The Decompositional Image

07/02/2012

Collapsed Dichotomies

It is a provocative claim, but has narrative cinema to some degree collapsed the difference between montage and mise-en-scene; have many films of the last few years decided that there is no longer a clear mise-en-scene into which one cuts, but an open space that can be cut into at will? The films one thinks of here include work by the Dardennes, certainly Dogme outings like Festen and The Idiots, and also odd, isolated parts of films like Bertolucci’s Besieged, Lynne Ramsay’s Morvern Callar, and, yes, perhaps even The Blair Witch Project. Let us say this is a variation on Richard Combs and Raymond Durgnat’s claims in an article in Film Comment that certain films are offering a steady disintegration of the image: “a collapse into surface, grain, texture, a rawness of technique that crosses certain avant-garde practices but is not part of them.” However, our variation rests in thinking more in terms of the viewer’s expectations in relation to the notion of following the image rather than perceiving it. Combs and Durgnat invoke similarities between dogme style films and the avant-garde in relation to the texture of the works, to the way they resemble, if you like, abstract art pieces that are studied not for their textual significance but their textural quality. It is a line of criticism we can trace back to Jean Louis Schefer, whom Gilles Deleuze invokes, and the comments Deleuze makes on Philippe Garrel’s early films, Kurosawa’s Macbeth, Welles’ Macbeth and other films in Cinema 2: The Time Image. In this approach, according to Schefer, “it is the suspension of the world, rather than movement, which gives the visible to thought.” But let us turn this inside out and explore instead how movement gives thought visibility, for many of the films here are nothing if not films of movement.

To clarify a little it might be useful to say which contemporary films we are not talking about, which films we feel still very much respect the distinction between editing and mise-en-scene. There are the long-take masters like Theo Angelopoulos and Bela Tarr, whose aesthetic still fit into a filmic modernism where style proves vital, the way actors are expected to meet their mark in relation to lighting and mise-en-scene. It is an approach von Trier very much respected earlier in his career, in The Element of Crime and Europa, and where he says of his recent work, “I have become more interested in people. It’s as simple as that.” ( Film Comment). “When we improvise, it’s because I want to take some of the qualities that are there already and use them, instead of making an actor stand in a corner, count to four, take one step to the left etc. which was the way I started.”

But just as the actors are freed up by such a style, then so is the director. And this is what we mean when we talk of the collapse of the distinction between mise-en-scene and editing. It is as though the filmmaker does not expect us to perceive the image ‘normally’, which we might assume in a conventional film, nor ‘abnormally’, in the avant-garde disintegration of the image, but bewilderingly, as we never know whether the filmmaker will cut and change the filmic environment, or will hold the shot and thus hold to the filmic space. To try and ground some of these comments it might be useful to look at specific films, at von Trier’s The Idiots, at the Dardennes’ Rosetta and The Child, at Ramsay’s Morvern Callar and Bertolucci’s Besieged.

In von Trier’s film the director is still loosely interested in what we can call the observational shot as opposed to the illustrative shot. Now generally we can see the observational shot, for example, in film movements as separated in time and space as neo-realism and Iranian cinema, just as in the illustrative shot we have anything from Hitchcock to James Bond films. One of the great defenders of the observational shot was of course Andre Bazin. And maybe never more so than in his article on Umberto D in What is Cinema? (Vol 2) where he believed “the narrative unit is not the episode, the event, the sudden turn of events, or the characters of its protagonists; it is the succession of concrete instants of life, no one of which can be said to be more important than another, for their ontological equality destroys drama at its very basis.” Bazin goes on to give the example from the film of the maid getting up and making coffee, which he sees as the opposite of the art of ellipsis – the art, if you like of editing – for the art of mise-en-scene. But while von Trier may respect the observational image, he constantly displaces us, as he forces us to re-examine our observational stance. This is of course an aspect we find in Godard, where in Breathless the director jump cuts within often observational scenes of the central character Michel and his girlfriend early in the film, but Godard’s device is both loosely Brechtian and at the same time curiously imitative of Michel’s perspective: he jump cuts as if out of an aesthetic nonchalance on his part, and a social disregard on Michel’s. We could say Godard pushes the idea of identification into the technical, but at the same time this has little to do with the way we usually assume identification and technique is formulated.

