An encounter with Patrick Keiller
In setting up the lecture series that was the seed for this book, we wanted to discuss the issue of film space, how it differs from architectural space and how it deals with the question of movement. Who better to invite than Patrick Keiller, educated and active as an architect before he became a film-maker, and author of a new kind of film depicting place and architecture in its political context, of which Robinson in Space and Robinson in Ruins are the best-known examples. Keiller arrived with a bag full of discs to show us samples of his own and other types of film and entered into a fascinating and wideranging disquisition about the use of film and the history and theory of cinema, followed by a question session that, among other things, revealed his attitudes to his own work. We recorded the occasion and made a transcript, but it is hard to follow without the visual examples that were the essential core, and it was an oral performance rather than a text. All the same, a number of valuable observations and arguments were made and examples were presented, and, as the topic should not go unrepresented in this book, we have put together this summary.
We first looked at brief excerpts from two films, both of which seem to portray place and architecture. I, as host,1 first showed the opening take from Orson Welles’s Touch of Evil (1958), which runs for minutes and involves a cast of hundreds. The camera, on a mobile crane, starts inside and moves outside, then up and over the street, then down and along it, all the time following the principal actors, with the life of the street flowing around them. It is an extreme case in the history of cinema of complex, continuous movement involving persons, vehicles and the camera, stretching to the limit the capabilities of the crane and lenses. It must have been a nightmare to organise and rehearse, all for the sake of a new sense of realism by eliminating cuts. It gets the film off to a thrilling start, but the view of the camera is not the view of a potential observer, its movement not the possible path of a pedestrian. A second film, chosen by Keiller, was the opening of Jean-Pierre Melville’s Un Flic (1972), about a bank robbery conducted in an off-season French Atlantic resort, with the winter waves crashing in. As we ran the film, Keiller pointed out the joins between takes and remarked that, although the exterior shots were evidently authentic, the bank interiors with which they were intercut were quite probably a studio set. So began his series of reflections about the contrivance of cinematic space and the juxtaposition of images:
Un Flic is more conventional than Touch of Evil in that it begins with what was probably a long take intercut with the opening titles, but mainly because the first location is represented using conventional narrative film montage, so it has reverse angles and multiple changes of viewpoint which together create a representation of the space in which the action takes place. There is not much verisimilitude, and I am pretty sure some of the interiors are sets. So although the exteriors are a real place, the interiors probably are not. One of the reasons I chose it was because it has both. Both Un Flic and Touch of Evil might be regarded as examples of film noir. They are of a similar mood and both very location-based. Melville is perhaps best known in this country for his film Le Samuraï (1967), but I thought this a better example of more-or-less conventional spatial montage, of the assembly of space in film, just to show how people do it. Both films contain examples of the survival of what film historian Tom Gunning describes as the ‘cinema of attractions’, which he argues existed at the turn of the last century, before the advent of what we now recognise as narrative cinema.2 Until the mid-to-late 1900s, films were generally very short, between 1 and 3 minutes long, and consisted of one or more fairly long takes, indeed, many consisted of only one take.
They were not much edited, though there is evidence of editorial thinking in some, as the camera sometimes stops and then starts again. This was done for a reason, perhaps because something was in the way. Some films have more than one camera subject, but most have a single subject which persists throughout. It might be somebody singing a song or doing a comic turn, but is as likely to be some sort of topographical actuality. Many are street scenes or what are known as ‘phantom rides’: where a forward-facing camera is mounted on the front of a train, though the term is also used more loosely to indicate a film from a camera set up on a moving vehicle, a boat, or something similar, which might also be facing backwards or sideways. As Gunning points out, after about 1907, people began to edit films in a new way. They had already started joining these attraction films together to make little stories. Sometimes the people who did this were not the film-makers themselves. The film-makers, perhaps, were happy enough just to film actuality because it was such a marvellous thing to do. It was those interested in making money out of the films who started to splice them together; the producers, as they became known. This is allegedly how narrative cinema evolved, and the change in my view was much more fundamental at this stage than the later change from silent to sound.
