|
Cinema and Architecture: Intersections of Illusionism
'Brian Henderson
Brian Henderson
Designing Dreams: Modern Architecture in the Movies by Donald Albrecht (Harper & Row: New York, 1986) represents one approach to the optic of "Architecture and Cinema", and, as far as it goes, a useful one. Albrecht's thesis is that the modern movement in architecture – Sullivan, Wright, Le Corbusier, Mies Van der Rohe, the Bauhaus designers and others -had a large and important impact on Hollywood films in the 20s and 30s, especially between the mid-20s and 1939. (He also devotes a chapter to the movement's impact on European cinema between 1916 and 1933.) Albrecht's brief account of the modern movement and its impact on some notable Hollywood set designs (and designers) is excellent. The book's photographs of buildings and film sets are very well chosen and, in many cases, well analyzed by the author. The photographs do not simply support the argument of the text – in considerable degree they are its argument.
The book was co-sponsored by the Museum of Modern Art although it was never, apparently, an exhibition there or elsewhere. The book's preface clarifies the Museum's participation and at the same time reveals the limitations of Albrecht's approach. Ludwig Glaeser, a curator in the Department of Architecture and Design at the Museum of Modern Art, proposed to his friend Albrecht, a practicing architect and film-lover:"… a study that would begin as a comparison of the Museum's vast files of architectural photographs with its enormous collection of movie stills." (1) This accounts first of all for the brilliant clarity and luminosity of the photographs of film sets in Designing Dreams. The "movie stills" in the Museum's collection are, of course, production photographs rather than frame enlargements. Movie stills were taken by professional photographers under contract to the studios, or hired for the occasion, who routinely took photographs of a film's leading players in many or most of the film's important scenes, primarily for publicity purposes. Notable sets were also sometimes photographed for their own interest, or emphasized in shots of the players. (2)
Albrecht's method is to compare two groups of photographs. An earlier generation of film critics would have said that what he omits is precisely cinema. The present generation, which might be called post-semiotic, will say that his analytic method bypasses the question of film discourse.
The legacy of naive realism may have been banished elsewhere but it still hangs over the entire question of architecture and set design in cinema. The signifiers fall away – that set in that photograph is the object of our analysis and may be addressed directly. Connotations may be deduced from it and confidently correlated with narrative, genre, ideology, and history. But architecture and set design never appear simply or directly in films. They are not present in or of themselves. Films create representations of architecture but it is film discourse that governs the production of such representations. Actual sets are only one signifying component in this production and never function autonomously. Films may not even be said to represent sets, which always operate in conjunction with other elements to produce a new representation. The "set" that appears in a film is thus at best the representation of a reprensentation and operates even as such only in a transformed way.
Critics who are concerned, like Albrecht, with the art-historical styles of set designs usually say little about their cinematic function, which is to create the illusion of reality, including the illusion of depth, to the eye of the spectator. (The spectator has two eyes open but the camera has only one.) Ordinary set designs routinely make use of windows, doorways, courtyards, and many other frame-within-a-frame constructions to heighten the sense of distance between the successive planes of the composition. They also use false perspective – objects and figures in the distance built in a smaller scale than those in the foreground – to create, in conjunction with other devices, the illusion of depth.
The very optical conditions that make possible spatial illusion within set designs make possible also the replacement of such sets, in part or whole, by miniatures, elaborately built to scale, and/or by various devices that create composite images – rear projection, glass shots, static and travelling mattes, the Schuefftan process, the optical printer, and a large number of variations and combinations of these.
Such devices are far more pervasive in classical cinema than most critics and theorists, even the most sophisticated, have been realized or, in any case, acknowledged. Andre Bazin's theoretical work has been much criticized but his critical defense of Jean Renoir, Orson Welles, and the neorealists, particularly Roberto Rossellini, still is current. Bazin praised the work of these directors for the maturity of their themes, the "novelistic" complexity of their narratives, and for the realism of their shooting styles, particularly their preference for the use of composition-in-depth, emphasizing the depth of the image, and their concomitant avoidance of editing.