Under the notion of suture the viewer is stitched into the film by the editing and a clear spatial laying out of the scene. Obviously the jump cut obliterates the latter but, in this instance, it nevertheless re-energizes the former. The insouciance of Michel is matched by the insouciance of the filmmaker, so just as Michel believes he’s living in a world that defies the laws of the nation, so Godard disobeys the laws of cinematic technique.  Godard generally moved more and more into the Brechtian and further away from identification. If we can say the Dardennes, von Trier, and the Bertolucci of Beseiged and others have taken something from Godard’s radical collapse of montage and mise-en-scene, they have done so by holding onto a strong sense of subjective camera. Yet it is a subjective camera that has little to do with Pasolini’s free indirect subjectivity, an idea that still suggests filmmakers respect the mise-en scene/montage separation.  What Pasolini proposes in the essay a ‘Cinema of Poetry’ is that in cinema, generally, “direct discourse corresponds to the point of view shot”, but in free indirect discourse the point of view shot is eschewed and the filmmaker takes on the perspective of the character through the entire filmic vocabulary, but removes the conventionality of that vocabulary. Thus the director can conjure up “acts of monstrous subjectivity”, and yet an example of this is Antonioni’s  “technique of making the characters enter and leave the frame, as a result of which, in an occasionally obsessive manner, the editing comes to consist of a series of “pictures” – which we can call informal – where the characters enter, say or do something, and then go out, leaving the picture once again to its pure, absolute significance as picture.”

But what we find in the recent films is that if anything the pure absolute significance would be the actor and the story. In this sense it is the opposite of Pasolini’s comment on Antonioni. When von Trier says “on this film [Dancer in the Dark] and the last one [The Idiots] I operate the camera myself”, he makes clear the importance of the actor in the shot, or more especially the shot as a piece of observation, a kind of radical perception that goes further technically than the observational shot. If we take for example the early scenes in Bertolucci’s Besieged, we see Thandie Newton’s leading character passing through the streets of an African town on a bike. But the camera seems neither to be first and foremost to be following Newton, nor first and foremost to be developing a story. Sure, the narrative shortly afterwards develops, with soldiers carting off a schoolteacher moments later, but in this early scene Bertolucci explores the freedom of not so much the radical subjectivity proposed by Pasolini, but instead  ‘radical observation.’ In Besieged, inThe Idiots, in Morvern Callar, we have almost monstrous observation, to paraphrase Pasolini, serving a Bazinian sense of attending to the moment. But the notion of observation no longer lies in the deep focus shot, where the story and character function in the foreground, with the background detail as verisimilitudinous as possible, but that the detail and observation are no longer canny, no longer there to be spotted in the background by the knowing viewer, but contains the uncanny, a sense of perverse observational focus that can throw the viewer epistemologically because we do not expect the detail to be of any significance.

From this perspective editing no longer tells but much more shows, but what it shows cannot be determined by narrative specifics or directorial flourish. If in a film like Night of the Hunter we have a close up of Shelley Winter’s hand hestitantly opening her bedroom door as she decides to go in and see the sleeping Robert Mitchum in her room, thus serving wonderfully well narrative specifics, in Taxi Driver we have the close-up as flourish.  Manny Farber in Negative Space, for example, wonders “was that overhead shot, of De Niro’s hand and arm sweeping over the blond campaigner’s cluttered desk, shot at the same time as their conversation?”, and adds,  “The amount of twisting questions that are thrown at a spectator highlights its director’s boldness on intricate visuals.” But the questions thrown are questions of virtuosity more than uncanniness. Even the Antonioni-esque shot of Bickle on the phone to Betsy as the camera retreats into an empty corridor doesn’t quite suggest the uncanniness we are talking of. It may still pass for an unmotivated shot, but it is still seems consistent with the imposition of style, not quite a virtuoso shot, a type of shot we find at the beginning of De Palma films, Snake Eyes and Mission to Mars, Scorsese’s snaking steadicam into the nightclub in Good Fellas, the beginning of P. T. Anderson’s Boogie Nights and Altman’s The Player, but a shot of style nevertheless.

