocentric structuro-semiologists reverted to “literary elements.” Scorpio Rising is as far from narrative as A Propos de Nice. There are overtones of “coming together” but in the end it’s not even clear which of the cyclists is killed.
Un Chien Andalou raises the same questions. Is the prologue (the eye-cutting) part of the same story as the rest? It promises to be insofar as they share a face; but they don’t share Buñuel; and the intertitle asserting a chronological link is crazy enough to be ironical. Spectators assume the rest is a narrative whole, but it’s only an assumption. The segments are related, but relationship isn’t a synonym for narrative. And after all, as soon as we talk about “theme” we’re accepting that the narrative itself is subordinated to a non-narrative relationship, a structure of ideas, or a super-entity. As when one treats a film as an auteur’s psychodrama: chronology is subordinate to simultaneity, it’s a dream-metaphor for structure. Even more clearly when one thinks of the audience’s or the culture’s psychodrama. As soon as you treat a story as symptoms of somebody’s existing Oedipus complex, you’re denarrativizing the narrative.
JR: What about the notion that narrative itself is Oedipal?
RD: Edible?
DE: Yeah, like Snow says: eating is believing.
JR: Oedipal — Barthes’ idea in The Pleasure of the Text (which I suppose is edible, too) of the narrative as a search for the solution to the riddle, the identity of the father, the possession of the mother, the binding structure that Deleuze and Guattari imply we should start trying to live in spite of. The guiltless pursuit of desire that they espouse is like a sensual dive into non-narrative, a constellation of wants that shine in different directions, make individual demands, create a stampede rather than get herded together on one cattle car that is headed for the slaughterhouse. That may sound like a brutal metaphor for narrative, but the functional structure of the cattle car and the story in relation to the occupant (whether it’s a character or a spectator) is quite similar. Like the relation of diamond and crystal ball to family plot in Hitchcock’s last film: each is a container of narrative, and each narrative points ultimately toward death (all stories have an end).
Anti-narrative, like anti-Oedipus, suggests a way to keep things going, a polymorphous perversity that grabs at Cary Grant’s socks or Celine and Julie’s toys, also grasping, say, the sensual impact of a soundtrack apart from its narrative function, or succumbing to the lure of an extra or a prop. The collapse of narrative in Gravity’s Rainbow is literally an anthology of methods for breaking out of the rocket’s trajectory, or the trajectory of the narrative, each bound for the same fatal destination.
RD: Many of the important structures of ideas around narrative are non-narrative. And if you’re looking to narrative to enjoy the atmospheres of the moment as it goes, then you’re not looking for mystery. The whole pro- or anti-Oedipus argument is very outdated. After all, the Kleinians have loomed very large in psychoanalysis for over twenty years, nearer thirty. They attribute paranoia to pre-Oedipal projections onto “the bad breast.” “The other” is in the mother long before the Oedipal father appears on the scene. It’s strange that feminists have made so little of this approach. Maybe a richly atmospheric narrative is a good feed at the breast. Many narratives involve little or no suspense, certainly not of the Oedipal type. No doubt, as Kleinians would say, a good feed quietens fantasies about “the bad breast.”
Narrative is run largely by the laws of music. When Truffaut compares a film to a circus in which there’s a sequence of contrasting moods, he’s absolutely right. In 42nd Street each of the Busby Berkeley numbers is an elongation of a static situation. They’re states lyricized rather than sequences of actions decisively changing states. Many writers, especially poets, begin with a feeling — what they’ve got to put before you is a state of mind, which in itself is complex and simultaneous, a vertical structure but in a sequential order. Verse form often functions as a kind of binding over the sequentiality, by regular repetition and rhythm.
JR: Like the action in The Blood of a Poet, all of which is supposed to take place during the crumbling of a chimney. It’s a kind of bracketing of narrative that Buñuel also resorts to. In The Discreet Charm of the Bourgeoisie you have nothing but one story after another, but because of the way that they’re bracketed, there are different ways you can read them, different ways that they dovetail into one another — become or don’t become part of each other — and consequently there’s a disintegration of a single continuous thread. Theoretically you could say that this process takes place at one point or another in every film, whenever one’s attention becomes discontinuous. The synopsis is therefore not only a life preserver for the less courageous, but a Platonic model of the way we’re “supposed” to read films, of films as they’re “supposed” to be read.
RD: Narrative as such has no way of handling the presence of two Conchitas in That Obscure Object of Desire. And they’re not just two facets of one woman. They’re different women. Differences of physique become differences of style, differences of style become class differences. The brunette is classically proletarian, the blonde is characteristically petit bourgeois. Then think how Buñuel doubles the framing narrative, for no narrative reason…
JR: What Nashville seems to promise but doesn’t deliver is a proliferation of narrative that deflects away from any center and spins out in all directions. What you get instead is a drawing together, an interweaving of all the strands into that phony climax and god-awful flag filling up the screen — multiple narratives boiled together into one stew that yields a single corny premise.
RD: I like Nashville in the sense that it’s like an Ealing Studios omnibus picture with parallel plots. It’s rather like Derby Day or Train of Events. The pacing of the plot is always lost amongst a swarm of conspicuous details that have to do with the kind of physical integrity of the scene, or act like digressions. I find that very interesting in all of Altman.
JR: It sounds like a description of Renoir.
RD: Yes, there are Renoirian sides to Altman. It’s almost as if the physical details on the screen become a rhythmic structure, and the plot disappears into the rhythmic pacing of the details, as in a mosaic. Finally temps morts are becoming important in Hollywood, which has finally picked up on neo-realism, and picked up on this interesting nitty-gritty mosaic of detail. Citizens Band is another Ealing Studios picture with parallel plots; and it’s got an Altmanesque flatness, for the same reason that the English cinema emphasizes supporting actors more than stars, so it used parallel plots and omnibus forms rather than a heavy foregrounding of a one-line hero.
DE: But parallel plots don’t stop the flow in order to bring in other things; they just add another flow. And both can be eroded from the inside out. In Not Reconciled, Straub is dealing with a family-through-the-years story — or as Stephen Heath via Sigmund Freud would have it, “Family Romance.” We’ve see it time and again in films like Giant and Since You Went Away and Imitation of Life. But here we have material that would make a solid “engrossing” two-hour film, reduced to fifty-five minutes! The “plot” is impossible to make sense of — even on repeated viewings. What you do get is a chance to deal with the gaps in between — narrative “overtone” so to speak.
Getting back to what Ray said about the New Hollywood, it’s true about the temps morts, but how morts can they be with Jack Nicholson or Robert DeNiro up there emoting? Where they’ve really begun to make some changes is in the way that all the niceties of character development and carefully controlled exposition have been thrown out for the sake of bringing on more and more spectacular effects. I noticed this in Close Encounters of the Third Kind: you never get to “know” any of the characters in the manner of commercial films of the past. But where it really shows most is in The Fury: nothing but action set pieces, with little or no connection between one bit and the next.
This could be the wave of the future. We’re talking about something totally outside of aesthetics (though it does have aesthetic consequences: DePalma may be Hollywood’s answer to Michael Snow). It’s just that the smart money is beginning to realize it doesn’t have to lie to itself about story values in terms of reaching an audience. Just give them one damn thing after another — they won’t be bored.
© Copyright Raymond Durgnat Estate, David Ehrenstein and Jonathan Rosenbaum 1978. No part of this article may be reprinted without permission of the authors
—Film Comment, July-August 1978