Monday, 16 May 2011

Some thoughts on Gilles Deleuze

Some time in the late 1990s I flicked through the two cinema books by Gilles Deleuze, since he had become famous enough to reach my ears. I did however put the books back on their shelf not feeling they were useful for me. Now that I'm in St Andrews I'm reminded of him almost on a daily basis so I might as well write down some thoughts. Although I'm not a Deleuzian we do have some mutual interests.

Deleuze's writings are famously opaque, and he (and Felix Guattari) have an annoying tendency to use words that already exist but give them a new meaning, and therefore confuse the reader even more. It is tempting to ask why. Why not come up with a new word instead, or at least use a word closer to the actual meaning they are aiming for. It could of course be part of the philosophical project to re-distribute meaning (becoming-bewildered?), but it's still annoying. But since this is a film blog I will not critically engage with his politics or philosophy, tempted that I am, but concern myself with some aspects of his writings on cinema. I make no claim to say anything original though.

When people try to sum up the ideas of cinema, they usually say that Deleuze divides cinema into two categories, "movement-image" and "time-image", and that before second world war there was only the movement-image. Movement-image means films with a clear cause and effect structure, with time subservient to action and/or space. After the war the time-image came about, which less (or no) causality and where time in itself could become a subject, or at least that scenes or whole films were showing time unaltered, letting time flow at its own pace. Some, in order to simplify, say that movement-image equals mainstream cinema in general and classical Hollywood cinema in particular and time-image equals European art cinema. This is not how Deleuze saw it. Movement-image was not just Hollywood but also for example French impressionism, Soviet montage, German expressionism and the films of Carl Th. Dreyer. This is actually part of what makes him distinguished.

This is also one instance where Deleuze's construction of the history of cinema becomes puzzling. He argues that there was a crisis in the movement-image, partly due to the films of Alfred Hitchcock, and that Italian neorealism brought about the birth of the time-image. This was due to the destruction of Italian cities during the war, and also because the idea of the American dream was shattered, and, argues Deleuzian scholar D.N. Rodowick, a lack of trust in democracy. This feels somewhat roundabout and unsatisfactory. If Hitchcock was a cause, wouldn't the time-image have emerged in the US rather than Italy? Also, since Rear Window (1954) is used as a prime example of the crisis in the movement-image it would seem that this crisis came upon after the birth of the time-image, and consequently could not be the cause. About politics, why would the dream of America be shattered? They had just won the war and was arguably at the height of their global status, or does Deleuze mean that the Italians were dreaming of America during the war, and then stopped in the clear light of peace? Since Italy hadn't had democracy for some 25 years (and only partially before that) it does sound strange to say that neorealism came about because of a new lack of trust in democracy. What they lacked was not trust, what they lacked was democracy, which now came about after the war. Might it not be said that neorealism came about because of democracy and the American dream, rather than because of a disentanglement from it?

If we return to the neorealist films it doesn't get clearer because what Deleuze describes as the time-image in them is not really there. Take two early neorealist films, Rome Open City (Roma, città aparta 1945) and Shoeshine (Sciuscià 1946). They tell fast, straightforward, conventional stories (movement-image), with Shoeshine in particular similar in style to a lot of Warner Bros. films of the 1930s. That their relationship to time would differ from earlier mainstream cinema is at best dubious. A central neorealist film such as Bicycle Thieves (Ladri di biciclette 1948) also follows a clear, linear storyline, with straightforward cause and effect. It's a slightly melodramatic story using simple cinematic devices to guide to viewer through it, and uses music for emphasis. Man gets a job, man needs bicycle to keep job, bicycle gets stolen, man looks for bicycle with son, man finds thief, thief escapes, man and son becomes estranged from each other, man tries to steal bicycle, man gets caught trying to steal bicycle, man and son re-connect. It is again not clear what is new in relation to time, even though the film is slower and a bit more meandering than a film like Dead End (1937), or Rome, Open City for that matter. I'm not saying that Bicycle Thieves isn't different from a lot of Hollywood films, because it is, but that is not particularly interesting. If it was different from all films that came before, from all countries, then that would be interesting. Alas, I don't see it.

At one point Deleuze says about the crisis in the movement-image that: "[t]hese are the five characteristics of the new image: the dispersive situation, the deliberately weak links, the voyage form, the consciousness of clichés, the condemnation of the plot." and then he adds that "It was Italian neo-realism which forged the five preceding characteristics." But that is actually a pretty definitive description of Hitchcock's The 39 Steps (1935), with the exception of the dispersive situation, rather than neorealism. Besides, some films made in classical Hollywood, such as Portrait of Jennie (1948), have a much more radical relationship to time than most, if not all, neorealist films.

One of Deleuze's favourite examples of the time-image is Last Year in Marienbad (L'année dernière à Marienbad 1961) and here the differences between it and classical cinema are clear. There is no question about it having a totally different relationship to time, space and causality then, say His Girl Friday (1940). But again, what exactly is the significance of this fact? Can we say something about society in general based on this difference, or about cinema? The Bicycle Thieves and Last Year in Marienbad are also equally different from each other. If Bicycle Thieves is time-image, it would be tempting to describe Last Year in Marienbad as anti-time-image, since it almost obliterates time. It has so completely broken down the concepts of past, present and future that it might be said to take place out of time.