Now we are maybe in a better position to define a number of different types of shot as we move towards describing what we mean by the collapse of mise-en-scene into montage. We have the Bazinian long-take, where the purpose is to allow the viewer an observational relationship with the character and the milieu, which will be minimally interrupted by editing. Then we have a disruptive style of free indirect subjectivity where the director and the neurotic character aren’t easily extricated. Here the director doesn’t seek unobtrusiveness as a goal, but obtrusiveness: this is the antithesis of Ken Loach’s idea that if he could film without a camera he would; Pasolini suggests the camera should be felt. Then we have Schefer’s idea that is in some ways similar to Combs and Durgnat’s – the way the image becomes abstract because of its disintegration, the way images becomes less about the clarity of the scene, than the fuzzy incomprehensibility of the image. This tending towards decomposition hints at film’s capacity for abstraction without becoming abstract cinema. In this area we might think of the way numerous films in recent years have moved towards increasingly low contrasts: films like Pola XLa vie nouvelle and The Time of the Wolf, as well as the Dogme films the writers invoke. Then we have the virtuoso films of De Palma, Altman and Scorsese, where to some degree the camera is felt, but not in an indeterminate way suggesting the psyches of confused characters, but much more the assertive confidence of the filmmakers.

Now these are of course just a small handful of the many filmic approaches available to the filmmaker, and even our own apocalyptic notion of the collapse of mise-en-scene into an arbitrary montage is loose, suggestive and merely a means by which to understand how readily some filmmakers are forcing us to view things differently, are forcing upon us a radical sense of observation. But if Dogme films and films of their ilk are fresh, as Durgant and Combs suggest, then they’re not only fresh because of visual decomposition, but also because of their epistemological troublesomeness. Now a filmmaker like Alfred Hitchcock is epistemologically conventional, in the best sense of the term. He wants us to follow the images he chooses, and he often expects us to be simultaneously ahead of the character but behind the filmmaker. Any epistemological troublesomeness is narratively created, even if by the end of his most complex films, like Suspicion and Vertigo, we feel there is an unlocatable element that suggests any crisis presented is much deeper than any resolution offered.

But the epistemological troublesomeness present in a film such as The Idiots, or the Dardennes’ Rosetta, comes less from a narrative sense of what might happen next, than an imagistic sense of what might happen next. Where will the director cut, and when he does, what is the significance of the cut? It’s as though the meaning of the scene lies less in the accumulation of images, than the mise-en scene snatchingly filmed. We need only to compare a very fine, in many ways radical seventies film like The Mother and the Whore toThe Idiots to see that however respectful the director Jean Eustache is to the Paris milieu in which he sets his films, he still films in blocked segments. For example when the central character played by Jean-Pierre Leaud meets up with his ex, Gilbertine, in a Paris park, Eustache opens the scene with a long, frontal shot, establishing we are watching the couple, then cuts to a shot over Leaud’s shoulder as he talks. Then cuts back and forth thereafter until closing on a two shot from behind the park bench as the couple get up and leave. There are touches in this scene that are more than just conventional – the opening establishing shot is so long a shot that the two characters are small in the distance, and the camera holds for a second or two at the end of the scene after the characters have moved out of frame. But the cinematic vocabulary is still within the realm of blocked segments.

If we take the scene in The Idiots where the idiots of the title try and sell some thrown together, absurdly useless Christmas toys in an upmarket neighbourhood, it is as though the scene rather than the blocked segments dictate the flow. Von Trier neither establishes the scene, nor does he cut in relation to a character talking. He’s as likely to cut to a gesture as to a remark, and so we are forced into a slightly different epistemological relation even than that in which we find ourselves in so radical a film as The Mother and the Whore. It is as though in The Idiots the standard blocks of cinema, the mise-en-scene and the cut, are semi-obliterated by the scene and by the actor. Actors are no longer Hitckcockian cattle, with the scenes carefully storyboarded, nor are they just part of an elaborate mise-en-scene, as proposed by Delicatessen and City of the Lost Childrendirectors Jeunet and Caro when they say in Premiere magazine, “Our universe is rather bizarre and strange, everything in it – actors, decoration – has to fit.”. No, the actors become agents, almost agent provocateurs, as they radicalise the scenes they’re within rather than just play them. As actress, “the Deneuve of Dogme”, Paprika Steen says, in The Name of this Book is Dogme 95, Von Trier “takes a lot more chances, Lars, with his actors: he makes them go further out there, where it’s really embarrassing…” Bodil Jorgensen, the lead in The Idiots, says the man playing her husband she knew, and knew he had an aggressive streak. “I put the plate up in front of my face because I was so afraid of him. And the plate cut my eyebrow, and Lars shouted “Dogme blood!” So they filmed this trickle of blood down the side of my face.”

