When thinking of the image, it is useful to observe how various filmmakers see it as a very different thing, evident in a number of comments offered in the book Film Forum, a discussion with thirty five filmmakers. Dusan Makajevev says, "What is important to understand is that in movies there is nothing else - whatever is in the frame is all there is." Claude Chabrol, though, thinks that "what is important in the frame is what you don't show". Are these contrary statements, or are they both saying the same thing in different ways: that the frame is what we see, and that is all we see at any given moment? As director and cinematographer Nestor Almendros reckons: "I think you do need the frame, otherwise, if there is no limits, there's no artistic transposition. I think the frame is a great discovery." Rainer Werner Fassbinder reckons, "It is very simple. I think the frame is like life. Life too offers only certain possibilities."
What we want to explore here are four of the possible options within the image. First, films that seem to make the frame a window onto the world, secondly those that see it as a frame containing the world to the detriment of conventional realism, third, films that use the frame chiefly to further the story, and, finally, films that use the shot as no more than an element within the editing schema. Of course such generalizations possess only so much validity, but they might usefully help us notice that while most directors appear to agree that the frame is all they have, they don't all see the frame possessing the same type of limitations.
Ken Loach would be nothing if not an exemplar of the realist frame, where in films like Kes, Raining Stones, Land and Freedom and My Name is Joe he wants to capture a plausible feel over a stylized effect. As he says, "making things look fake is the easy way. A technical demand like getting actors to hit their marks may not seem very important, but it can be a distraction, because no matter how skilled they are, it creates another level of consciousness in their minds, and I think that must crucially detract from their capacity to be totally involved." (Loach on Loach) If in the lengthy land rights discussion scene in Land and Freedom we feel we're observing and eavesdropping, it lies partly in the looseness of Loach's approach: it is as if he wants less to frame than to capture.
Other realists for the purposes of our argument would include Jean Renoir, Roberto Rossellini, and Robert Altman. Watching scenes from Renoir's The Rules of the Game we see how open its possibilities are. Characters constantly enter and exit, left and right, top and bottom. When Renoir talks of directors who fall into two categories he notes that there are those who put their "camera in a certain spot which is carefully chosen. It gives you a beautiful background; it gives you with the props a certain idea which can symbolically help the telling of the story; and then you take the actors and you put them in front of the lens, and you go on." "But I am the opposite. I like to start with the actors. I like to put them in a certain mood. We [the directors] are simply midwives". As he adds, "I always try not to cut during the shooting. It's why I use so often tracking shots, pans, etc. It is for no other reason than I hate to cut the acting of an actor during his inspiration." (Directing theFilm) If Renoir often wants an image dense with exits and entrances, this creates a slight contradiction. It contributes to a feeling of realism through the refusal to segment the film into chunks of narrative meaning, but might undermine realism by echoing the theatrical. Is it not in theatre that we so often have a sense of entrances and exits, of the stage as a limited space, but one utilising the wings?
Though Renoir and Rossellini are often talked about as filmmakers in the realist tradition, a clip from Voyage to Italy next to one from Rules of the Game indicates a few differences. In a scene in Rossellini's film where the husband meets his wife on the rooftop sunbathing, the camera tracks his movements as he comes and sits next to her. The film then settles for conventional shot/counter shot and two shots; but nevertheless plays up the idea that this is a found reality, a location more than a set. Renoir's house in The Rules of the Game maybe still seems more a set than a location, and that Renoir's fascination with actors expressed in his earlier remark doesn't, in this instance, quite stretch to a felt reality for locale. Rossellini, however, was often more interested in the location than in the plot, or for that matter the acting. As his leading lady Ingrid Bergman would say: "In Hollywood, I was accustomed to the scripts being prepared meticulously, well in advance. Every detail was set down, every shot, every angle, every camera movement was written down on paper....With Roberto it was more like a battlefield where only the general knows what the soldiers are supposed to do. (Ingrid Bergman: A Personal Biography) When asked by Eric Rohmer and Francois Truffaut in Cahiers du cinema whether he would work from any preconceived idea, Rossellini insisted: "None at all. I have no fixed plan. What I do have, rather, is a particular speed of observation, and I work according to what I see." In Voyage to Italy Naples becomes a character in the film, almost to the detriment of the actual characters, taking into account Bergman's remark.