Deleuze also suggests that this new cinema of the time-image consists of "seers", people who do not act or react but just observe. This to is debatable. In the neorealist films mentioned above the characters are active, both taking charge and/or reacting constructively (more or less) to what is happening. Deleuze also uses as an example the boy in Germany Year Zero (Germania anno zero 1948). While it is true that the boy walks around a lot, just watching, he finally acts upon that which he has seen, killing first his father and then himself, and this has been clearly signalled all the way from the first scene (where two women talk about suicide), and enhanced by the dramatic music. In a sense he's not that different from many a hapless detective in film noir, or Ole in The Killers (1946). Besides, don't films in general often have a combination of "seers" and doers"? Like The Searchers (1956), where Ethan Edwards is an active agent, a "doer" and Martin Pawley a "seer", who also reports back, in letters to his adoptive family, what he has seen. In The Little Foxes (1941), Birdie is very much a "seer", or perhaps more correctly a "listener", whereas Regina Giddens is not (how could she be, when played by Bette Davis).

So it is in no way clear why neorealism would be a break with the past and an introduction of a time-image whereas all else before would be movement-image. Deleuze is not that dogmatic though, since Citizen Kane (1941) for him is time-image, despite being made in the US before they even entered the war. In general, Deleuze is a bit more flexible than his own writing suggests (and the way he's often interpreted). He does for example discuss the films of Vincente Minnelli as if they were time-image, despite him working in the US and having begun before the neorealists. And he discusses the films of Yasujiro Ozu as being time-image, even before the war.

But maybe it is actually a mistake to divide films into either movement-image or time-image. Deleuze divides movement-image into three sub-categories, one of which is affection-image, which is the close-up. Should we take this to mean that there are no close-ups in the time-image? In that case few films are time-image. But if time-images are about time, then having close-ups shouldn't matter. Of these three sub-categories Rodowick says that "every film combines these three sorts of images." This then would mean that all films are movement-image? On the other hand Deleuze says that the time-image was inherent in the movement-image from the beginning, occasionally peeking through but that "it could only realize that it had it in the course of its evolution, thanks to a crisis in the movement-image".

Deleuze at one point says that the time-image has become the "soul of cinema", whereas now only the mainstream, commercial films are movement-image. To speak of the "soul of cinema" is rather esoteric and unhelpful in the extreme, and besides, it would appear that most films still is movement-image. Maybe movement-image, since it dominated completely for the first 50 years and still continues to dominate is the true "soul of cinema". But of course neither image is the soul of anything.

Is it not possible that there never was a shift from movement-image to time-image but that different movements and different filmmakers, all over the world and all through film history, have had different ways of dealing with time and space, either out of social context, stylistic clichés or personal sensibilities? That there was always both the time-image and the movement-image, side by side, sometimes in the same film. Could you not say for example that in Don Siegel's best films movement-image and time-image merge. Or perhaps it would be easier to get rid of these confusing and contradictory terms all together.

But I still want to rescue one half of the idea of time-image, because there is something worthwhile in that concept. Not when it refers to Last Year in Marienbad but when it refers to the breaks in the narrative flow that Deleuze writes about. One example of a time-image sequence that he mentions, inspired by André Bazin, is the pregnant maid, alone in the kitchen, in Umberto D. (1952). It has been used by so many others for explaining the time-image that I'm sometimes tempted to ask if it is the only example. But that would be unfair because films are full of those scenes, and besides it is a lovely scene (although it is still embedded in the story, linked casually to what happened before). You can find them all the way from the beginnings of narrative feature film in the mid-1910s, scenes when the story, the narrative, stands still, takes a step back, just observes time, or the movement of bodies in unaltered time and space. They are plentiful in the films of John Ford, one of the foremost of "time-artists", in the films of Japanese directors such as Yasujiro Ozu, Mikio Naruse, Sadao Yamanaka, in films of Frank Borzage, Ernst Lubitsch, Howard Hawks, Leo McCarey, Powell and Pressburger and Jean Renoir, to name some obvious examples from pre-second world war cinema. Possibly George Cukor as well. But these sequences don't need Deleuze to be acknowledged, in fact arguably Deleuze forgets about them in his eagerness to give a post-war birth to this time-image.

There are many terms, concepts and ideas in Deleuze's cinema books that I haven't mentioned, and I won't either, not this time. There are aspects of the "crystal-image" which I like, but I would use it in a different way from how he does it, and I might come back to that. Now I will just say that even though the cinema books might be useful for the study of the philosophy of Deleuze, or philosophies on time, it becomes a problem when he tries to force his theories about time and life in general on to the films, and his analysis and historical perspective suffers as a result.

--------------
Deleuze could of course easily argue that I'm wrong in all that I'm saying because, on a certain level, he invented these images. He decided that (almost) all films prior to 1945 were this movement-image. Then he decided that there was a crisis in this movement-image that he had invented, and then he came up with the time-image. Who am I to argue with this, these were his concepts and he should be allowed to use it as he sees fit. Since he has only explained the rules of his images in opaque terms, he has given himself some wiggle room.

I would also like to suggest that the cinema of Borzage is the love-image.

The quotes from D.N. Rodowick are from Gilles Deleuze's Time Machine.
The quotes from Deleuze are from Cinema 2: The Time-Image.