This change of form did not just affect fiction. There were similar things that looked like newsreels, often called ‘topicals’. So by the end of the first decade of the twentieth century the form of film had changed. But as Gunning writes, ‘the cinema of attractions does not disappear with the dominance of narrative but rather goes underground, both into certain avant-garde practices and as a component of narrative films’.3 That is why I think it is interesting to look at these two feature films, because their opening sequences can be seen to be descended from phantom rides and other kinds of early film.
I was reminded of all this recently by a review, by John Naughton, Professor of the Public Understanding of technology at the Open University, of Tim Wu’s The Master Switch: The rise and fall of information empires. The book is about the Internet, but also discusses the way previous technologies were similarly colonised by large corporations. As Naughton put it: ‘A freewheeling, chaotic and creative industry was cornered by a cartel of vertically integrated corporations which for decades channelled all cinematic creativity through a set of narrow apertures.’4
My interest in Tom Gunning’s proposition is that it helps explain to me why I make such films. I am perhaps trying to recover something of ‘the cinema of attractions’, though I’m not quite sure what. My films do not really look like early cinema, but they do have some things in common with it. They do not follow the normal tradition of representing architectural or landscape space, even in the way Melville does. Melville manages to make architectural documentaries, even if he does it inadvertently. But I am doing something else, attempting to perform a transformation on everyday surroundings, so I point my camera at unremarkable places in the hope of enacting a transformation upon them.
Henri Lefebvre seems to describe something similar. In The Production of Space he writes: ‘The fact is that the space which contains the realised preconditions of another life is the same one as prohibits what those preconditions make possible’.5 I understand this to mean that if you look at everyday surroundings in a certain way, with a certain subjectivity, they seem to contain ‘the realised preconditions of another life’, but this space somehow also prevents the revolution, the transformation, actually occurring. It seems to me that in films, the ‘prohibition’ is sometimes suspended. So there is something potentially utopian about film space, though utopian is perhaps not quite the right word.
Keiller then showed us excerpts from Robinson in Ruins, which is structured in months, and he chose to begin with October. He uses a static camera set on a tripod, so that any movement is either what occurs within the frame or the shift from one camera set-up to the next, and the story is held together by the narrator’s voice-over about the fictional Robinson, informing us of aspects of the political and historical background and building a narrative around the place. Within this disciplined format, the choice and juxtaposition of images are of crucial importance, and he began to reflect on this:
There are two points I’d like to make. The first concerns the attempt, because some locations are intervisible, to orient each view in relation to the others through recurrent elements, such as the factory chimney and the railway. The railway appears in three set-ups, the factory chimney in at least two, so you get a kind of reverse angle. First you see a quarry from across a railway line, then from the other side, and this creates a fictional landscape, though perhaps not much like the real one. If there was action it would be easier to interrelate the pictures because people would, for example, walk out of the right side of one and into the left of another, so one would assume that the space in the second picture lay to the right of that in the first, which is the way spatial cues function in played cinema. The other point, which was implied or exemplified by the pictures of the maize harvest, is that a cinematic image covers very little of the normal field of vision, even with a relatively wide-angle lens. Everyday experience involves much more architectural or landscape space than you could possibly depict with a camera, which is why extensive spaces in film are most often assembled through montage. Sound is very important too, because its continuity can imply the proximity of partial views so as to create an impression of a wider landscape.