The contrast here between Renoir and Rossellini is the difference between seeing the found reality as the actor in the first instance, the space in the second. Both filmmakers are seen as realist, but their realism takes on different forms. However, we should be careful about generalising too much; Rosselllini did after all make a series of films with Anna Magnani as well as Ingrid Bergman, with proper film actors who are important to the film, and we cannot easily separate films like Stromboli and Voyage to Italy from the beseeching and sometimes petulant tone Bergman offers in these performances. Equally Renoir's Indian set The River, a film Jacques Rivette mentions in the context of Voyage to Italy in an article in Cahiers, was admired for its sense of place, and Rivette himself would go on to become a filmmaker fascinated by both performance and filming real spaces in Paris Belongs to Us, L'amour Fou, Celine and Julie Go Boating, la belle noiseuse, VaSavoir. He seemed as influenced by Renoir as by Rossellini. What he added, though, in his own work, was the question of real time to found realities. Rivette's longest film runs to thirteen hours, and they are rarely much less than three. In La belle noiseuse the director adapts a story by Balzac about an artist and a model, and illustrates the hesitations, the progressions, the retreats through making not only the locations present, but also time felt.
This combination of the acting, or more precisely the behaviour of the characters, the interest in location, and the passage of time, are all elements central to notions of the realist image. Whether there is a concern for the frame or not, there is an interest in capturing over framing. As Robert Altman insists, "I don't feel frustrated by the frame at all. That's the medium I've chosen to work in." When he says he likes the anamorphic lens, a lens that would allow him to give the film "a sense of space; I wanted the audience to feel diminished" (Film Forum), equally, perhaps, it is to feel that the viewer can observe a wide range of information in front of their eyes.
Rivette and Altman are very different filmmakers, who are very different from Renoir and Rossellini, but for argument's sake let us say they share this interest in the frame as a realist device, in the 'realist image', and by way of contrast, we can look at other filmmakers who use the image as a formalist procedure, as a means by which the director paints with light and sees the actors and the setting, even the film's use of time, as a property not first and foremost of the world, but of the filmmaker who attempts to create an 'aesthetic image'. In this approach the film often doesn't so much imitate life or theatre, but much more painting, and numerous critics have noticed in, say, Michelangelo Antonioni's work, the influence of abstract expressionism, and in Jean-Luc Godard's pop art. As Rosalind Krauss says in The Optical Unconscious, Antonioni shared with abstract expressionism the foregrounding of the background over the foregrounded figure. In Lanotte there is a scene where Antonioni frames the building as huge next to the tiny figure of Jeanne Moreau. Yet the abstract expressionist influence is there also in The Red Desert. Here the walls that are spattered with paint suggest an action painting. Antonioni will also create framings that call attention to the scene by suggesting the split screen within the same shot, like a diptych. When Thomas takes photos in the park in Blow-Up, Antonioni's often empty frame calls attention to itself not only in the fact that this is a London park on a Saturday afternoon all but empty, but also in slicing the frame in two by putting a tree in the middle of the image. This approach forces upon us questions about the frame. If the great French film critic Andre Bazin thought the image was a window onto the world, Antonioni consistently blocks the view. Asked about the importance of painting on his films, Antonioni would say in an interview with none other than Godard: in The RedDesert "I feel the need to express reality in terms that are not completely realistic. The white abstract line that breaks into the shot of the little gray road interests me much more than the car which is coming towards us." It is a wonderful incapsulation of the difference between a realist who uses the frame to capture the world, and a director fascinated by the painterly possibilities in the frame. Indeed, we need only compare Rossellini's investigation of Naples in Voyage to Italy and Antonioni's exploration of Barcelona in ThePassenger to see the difference. Rossellini's camera shows Naples through the eyes of Ingrid Bergman's character; Antonioni searches for a cinematic space beyond the character's. As he sometimes frames the Gaudi architecture to shrink the figures in the frame, there is the feeling that the character is present as an aspect of scale more than a figure of focus.
When Godard asks Antonioni,"when you start or end a shot of an abstract shape, of an object or detail, do you do so in the same spirit as a painter?", he could have been asking a question to himself. As Richard Brody's biography of the director, Everything is Cinema, noted, quoting cinematographer Ed Lachmann: "Godard "told me that he wanted to have a director of photography who would work with him in a different way, who would be with him not just when the film was being shot, but all the time". Here Godard wanted a cameraman as a tool at his ready disposal, not to capture life on the hoof, but to capture the general perceptions of the filmmaker, with Lachmann the brush and oils. In a brilliant article by Alain Bergala, The Other Side of the Bouquet, Bergala talks about Godard's work in the eighties, and reckons: "by the time he began making Passion, just after taking stock of the field, Godard had come to think of himself, within the cinema surrounding him, as one of the last dinosaurs, a strange specimen, the sole survivor of his species...Thus, for Passion, Godard had to search outside cinema for masters to measure himself against, and for Great Classics to lean on and pervert."