Linwood G. Dunn, who did the optical printing for Citizen Kane, recalls in a 1983 essay:"I was asked during post-production to make radical alterations in certain filmed scenes . . . There are probably not more than three or four persons living today who really know the great extent of post-production modifications made throughout this film … as well as the many other photographic effects techniques that were utilized." (3) That more people now know Dunn's story is due chiefly to Robert L. Carringer's 1985 book, "The Making of Citizen Kane" (4) , which synthesizes the testimony of the film's many collaborators. The "deep-focus" shot of Susan's suicide attempt – glass and poison in extreme foreground, Kane and doctor bursting into the room in the background (Plate I) – is in fact an in-camera matte shot. (The foreground was light and focused with the background dark, then the foreground was darkened, the background light, the lens refocussed, the film rewound, and the scene reshot.) The shot of a tiny Kane at the end of a long corridor at Xanadu is a composite of three separately photographed elements. The opening shot of Xanadu "begins with a moving camera shot up a prop fence, and it continues with a series of shots, connected by fades and dissolves, of miniatures, models, and background paintings." (5) (Plate II) Carringer estimates that more than 50% of the film's total footage involves special effects of one kind or another; Dunn says that in some reels the percentage of optically printed work is as high as 80%.(6)
Although some of his statements denounce special effects, Roberto Rossellini himself invented a modification of the Schüfftan process which permits its use with zoom shots. It uses "a special mixture of silver salts and glucose for partly silvering glass to form a mirror which makes it possible … to combine action with scale models".(7) Rossellini used this device particularly in his historical films. Among other instances, the Versailles scenes in "The Rise to Power of Louis XIV" (1966) were created by virtue of this device. Bazin's third canonized realist, Jean Renoir, used a glass shot in Nana in conjunction with a partially constructed set by Claude Autant-Lara. The floor and grand stairway of a mansion were constructed by the set designer; its ornate upper story and ceiling were painted on the upper part of a sheet of glass that was placed close to the camera. When the camera photographed the partial set through the glass, the painted image merged with the partial set in a seamless illusion. (8) (Plate III)
The films of Alfred Hitchcock pose a very different set of questions regarding cinema and architecture. Hitchcock has never presented himself, or been presented as a realist but, as is well-known, he has a predilection for familiar locales as backdrops to the melodramatic actions of his films. One might call them, indeed, superfamiliar settings, even clichés – the British Museum in Blackmail, the Dutch windmills in Foreign Correspondent, the Statue of Liberty in Saboteur, a tourist's Rio (Notorious), Nice (To Catch a Thief), and San Francisco (Vertigo), and, of course, Mount Rushmore, the United Nations building, a Frank Lloyd Wright house, et al. in North by Northwest.
These postcard locales serve narrative and cinematic functions that are less readily recognized than the landmarks themselves. Some modern filmmakers use space to disorient viewers, to keep them off-balance and uncomfortable. Making sense of the film's spaces is a process in which the viewer must actively engage, a struggle that may not succeed but that is necessary to give the film meaning. Hitchcock does very nearly the reverse. His films propose again and again very familiar spaces, the effect of which is to lull the viewer into a false security. We know where we are, and therefore we relax. In each case, of course, he pulls out the rug from under us and in each case it works.
It is also true, from the standpoint of narrative construction, that Hitchcock's films move quickly from event to event. There simply is not time to introduce spaces before the action that is to occur there; he uses landmarks as a sort of shorthand that allows successive phases of the action to proceed without intervals or preliminaries. Finally, though the point may be obvious, Hitchcock uses his striking variety of locations to keep the eye of the viewer entertained. Most directors shuttle back and forth between identical set-ups, at least in dialogue scenes. Hitchcock almost never does this – he changes the camera angle or his characters' positions, revealing new views of the background, to insure a series of ever fresh compositions.
Precisely how Hitchcock articulates his landmark settings in the course of a scene or sequence illuminates the relationship of architecture and set design to filmic discourse. Hitchcock typically begins with a location shot of the landmark in question – Mt. Rushmore in North by Northwest, for example. After the CIA man explains the film's plot to Roger Thornhill (Cary Grant) in a Chicago airport, Hitchcock cuts from Grant's face, as he realizes Eve's precarious state, to a shot of Mount Rushmore in South Dakota. The camera zooms to a closer shot of the heads of the four Presidents, which is emphasized further by an iris (a circular matte) that suddenly appears. This is followed by a shot of Thornhill, framed against a long shot of the mountain, looking through a telescope viewer. We know from this shot that Thornhill has decided to cooperate with the CIA to protect Eve – he is talking to the CIA man as he looks through the viewer. We know, at another level, that the mountain will have a role in the action. (There are nine shots of the mountain in this scene and two more in the next.) Chekhov said that if you introduce a pistol in Act I, it must go off by Act III. Hitchcock adapts this – if you see Mount Rushmore at the beginning of a sequence you know that his characters will climb down it at its end.