Again this isn’t ostensibly entirely new – we might think of Rivette’s extraction of extreme performances from Jean-Pierre Kalfon in L’Amour Fou and Out One: Spectre, and Philippe Garrel’s observational work in the seventies, especially with his own partner Nico, in films like Le Berceau de Cristal. But we do feel in The Idiots, and also a Dardennes film such as Rosetta, that the performance is no longer going on within the frame, but is dictating the very framing itself. When the Dardennes insist that they were working very much from the character of their title character, and the actress playing her, this creates in the viewer some initial confusion. As Michele Halberstadt says, interviewing the directors in Enthusiasm magazine, “from the first shot, from the first scene you leave us disorientated. We always have the feeling that you are only showing us part of the action, never all of it.”  But as Jean-Pierre Dardenne insists, the important thing is to “follow a character, as well as you can, without knowing everything that happens to him.” This is a clear illustration of the frame giving way to the performance. We also see it frequently in the Dardennes’ The Child. This may be through the way that Jeremie Renier looks to off-screen space that makes clear what else is going on in the mise-en-scene beyond the immediate frame, rather than the directors utilising an establishing shot and placing the actors within the mise-en-scene. Or it may come in the way that the Dardennes generate a form of chase sequence suspense without using the typical cross-cutting. Instead the directors hold close to Renier and his friend on their moped escaping after snatching someone’s bag, and allows us to see the danger they’re in not through a clear logistical cross-cutting that respects conventions of time (editing) and space (mise-en-scene), but through close attention to Renier and his friend’s immediate space, and how we can only see beyond that space what they themselves will see. The perspective is not so much the ‘privileged’ filmmakers’ view, so much as the unprivileged, fretting boys’.

In a key article in Cahiers du Cinema in the late sixties, Jacques Rivette, Sylvie Pierre and Jean Narboni proposed three “characteristic types of situation with regard to montage”. One was based on films dependent on “montage texture”, where the films were concerned with dialectical montage, or where the editing had a clear discursive purpose, examples given including Eisenstein’s work and Solanas’s Hour of the Furnaces. The second type were films that would not have so directly conspicuous a relationship with montage, but where thinking about montage is still vital. So the way Renoir and Mizoguchi eschew montage still suggests its importance even as it shows its relative absence. The third type consists of films where montage is used not to generate meaning but to block, to break down the mise-en-scene to the point that meaning cannot be readily discerned. The example given is Jean-Daniel Pollet and Mediterranee, but we might just as readily think of Carmelo Bene, with his rapid cutting, a cutting that wouldn’t create a socio-political meaning, à la Eisenstein, but would radicalise in meaning’s absence.

We could say that in the first and third examples given, montage is active, active in the sense that the viewer cannot but be confronted by the editing style. In the second it is much more likely to be passive – a viewer could easily watch a Renoir film and not think about the way the film was cut at all. But in the recent examples we’re giving, it’s as though the films are both active and passive, active in the sense that the editing owes little to conventionality, nor to long-take Bazinian invisibility, but that it is visible but not actually intrusive. For example students writing essays on The Idiots have talked about its use of long takes, and a friend commented on how many long-takes there seemed to be in Festen. Both films however usually use many, many cuts. This confusion, though, can come about because the films respect the scene, as we’ve suggested, more than the mise-en-scene – that the films are no longer working within the dichotomies of montage and mise-en-scene, but the scene itself, we feel, the staging of the scene, will be lengthy, perhaps even almost stage-like in terms of how much shooting time the actor has to get into the part. Then the editing will be merely the means by which the director will cut into the scene, as he protects the freedom of the performers, and the moment of the sequence. And what we look for in the scene, and in the cuts within the scene, is a radical observation because the filmmakers’ cutting, or for that matter non-cutting – it no longer matters as von Trier and the Dardennes reinvent Bazinian realism by other means – constantly keeps the viewer on the alert. Everything matters because nothing is given, as the viewer cannot anticipate the form of the film, but instead must try and anticipate much more the movements of those figures within the frame who have the opportunity to change the mise-en-scene based on their immediate behaviour. One might hear often enough of the star in classic Hollywood and beyond who would make all sorts of demands on the production as the film’s star, but is this not a much more modest and at the same time invigorating actor led cinema?