Yet even in Godard's sixties work there was often the sense that the frame was a canvas upon which to paint the image. In films like Contempt, Pierrot le Fou and La ChinoiseGodard's use of primary colours shows the director to be a colourist, someone for whom, as he would say, "that's not blood; it's red". The apartment in Contempt is less realist space than colourist experiment, with Godard playing up the white walls and empty space against Brigitte Bardot's red bathrobe, the blue chairs, the red couch. The framing often isolates a moment so that it becomes a dwelt image, rather as we might pause over a painting in a gallery. There is the shot of the stairs by the beach with the characters tiny within the left hand side of the frame; the image of the rock framed to the left with the open sea on the right, another with Bardot lying in bed in the bottom third of the frame, an expanse of white wall and a door above and around her. There was also, of course, the influence of pop art, evident in Made in USA and La Chinoise.
A director unequivocally acknowledging the painterly is Peter Greenaway - a recent work called itself Nightwatching after Rembrandt's 'Night Watch' (a painting central also to Godard's Passion), and moves towards less narrative development than aesthetic composition. The coming together of 'Night Watch' is like hyperlink cinema as painterly revelation, with Greenaway gathering together the various people who will appear in the painting, a painting that may allegorically contain the key to a conspiracy. Similarly, A Draughtsman's Contract combines revelation within composition. There are all sorts of hints in the drawings the central character makes but that he is too blind to see, as Greenaway emphasizes the frame within the frame. Greenaway haughtily insists "I had for a long time resisted the idea of the classic European art movie, which, as I saw it then, was too much related to the business of writing literary scripts, processing narrative in predictable formulaic terms, narrowing down the filmic vocabulary, and obeying all the orthodox narrative verities, but I was persuaded, indeed challenged, to create a film world where the characters no longer talked directly into the microphone and the camera, as in the earlier film-essays, but to one other. The result was The Draughtsman's Contract." (Shooting Down Pictures) The film puts its mystery quite literally into the frame. It is as if Greenaway's career has been a certain type of draughtsman's contract of his own: insistently seeing film as a medium not of verisimilitude and with a respect for the real, but an art form giving him the opportunity to search out artifice in various manifestations. With numerous books insisting on the importance of the script as the movie's blueprint, Greenaway proposes it ought to lie elsewhere. Greenaway's work often points up the importance of the cartographic, and the sketched, as if screen space is to be mapped or coloured in rather than reality filmed. In The Cook, the Thief, his Wife and her LoverGreenaway offers a tracking shot, with Helen Mirren's costume changing as she moves from one room to the next, the costume matching the changing room colours, from white in the bathroom, to red in the hall.
Greenaway is a certain type of master, and yet the flipside of insisting that film has too often been at the service of the script, is to wonder why it might have been. Unlike painting, film is, if not necessarily a narrative form, then at least a temporal one: it is an act of defiance to walk out of a film, a casual and inevitable gesture to walk away from a painting. There is no designated amount of time by which we are expected to look at an art work; there is when it comes to watching a film. It is all very well radically changing cinema, but shouldn't one be aware of the ontological nature of the form at the same time (no matter if this is obviously changing with films in the gallery space)? Kafka once referred to cinema as putting one's eyes in uniform, and there is the danger that just as a filmmaker restricts the viewer with a formulaic story, can the uniform not be equally constraining if one feels time as a fundamental in cinema hasn't been acknowledged? It is a question filmmakers concerned more with the frame than the story might fail to address. If we feel Godard and Antonioni are masters, and Greenaway not, it lies in part in Godard and Antonioni's interest in the ontology of film, Greenaway's dismissive attitude towards certain key questions over its nature. As Zach Campbell says of Greenaway in ElusiveLucidity, "what bothers me about this running-off-at-the-mouth is that it poses as a high-cultural stance but it betrays a profound ignorance of the cinema..."