As in most of his films, special effects were used throughout North by Northwest, but never more intensively or in so many combinations as in the climax on Mt. Rushmore. The most interesting feature of the sequence is that Hitchcock changes angles as frequently here as elsewhere, providing the viewer with ever new perspectives but putting unusual burdens on his set designers and special effects people. The first sight of the monument by the fleeing couple – the back of the President's heads – seems to be a constructed set. Six minutes and 108 shots elapse between this shot and the end of the film. The couple's descent and the villains' pursuit down the face of the monument, moving around and between the heads and various rock formations, is accomplished by, among others, the following devices: Partial sets of rock formations are shot flush against vast rear projections of a presidential face or faces (in various angles) or a vista of surrounding bluffs. (Plate IV) Sometimes the characters are shot directly against rear projections. In all cases, wind effects on the clothing of the actors supplement the illusion. In shots showing the position of some or all of the characters on the mountainside – in order to show their position relative to the other characters and/or to the monument – some or all of them are matted in through multiple exposure. The long shot of the police and the CIA man when they first appear on the mountaintop at the end of the sequence seems to be a shot of the mountain with figures, the actors or stand-ins, matted in. This is followed by a closer shot of the actors standing on a set designed to match the mountain. The last shot on the mountain shows Thornhill starting to pull up Eve, who is hanging from the rock below him. A "cut on motion" completes this action in a sleeping compartment on a train back to New York, as Thornhill pulls Eve to an upper bunk. (9)
It remains to discuss briefly the question of subversion or transgression in the cinematic representation of architecture. There are many filmmakers who have denied the viewer spatial orientation, who have constructed incoherent, dreamlike, or impossible spaces. This is most radically pursued, no doubt, in "experimental narratives" – that sector of the avantgarde that has some division of spaces according to a series of events, however dreamlike or disconnected. Works in this large category include, besides Bunuel, Cocteau, Delluc, Dulac, Epstein, etc., in the twenties, Maya Deren's "Meshes of the Afternoon" (1943), Kenneth Anger's "Fireworks" (1947), some of Andy Warhol's films and, more recently, the work of Yvonne Rainer, Erika Beckman, and many others.
From one point of view, these works "subvert" the spatial and architectural codes of the Hollywood and mainstream European cinema. From another point of view, it diminishes such works to see their primary operation as a critique of commercial cinema. They do something much more important than that – they use space in new ways to create new kinds of aesthetic entities.
To speak of transgression we must perhaps speak of practices within a signifying field, not of those that lie outside it. This, at least, is Roland Barthes' emphasis when he defines semiotic transgression as "a paradigm is extended into a syntagm".
"There is then a defiance of the usual distribution syntagm/system, and it is probably around this transgression that a great number of creative phenomena are situated, as if perhaps there were here a junction between the field of aesthetics and the defections from the semantic system. The chief transgression is obviously the extension of a paradigm onto the syntagmatic plane, since normally only one term of the operation is actualized, the other (or others) remaining potential: this is what would happen, broadly speaking, if one attempted to elaborate a discourse by putting one after the other all the terms of the same declension." (10)
We know that the methods for creating, or enhancing, representations of architecture in cinema were carefully kept from the public. See, for instance, the publicity still for the 1933 film Deluge (Plate V), which reveals the true scale of an elaborate miniature of New York City by showing technicians working on it. "Information Intended for Exhibitors Only" is the warning printed across the top of the photo. To show objects of disparate scale in the same image was transgressive in classical cinema and was carefully avoided, even in publicity photos.
The films of Terry Gilliam ("Time Bandits", "Brazil", the ship/building section of Monty Python's "The Meaning of Life") concern themselves, intermittently at least, with two illusionistic qualities of the cinematic image, the relativity of scale within it and its illusion of depth. These practices are transgressive in the semiotic sense defined by Barthes. They thereby reveal to viewers what has almost always been hidden from them. (Whether they are actually perceived by audiences as exposures of illusion or, on the contrary, as new kinds of spectacle, is an interesting question.)
In the little boy's bedroom in "Time Bandits", the picture of a mediaeval knight tacked to his wardrobe suddenly bursts into three dimensions and menaces him with charging steed and waving sword. When the attack is over, the boy examines the picture by taking it in his two hands: it is flat.
The next night five dwarfs tumble into his room out of the same wardrobe. Its apparent shallowness reveals upon closer inspection a surprising depth, a passage, indeed, from and to other worlds. When the dwarfs and boy attempt to leave, escaping the Supreme Being, the wardrobe and the wall behind it move back at an alarming rate, stretching the boy's room to a virtually infinite depth, bounded only by a hole in space that they all tumble through.