© Tony McKibbin

Tony McKibbin Tony McKibbin

The Decompositional Image

Collapsed Dichotomies

It is a provocative claim, but has narrative cinema to some degree collapsed the difference between montage and mise-en-scene; have many films of the last few years decided that there is no longer a clear mise-en-scene into which one cuts, but an open space that can be cut into at will? The films one thinks of here include work by the Dardennes, certainly Dogme outings like Festen and The Idiots, and also odd, isolated parts of films like Bertolucci's Besieged, Lynne Ramsay's Morvern Callar, and, yes, perhaps even The Blair Witch Project. Let us say this is a variation on Richard Combs and Raymond Durgnat's claims in an article in Film Comment that certain films are offering a steady disintegration of the image: "a collapse into surface, grain, texture, a rawness of technique that crosses certain avant-garde practices but is not part of them." However, our variation rests in thinking more in terms of the viewer's expectations in relation to the notion of following the image rather than perceiving it. Combs and Durgnat invoke similarities between dogme style films and the avant-garde in relation to the texture of the works, to the way they resemble, if you like, abstract art pieces that are studied not for their textual significance but their textural quality. It is a line of criticism we can trace back to Jean Louis Schefer, whom Gilles Deleuze invokes, and the comments Deleuze makes on Philippe Garrel's early films, Kurosawa's Macbeth, Welles' Macbeth and other films in Cinema 2: The Time Image. In this approach, according to Schefer, "it is the suspension of the world, rather than movement, which gives the visible to thought." But let us turn this inside out and explore instead how movement gives thought visibility, for many of the films here are nothing if not films of movement.

To clarify a little it might be useful to say which contemporary films we are not talking about, which films we feel still very much respect the distinction between editing and mise-en-scene. There are the long-take masters like Theo Angelopoulos and Bela Tarr, whose aesthetic still fit into a filmic modernism where style proves vital, the way actors are expected to meet their mark in relation to lighting and mise-en-scene. It is an approach von Trier very much respected earlier in his career, in The Element of Crime and Europa, and where he says of his recent work, "I have become more interested in people. It's as simple as that." ( Film Comment). "When we improvise, it's because I want to take some of the qualities that are there already and use them, instead of making an actor stand in a corner, count to four, take one step to the left etc. which was the way I started."

But just as the actors are freed up by such a style, then so is the director. And this is what we mean when we talk of the collapse of the distinction between mise-en-scene and editing. It is as though the filmmaker does not expect us to perceive the image 'normally', which we might assume in a conventional film, nor 'abnormally', in the avant-garde disintegration of the image, but bewilderingly, as we never know whether the filmmaker will cut and change the filmic environment, or will hold the shot and thus hold to the filmic space. To try and ground some of these comments it might be useful to look at specific films, at von Trier's The Idiots, at the Dardennes' Rosetta and The Child, at Ramsay's Morvern Callar and Bertolucci's Besieged.

In von Trier's film the director is still loosely interested in what we can call the observational shot as opposed to the illustrative shot. Now generally we can see the observational shot, for example, in film movements as separated in time and space as neo-realism and Iranian cinema, just as in the illustrative shot we have anything from Hitchcock to James Bond films. One of the great defenders of the observational shot was of course Andre Bazin. And maybe never more so than in his article on Umberto D in What is Cinema? (Vol 2) where he believed "the narrative unit is not the episode, the event, the sudden turn of events, or the characters of its protagonists; it is the succession of concrete instants of life, no one of which can be said to be more important than another, for their ontological equality destroys drama at its very basis." Bazin goes on to give the example from the film of the maid getting up and making coffee, which he sees as the opposite of the art of ellipsis - the art, if you like of editing - for the art of mise-en-scene. But while von Trier may respect the observational image, he constantly displaces us, as he forces us to re-examine our observational stance. This is of course an aspect we find in Godard, where in Breathless the director jump cuts within often observational scenes of the central character Michel and his girlfriend early in the film, but Godard's device is both loosely Brechtian and at the same time curiously imitative of Michel's perspective: he jump cuts as if out of an aesthetic nonchalance on his part, and a social disregard on Michel's. We could say Godard pushes the idea of identification into the technical, but at the same time this has little to do with the way we usually assume identification and technique is formulated.