If there are those who insist on the importance of the frame to the detriment of the story, are there not numerous filmmakers who see the frame as no more than an element of the narrative; filmmakers who use the frame as a pragmatic shot choice to push the story forward and generate the 'constructive image'? As David Mamet says, ""The question is, where do I put the camera?" That's the simple question, and the answer is, "over there in that place in which it will capture the uninflected shot necessary to move the story along." Mamet does admit there are filmmakers with great visual acuity, but, since he happens not to be one of them, his take on cinema is to film the simplest possible way for the furtherance of the story. "If a student needs a retraction he will pursue a series of actions that will lead him to the retraction or to an irrevocable denial of the retraction. And then he will be at rest; a condition of entropy will have been achieved." For Mamet, any deviation is wasted footage. "Why do directors...shoot this many [up to sixty takes]? Because they don't know what they want to take a picture of?" (On Directing Film) They are shooting footage because they are unfocused, according to Mamet, and a filmmaker's purpose is to know what they need: the shots required to tell the story. Our purpose isn't to elevate or denigrate: we can all think of great examples of visual storytelling where the films move along quickly because of the director's genius for constructing the shots according to maximum dramatic impact. Hitchcock's cutting, panning and tracking in the Pyscho sequence, where Marion Crane decides to keep her boss's money rather than depositing it in the bank, is a great example, and the opening of Vertigo that establishes Scottie's fear of heights another. There might have been a lot of fat on Hitchcock, but rarely in his images.
When Mamet implies that filmmakers who shoot a lot of footage don't quite know what they're doing, Arthur Penn's approach indicates the opposite. When he says "I shoot at a very high ratio - almost indecently high" (Directing the Film), then nevertheless a sequence like the end of Bonnie and Clyde seems based on absolute economy. As the film shows Bonnie and Clyde shot down, the film eschews shots of the killers, and relies instead on putting us in the position the characters are in: a slightly paranoaic sense that something is untoward. The sequence could have been based on parallel montage, on shot/counter shot as we know that they are in danger from killers behind the bush, but instead Penn's economical editing makes us sense there is something wrong, but doesn't give us the knowing sense that the characters are about to die. Another fine example of editing with no waste is a brief moment from Michel Deville's Death in a French Garden. Here a son arrives at his father's house in a series of match cuts, with the sound of car doors, phones and apartment doors allowing for smooth, quick, transitions. In both Penn and Deville's films there is great visual economy.
Such examples still work within the great Russian filmmaker Pudovkin's notion of building the scene. When he says that "between the natural event and its appearance upon the screen there is a marked difference. It is exactly this difference that makes film an art" (Film Technique and Film Acting), he is also proposing that the filmmaker works from an established situation that is then edited into a new way. The scene itself is intact as an event; it just gets edited to make it more cinematic. Pudovkin gives an example of this from an American film Daddy, and a car crash in the movie, but the scene in the bar in his own Mother would work equally well. Here he sets up a situation that has been used over and over again, with at one moment someone looking as if they will start a fight, and Pudovkin cuts between the variables in the situation as the music stops and people look on. The scene still seems to us almost ninety years later as both perceptually conventional and very effective.
However, fellow Soviet filmmaker Sergei Eisenstein believed that films should be shot based not on rearranging the scene, but making the scene so dynamic that you can't easily put the bits together again. If Pudovkin believed that images should be bricks put together to build a powerful edifice, Eisenstein's buildings defy physical laws. If it takes the Tsarist troops so long to walk down the steps at Odessa in Battleship Potemkin, it isn't because there are so many; more that Eisenstein wants to play with space and time to emphasize the remorseless movement of the troops as they mow down the people. Equally, we may find ourselves wondering where certain people are within the scene in relation to others as the baby carriage falls down the steps. As the carriage is looked upon by other characters, we're not quite sure where they are looking on from. In a film based on the principles Pudovkin adhered to, we would know how near or how far people would be as the film would have established the scene, and re-established it for the purposes of ready cognition. Pudovkin might have admired this scene from Eisenstein's film, but it was more for its similarities with Pudovkin's method rather than its deviations from it.
Eisenstein's experiments with montage were taken still further in films of the sixties and seventies, by filmmakers like Alain Resnais, and also Nic Roeg, John Boorman, Joseph Losey and Richard Lester, and, in a different way by Robert Bresson, from the fifties onwards. They all searched out the 'reconstructive image'. Early on in Resnais's Hiroshima mon amour we might wonder where the voice-over is coming from as we hear a woman insisting that she had seen something in Hiroshima, and a man insisting she had seen nothing in Hiroshima. We have seen two bodies intermingling, and we may assume the voice-overs belong to these bodies, but at this moment we cannot say with certainty as the film then gives us images of what the woman says she has seen: Hiroshima hospitals and museums. If Eisenstein rearranged a scene in such a way that we couldn't easily put it back together again, then Resnais is still more ambitiously dislocating time and space so that we are wondering where the characters are, and who they might happen to be. If Eisenstein dissolves space for the purposes of creating a rhythmic force in cinema based on the collision of images that makes the event more immediate, Resnais' approach tends to make events more remote. The problem of time has entered the image so that there is no easy present tense from which we can work. He would take this further in his following film, Last Year at Marienbad, where the notion of this year and last year, and the reliability of characters' memories, disintegrates.