As well as his own films, Keiller brought along examples of early cinema, mostly views of British cities, that he had included in an installation at the BFI in 2007, interrelating them through a series of maps. He presented several variations of what he had called the ‘cinema of attractions’, the age of innocence of the medium, before the imposition of commercially driven narrative:
The Lumières’ first public screening of films was in December 1895. It was not long before cinematographers began mounting their cameras on things that moved: the first moving-camera film is supposed to have been a Lumière film – which does not survive – photographed by Alexandre Promio from a gondola in Venice in September or October 1896. The first film photographed from a train was probably Départ de Jérusalem en chemin de fer, an oblique, rearward view of a station platform seen from a departing train photographed, also by Promio for the Lumière company, in January or February 1897. More typical is Through Miller’s Dale (near Buxton, Derbyshire) Midland Rail (1899), a forward-facing view from the front of a locomotive, one of the first phantom rides made in the UK. I am not sure exactly why these films were called phantom rides. I imagine it was a reference to virtuality, but the disembodied point of view, from a place where a person could not normally stand, perhaps suggests the view of a ghost. As far as I know, phantom rides were so called at the time.
Through Miller’s Dale is about two-and-a-half minutes long, three takes along the Midland railway as it passes through Miller’s Dale junction. It was made by the Biograph Company, photographed by William Kennedy Laurie-Dickson, one of the co-founders of the American Mutoscope and Biograph Company, who then came to the UK to start up its British off-shoot, the British Mutoscope and Biograph Company. He had previously developed Edison’s motion-picture technology, so is one of the founding figures of moving pictures. The Biograph camera was very big, probably too big to be placed on the front of the locomotive, more likely in its own truck. It photographed at 40 frames a second on to 68-mm film, which is why the best of the surviving Biograph films are so impressively sharp and detailed, and gives you some idea of the technological ambition of early film pioneers. This was by no means a cottage industry, and by no means undercapitalised. The Lumière company, too, had a global reach.
The screen is also flat, another big difference between cinema and architecture – film represents architectural space by creating an illusion of depth. One of the terms one comes across in the early period is stereoscopy, though it was not of the binocular kind. The true stereoscope with an image for each eye was a competitor of the early cinema, but certain film-makers seem to have sought what they called the stereoscopic effect, typically the result of differential parallax. In Through Three Reigns (1922), a retrospective compilation of his early films, Cecil Hepworth refers, in an intertitle introducing Thames River Scene (1899), to ‘stereoscopic cinema’. The film is a panoramic sideways view from a launch moving downstream on the river at Henley. The camera is moving among other boats moving at different speeds, which creates an illusion of depth. More generally, the illusion of depth in cinema, as in photography, seems to be enhanced by contrast, fine detail, and monochrome.
In the final part of his lecture, Keiller reflected particularly on the changing view of space brought about by modernism, as a result of technical advances, developing science, mechanical reproduction and new methods of communication, but he sets in the midst of all this László Moholy-Nagy and the inherent contradictions of his role as film-maker:
Reyner Banham identifies two kinds of architectural space. The transition between them begins towards the end of the first decade of the twentieth century and continues after the First World War. He refers to the writings of Theodor Lipps and the concept of Raumempfindung or as he puts it in English, ‘felt volume’, a conception of architectural space that we would understand as rooms, streets, squares, etc. He writes that Lippsian space, the ‘felt volume’, was superseded by something eventually defined in 1929 in Moholy-Nagy’s ‘minimum definition’ as ‘space is the relation between the position of bodies’,6 a way of thinking about architectural and three-dimensional space in which there is no container. It is interesting that this concept arose in a period when the solidity of matter had become less certain. A number of comparable transitions occurred at around the same time: for Virginia Woolf, ‘on or about December 1910, human character changed’; for John Berger, ‘The Moment of Cubism’ was located between 1907 and 1914, while something similar is described by Henri Lefebvre:
The fact is that around 1910 a certain space was shattered. It was the space of common sense, of knowledge (savoir), of social practice, of political power, a space thitherto enshrined in everyday discourse, just as in abstract thought, as the environment of and channel for communications; the space, too, of classical perspective and geometry, developed from the Renaissance onwards on the basis of the Greek tradition (Euclid, logic) and bodied forth in western art and philosophy, as in the form of the city and town [. . .] Euclidian and perspectivist space have disappeared as systems of reference, along with other former ‘commonplaces’ such as the town, history, paternity, the tonal system of music, traditional morality, and so forth. This was truly a crucial moment.7
Lefebvre’s book was first published in 1974. He was writing before people were aware of what was subsequently identified as the postmodern era, during which aspects of the earlier Lippsian concept of space have been revived. Banham’s book was published in 1960, and I do not think that he ever thought that Raumempfindung would return, although it is not altogether clear that it ever went away. Banham identified the second modernist space, the space of the modern movement, as that of Moholy-Nagy’s ‘minimum definition’. Moholy-Nagy was not an architect but he was, among other things, a film-maker, which offers an intriguing connection between architecture and film. His first film proposal, ‘Dynamic of the Metropolis’, was never realised.8 Begun in 1921, it was perhaps intended to be something like Vertov’s Man with a Movie Camera (1929). In his book From Material to Architecture, he formulated the concept of space identified by Banham, for whom an iconic image is a group of seaplanes montaged into an aerial view of a traffic intersection in Los Angeles. Moholy-Nagy’s book was published in 1929, the year of his film Marseille Vieux Port, although this, like his other realised films, is strikingly unlike the earlier film proposal, perhaps because of the limited means available. He described it later:
I had a predetermined length (300 m) of film stock, and decided that there was no way to capture a large city in so few metres. So I chose a small slice of the city, the old port, one hitherto unknown to the public due to its bad social circumstances, poverty and dangerous character. I tried to approach it with reportage that was not merely impressionistic; but finally I had to make do with a sketchy picture of the situation, since I was not even able to shoot from some higher point in order to better portray the totality of the processes before me. In this gloomy quarter, when after a long haggle I finally entered an apartment several storeys up, I was physically threatened, and so had to flee.9
The reality of film-making seemed to have got the better of Moholy-Nagy, but it did not necessarily do the films any harm as I wish I could show you. The other thing I would like to say about this transition between the two kinds of film-making is to quote, as we so often do, Walter Benjamin, writing in the 1930s in The Work of Art in the Age of Mechanical Reproduction:
Our taverns and our metropolitan streets, our offices and furnished rooms, our railroad stations and our factories appeared to have us locked up hopelessly. Then came the film and burst this prison-world asunder by the dynamite of the tenth of a second, so that now, in the midst of its far-flung ruins and debris, we calmly and adventurously go travelling.10
What can Benjamin have meant? What is this ‘dynamite of the tenth of a second’? He must have known that films were photographed at 18 if not 24 frames a second by the time he was writing. I do not think it has to do with frames, but is probably a reference to montage. So, what blew apart the prison world, of offices and furnished rooms, railway stations and factories was not the camera’s fragmenting duration into individual frames. Rather, the dynamite of the tenth of a second is the cut, which brings us back to the narrative cinema at the beginning. He was writing in the era of narrative cinema, and may not have had much experience of early cinema, which by that time was, I think, pretty much forgotten. Most of the people who made the early films had by then given up, often returning to their original professions as instrument makers, electrical engineers and so on.
Notes
1 I added this opening to the reworked version of the lecture, edited jointly by myself and Keiller, and he has asked me to make clear that Touch of Evil was my choice – Peter Blundell Jones.
2 Tom Gunning, ‘The cinema of attractions: Early film, its spectator and the avant-garde’, in Elsaesser, 1990, pp. 56–62.
3 Ibid., p. 57.
4 Guardian Review, 2 April 2011, p. 9.
5 Lefebvre 1991, pp. 189–190.
6 Banham 1960, pp.66–7, p.317.
7 Lefebvre 1991 (original 1974), p. 25.
8 Apparently it was first published in Hungarian. The title is usually written in English.
9 Quoted in programme notes by A.L. Rees and David Curtis for ‘Films by Moholy-Nagy: A medium of plastic expression’, NFT, 29 July 1987.
10 Walter Benjamin, ‘The work of art in the age of mechanical reproduction’, in Benjamin 1973, p. 238. See also Patrick Keiller 2013.