Under the notion of suture the viewer is stitched into the film by the editing and a clear spatial laying out of the scene. Obviously the jump cut obliterates the latter but, in this instance, it nevertheless re-energizes the former. The insouciance of Michel is matched by the insouciance of the filmmaker, so just as Michel believes he's living in a world that defies the laws of the nation, so Godard disobeys the laws of cinematic technique. Godard generally moved more and more into the Brechtian and further away from identification. If we can say the Dardennes, von Trier, and the Bertolucci of Beseiged and others have taken something from Godard's radical collapse of montage and mise-en-scene, they have done so by holding onto a strong sense of subjective camera. Yet it is a subjective camera that has little to do with Pasolini's free indirect subjectivity, an idea that still suggests filmmakers respect the mise-en scene/montage separation. What Pasolini proposes in the essay a 'Cinema of Poetry' is that in cinema, generally, "direct discourse corresponds to the point of view shot", but in free indirect discourse the point of view shot is eschewed and the filmmaker takes on the perspective of the character through the entire filmic vocabulary, but removes the conventionality of that vocabulary. Thus the director can conjure up "acts of monstrous subjectivity", and yet an example of this is Antonioni's "technique of making the characters enter and leave the frame, as a result of which, in an occasionally obsessive manner, the editing comes to consist of a series of "pictures" - which we can call informal - where the characters enter, say or do something, and then go out, leaving the picture once again to its pure, absolute significance as picture."

But what we find in the recent films is that if anything the pure absolute significance would be the actor and the story. In this sense it is the opposite of Pasolini's comment on Antonioni. When von Trier says "on this film [Dancer in the Dark] and the last one [The Idiots] I operate the camera myself", he makes clear the importance of the actor in the shot, or more especially the shot as a piece of observation, a kind of radical perception that goes further technically than the observational shot. If we take for example the early scenes in Bertolucci's Besieged, we see Thandie Newton's leading character passing through the streets of an African town on a bike. But the camera seems neither to be first and foremost to be following Newton, nor first and foremost to be developing a story. Sure, the narrative shortly afterwards develops, with soldiers carting off a schoolteacher moments later, but in this early scene Bertolucci explores the freedom of not so much the radical subjectivity proposed by Pasolini, but instead 'radical observation.' In Besieged, inThe Idiots, in Morvern Callar, we have almost monstrous observation, to paraphrase Pasolini, serving a Bazinian sense of attending to the moment. But the notion of observation no longer lies in the deep focus shot, where the story and character function in the foreground, with the background detail as verisimilitudinous as possible, but that the detail and observation are no longer canny, no longer there to be spotted in the background by the knowing viewer, but contains the uncanny, a sense of perverse observational focus that can throw the viewer epistemologically because we do not expect the detail to be of any significance.

From this perspective editing no longer tells but much more shows, but what it shows cannot be determined by narrative specifics or directorial flourish. If in a film like Night of the Hunter we have a close up of Shelley Winter's hand hestitantly opening her bedroom door as she decides to go in and see the sleeping Robert Mitchum in her room, thus serving wonderfully well narrative specifics, in Taxi Driver we have the close-up as flourish. Manny Farber in Negative Space, for example, wonders "was that overhead shot, of De Niro's hand and arm sweeping over the blond campaigner's cluttered desk, shot at the same time as their conversation?", and adds, "The amount of twisting questions that are thrown at a spectator highlights its director's boldness on intricate visuals." But the questions thrown are questions of virtuosity more than uncanniness. Even the Antonioni-esque shot of Bickle on the phone to Betsy as the camera retreats into an empty corridor doesn't quite suggest the uncanniness we are talking of. It may still pass for an unmotivated shot, but it is still seems consistent with the imposition of style, not quite a virtuoso shot, a type of shot we find at the beginning of De Palma films, Snake Eyes and Mission to Mars, Scorsese's snaking steadicam into the nightclub in Good Fellas, the beginning of P. T. Anderson's Boogie Nights and Altman's The Player, but a shot of style nevertheless.