Nic Roeg and others would be indebted to such innovations. In Bad Timing, the film opens with a suicide attempt and the film in fragments puts together the story so that we can understand something of the motivations behind the act. But this isn't the sort of flashback we see in Double Indemnity, where the core of the film takes place in the past, voice-overed in the present, but a film that flashes back and forward, to understand also the motives of the lover who some might think drove her to the action, and a police detective who investigates. In John Boorman's Point Blank, the central character (Lee Marvin) is in bed with his ex's sister (Angie Dickinson), and in a series of shots, as they roll over in bed, we see other lovers coming into view: Marvin in bed with his ex; Dickinson in bed with a former friend of Marvin's. With each roll another couple formation appears.
In Lancelot du Lac and L'argent, Bresson in some ways returns to the argument between Eisenstein and Pudovkin, but it is though the problem passes through Hitchcock, and saying a bit more about Hitchcock will help us, perhaps paradoxically, to understand something of Bresson's brilliance. Critics have noted that Hitchcock is a great director of relations, brilliant at making the viewer infer an event without knowing the details that can make someone know for sure. Of course, detective thrillers have always been based on inferred knowledge as the detective finds out 'whodunit' beyond a reasonable doubt, but usually with reasonable doubts along the way. Hitchcock more than most, though, seemed to want the passivity of action in relation to the process of working out an inference, and this is maybe most evident in two of his 'experimental' films, Rope and Rear Window. In both films, James Stewart's character is a figure of inactive inference. In Rope, two of his ex-students have a body in a box in an apartment where various characters are having dinner. Will Stewart work out what has happened? The viewer already knows, but when will Stewart realize? In Rear Window we are no wiser than Jefferies, with Stewart on this occasion not only limited to the apartment (as in Rope), but further incapacitated by a broken leg that keeps him in his chair. In the film we piece together the story as he does, locked into his point of view as he makes inferences about events in the flats across the way. In other instances, as Gilles Deleuze notes in Cinema 1: The Movement Image, Hitchcock uses what Deleuze calls a demark: a detail that foreshadows and forces the viewer to realize something untoward - the windmill turning in the opposite direction to the wind in Foreign Correspondent, the key which doesn't fit the lock in Dial M for Murder, the luminous glass of milk in Suspicion.
Hitchcock uses these relations for narrative development and psychological cognition, but Bresson is interested in a very different form of inference. At the end of Lancelot du Lac, the battle scene is presented elliptically as we see arrows being shot, horses trotting through the woods without their riders, and the knights wandering around mortally wounded. Where in another film these moments would be presented as the arrow being released, the knight being hit, falling off the horse with the horse riding off without its rider, and the knight injured, they are offered here in a series of dislocated shots where cause and effect are jumbled up, but where we can still infer meaning. But the inferences deployed in this brief battle scene are closer to Eisenstein's event than Hitchcock's plot. We infer to generate meaning; not to second guess motive and action, as in the constructive image. Bresson pushes certain problems of inference much further than either Eisenstein or Hitchcock and arrives at the radically reconstructive.
What we have offered here is no more than an attempt at seeing the image and the frame in some of its various manifestations, to propose that though the frame is all there is, there are many ways in which to create images and put them together, and thus we have looked at four of these possibilities. If Rossellini, Renoir and others seek the realist image as window to the world, Antonioni, Godard and Greenaway often pursue the idea that the frame is a canvas more than a window. Others see it as a plot cell, a unit of information to be aligned with other units of information to create a very strong story with minimum visual fat, from Hitchcock to Arthur Penn. Eisenstein and Bresson are no less interested in the idea that the image is a component of the editing than Penn and Hitchcock (no matter Hitchcock's occasional experiments with the long take, most notably in Rope and UnderCapricorn), but utilise the image to defy our usual perception of physical properties rather than to accelerate them. Then we have directors like Resnais and Roeg, dislocating time and making us unsure where we are in terms of the time frame. The above is by no means exhaustive, of course, merely suggestive - a way into seeing the frame utilised in different ways, in seeing windows to the world as glass that needn't only be transparent but that can be opaque, kaleidoscopic and mosaic-like too.
© Tony McKibbin