Now we are maybe in a better position to define a number of different types of shot as we move towards describing what we mean by the collapse of mise-en-scene into montage. We have the Bazinian long-take, where the purpose is to allow the viewer an observational relationship with the character and the milieu, which will be minimally interrupted by editing. Then we have a disruptive style of free indirect subjectivity where the director and the neurotic character aren't easily extricated. Here the director doesn't seek unobtrusiveness as a goal, but obtrusiveness: this is the antithesis of Ken Loach's idea that if he could film without a camera he would; Pasolini suggests the camera should be felt. Then we have Schefer's idea that is in some ways similar to Combs and Durgnat's - the way the image becomes abstract because of its disintegration, the way images becomes less about the clarity of the scene, than the fuzzy incomprehensibility of the image. This tending towards decomposition hints at film's capacity for abstraction without becoming abstract cinema. In this area we might think of the way numerous films in recent years have moved towards increasingly low contrasts: films like Pola X, La vie nouvelle and The Time of the Wolf, as well as the Dogme films the writers invoke. Then we have the virtuoso films of De Palma, Altman and Scorsese, where to some degree the camera is felt, but not in an indeterminate way suggesting the psyches of confused characters, but much more the assertive confidence of the filmmakers.

Now these are of course just a small handful of the many filmic approaches available to the filmmaker, and even our own apocalyptic notion of the collapse of mise-en-scene into an arbitrary montage is loose, suggestive and merely a means by which to understand how readily some filmmakers are forcing us to view things differently, are forcing upon us a radical sense of observation. But if Dogme films and films of their ilk are fresh, as Durgant and Combs suggest, then they're not only fresh because of visual decomposition, but also because of their epistemological troublesomeness. Now a filmmaker like Alfred Hitchcock is epistemologically conventional, in the best sense of the term. He wants us to follow the images he chooses, and he often expects us to be simultaneously ahead of the character but behind the filmmaker. Any epistemological troublesomeness is narratively created, even if by the end of his most complex films, like Suspicion and Vertigo, we feel there is an unlocatable element that suggests any crisis presented is much deeper than any resolution offered.

But the epistemological troublesomeness present in a film such as The Idiots, or the Dardennes' Rosetta, comes less from a narrative sense of what might happen next, than an imagistic sense of what might happen next. Where will the director cut, and when he does, what is the significance of the cut? It's as though the meaning of the scene lies less in the accumulation of images, than the mise-en scene snatchingly filmed. We need only to compare a very fine, in many ways radical seventies film like The Mother and the Whore toThe Idiots to see that however respectful the director Jean Eustache is to the Paris milieu in which he sets his films, he still films in blocked segments. For example when the central character played by Jean-Pierre Leaud meets up with his ex, Gilbertine, in a Paris park, Eustache opens the scene with a long, frontal shot, establishing we are watching the couple, then cuts to a shot over Leaud's shoulder as he talks. Then cuts back and forth thereafter until closing on a two shot from behind the park bench as the couple get up and leave. There are touches in this scene that are more than just conventional - the opening establishing shot is so long a shot that the two characters are small in the distance, and the camera holds for a second or two at the end of the scene after the characters have moved out of frame. But the cinematic vocabulary is still within the realm of blocked segments.

If we take the scene in The Idiots where the idiots of the title try and sell some thrown together, absurdly useless Christmas toys in an upmarket neighbourhood, it is as though the scene rather than the blocked segments dictate the flow. Von Trier neither establishes the scene, nor does he cut in relation to a character talking. He's as likely to cut to a gesture as to a remark, and so we are forced into a slightly different epistemological relation even than that in which we find ourselves in so radical a film as The Mother and the Whore. It is as though in The Idiots the standard blocks of cinema, the mise-en-scene and the cut, are semi-obliterated by the scene and by the actor. Actors are no longer Hitckcockian cattle, with the scenes carefully storyboarded, nor are they just part of an elaborate mise-en-scene, as proposed by Delicatessen and City of the Lost Childrendirectors Jeunet and Caro when they say in Premiere magazine, "Our universe is rather bizarre and strange, everything in it - actors, decoration - has to fit.". No, the actors become agents, almost agent provocateurs, as they radicalise the scenes they're within rather than just play them. As actress, "the Deneuve of Dogme", Paprika Steen says, in The Name of this Book is Dogme 95, Von Trier "takes a lot more chances, Lars, with his actors: he makes them go further out there, where it's really embarrassing..." Bodil Jorgensen, the lead in The Idiots, says the man playing her husband she knew, and knew he had an aggressive streak. "I put the plate up in front of my face because I was so afraid of him. And the plate cut my eyebrow, and Lars shouted "Dogme blood!" So they filmed this trickle of blood down the side of my face."

Again this isn't ostensibly entirely new - we might think of Rivette's extraction of extreme performances from Jean-Pierre Kalfon in L'Amour Fou and Out One: Spectre, and Philippe Garrel's observational work in the seventies, especially with his own partner Nico, in films like Le Berceau de Cristal. But we do feel in The Idiots, and also a Dardennes film such as Rosetta, that the performance is no longer going on within the frame, but is dictating the very framing itself. When the Dardennes insist that they were working very much from the character of their title character, and the actress playing her, this creates in the viewer some initial confusion. As Michele Halberstadt says, interviewing the directors in Enthusiasm magazine, "from the first shot, from the first scene you leave us disorientated. We always have the feeling that you are only showing us part of the action, never all of it." But as Jean-Pierre Dardenne insists, the important thing is to "follow a character, as well as you can, without knowing everything that happens to him." This is a clear illustration of the frame giving way to the performance. We also see it frequently in the Dardennes' The Child. This may be through the way that Jeremie Renier looks to off-screen space that makes clear what else is going on in the mise-en-scene beyond the immediate frame, rather than the directors utilising an establishing shot and placing the actors within the mise-en-scene. Or it may come in the way that the Dardennes generate a form of chase sequence suspense without using the typical cross-cutting. Instead the directors hold close to Renier and his friend on their moped escaping after snatching someone's bag, and allows us to see the danger they're in not through a clear logistical cross-cutting that respects conventions of time (editing) and space (mise-en-scene), but through close attention to Renier and his friend's immediate space, and how we can only see beyond that space what they themselves will see. The perspective is not so much the 'privileged' filmmakers' view, so much as the unprivileged, fretting boys'.

In a key article in Cahiers du Cinema in the late sixties, Jacques Rivette, Sylvie Pierre and Jean Narboni proposed three "characteristic types of situation with regard to montage". One was based on films dependent on "montage texture", where the films were concerned with dialectical montage, or where the editing had a clear discursive purpose, examples given including Eisenstein's work and Solanas's Hour of the Furnaces. The second type were films that would not have so directly conspicuous a relationship with montage, but where thinking about montage is still vital. So the way Renoir and Mizoguchi eschew montage still suggests its importance even as it shows its relative absence. The third type consists of films where montage is used not to generate meaning but to block, to break down the mise-en-scene to the point that meaning cannot be readily discerned. The example given is Jean-Daniel Pollet and Mediterranee, but we might just as readily think of Carmelo Bene, with his rapid cutting, a cutting that wouldn't create a socio-political meaning, la Eisenstein, but would radicalise in meaning's absence.

We could say that in the first and third examples given, montage is active, active in the sense that the viewer cannot but be confronted by the editing style. In the second it is much more likely to be passive - a viewer could easily watch a Renoir film and not think about the way the film was cut at all. But in the recent examples we're giving, it's as though the films are both active and passive, active in the sense that the editing owes little to conventionality, nor to long-take Bazinian invisibility, but that it is visible but not actually intrusive. For example students writing essays on The Idiots have talked about its use of long takes, and a friend commented on how many long-takes there seemed to be in Festen. Both films however usually use many, many cuts. This confusion, though, can come about because the films respect the scene, as we've suggested, more than the mise-en-scene - that the films are no longer working within the dichotomies of montage and mise-en-scene, but the scene itself, we feel, the staging of the scene, will be lengthy, perhaps even almost stage-like in terms of how much shooting time the actor has to get into the part. Then the editing will be merely the means by which the director will cut into the scene, as he protects the freedom of the performers, and the moment of the sequence. And what we look for in the scene, and in the cuts within the scene, is a radical observation because the filmmakers' cutting, or for that matter non-cutting - it no longer matters as von Trier and the Dardennes reinvent Bazinian realism by other means - constantly keeps the viewer on the alert. Everything matters because nothing is given, as the viewer cannot anticipate the form of the film, but instead must try and anticipate much more the movements of those figures within the frame who have the opportunity to change the mise-en-scene based on their immediate behaviour. One might hear often enough of the star in classic Hollywood and beyond who would make all sorts of demands on the production as the film's star, but is this not a much more modest and at the same time invigorating actor led cinema?


© Tony McKibbin