In the Napoleonic section of the film, we see a recreation of "The Execution of Maximilian" (1868–69) by Edouard Manet, as soldiers execute citizens following the battle of Castiglione (1796). The choice of this image and the mode of rendering it cinematically emphasize the flatness or two-dimensionality of the painted image, but the tour de force does not quite work – we are aware of three dimensions. (This is reinforced by cinema's time dimension – we hear the rifles fire and we see the bodies fall.) Thus a medium which is two-dimensional but pretends to be three (cinema) reproduces an image from another illusory medium only to emphasize, finally, its own superior illusionism.
Many of the film's situations and gags reveal a preoccupation with scale. Napoleon, for instance, is so obsessed with his shortness that he talks of nothing but height. Watching a Punch and Judy show, he is apoplectic when full-sized performers take the stage. "What I like is little people hitting each other", he says. Embarrassed by his tall staff members, he orders them to give their uniforms to the dwarfs. He invites them to dinner and regales them with tales of other short men who made history. Perhaps the film's boldest play with scale concerns the sailing ship piloted by the dwarfs. A storm that rocks the boat turns out to be a giant emerging from the water – he walks onto land wearing the ship on his head like a hat. (Plate VI)
There is a moment of transgression in "Brazil" that is all the more interesting for being tangential, at best, to the film's plot. We see Sam Lowry driving his tiny one-man automobile amidst trucks so huge that we cannot see the environment at all. (He is on his way to deliver a check to the widowed Mrs. Buttle, in whose house he will meet the love of his life; she, or rather his own fantasies, will soon cost him his own life.) Lowry pulls ahead and Gilliam cuts to a moving road shot down a row of what look like huge nuclear reactors. We assume that this is what Lowry sees as he drives into the clear. The road comes to a stop, however, in front of a reactor standing in the middle of the road. As the moving shot – and Lowry's implied motion – slows, the huge face of a drunk holding a beer bottle suddenly emerges to the side of the central reactor. The next shot reveals that it is a glass-enclosed model with a normal-sized drunk pressed against it. Behind the model, at a great distance, we see Lowry's car continuing on its way. He has escaped the huge trucks and the huge buildings for the moment but we know that they will overwhelm him before long. In "Brazil", no less than "Time Bandits", disparities in scale are a central concern.
(1)
Donald Albrecht, Designing Dream: Modern Architecture in the Movies (Harper & Row: New York, 1986), page VII.  back
(2)
The fruits of this practice, which are preserved in large collections, public and private, have been a mixed blessing for film historians. Such photos are sometimes the only trace that remains of lost films, and therefore comprise an invaluable record. On the other hand, most film analysts have at some time or other found themselves searching a film in vain to find an image that appeared in a book, magazine, newspaper or photo archive. The explanation, of course, is that the images were not taken from the film at all, but from production stills shot by a photographer on the set. The alternative to the beguiling lies of production stills is to use only reproductions from the film itself, that is, enlargements of individual frames. This is not a fully satisfactory solution, however, because frame enlargements are invariably inferior to production stills in photographic quality. They are deficient in contrast and definition (they're "fuzzy") and they lack tonal values (they're "muddy"), even when taken from a 35 mm print. Careful film scholars nevertheless insist upon frame enlargements in their published work and refuse to use production stills – cinema is what they propose to analyze, not photographs. back
(3)
Linwood G. Dunn, "Cinemagic of the Optical Printer", in Dunn and George E. Turner (Ed.), The ASC Treasury of Visual Effects (American Society of Cinematographers: Hollywood, 1983), page 240. back
(4)
Robert L Carringer, The Making of Citizen Kane (California: Berkeley, 1985). back
(5)
Ibid., page 94. back
(6)
Ibid., page 99. back
(7)
Jose Luis Guarner, Roberto Rossellini, tr. by Elisabeth Cameron (Praeger: New York, 1970), page 6. back
(8)
Leon Barsacq, A History of Set Design, tr. by Michael Bullock, revised and edited by Elliott Stein (New York Graphic Society: Boston, 1976), page 74. back
(9)
François Truffaut remarked to Hitchcock, "It seems to me that there were many trick shots in that picture, lots of them almost invisible, and also many special effects, like miniatures and fake sets." Hitchcock does not deny this but cites only an example – the "exact copy" of the United Nations lobby he had made – in his answer. François Truffaut, Hitchcock (Simon and Schuster: New York, 1983), page 251. back
(10)
Roland Barthes, Writing Degree Zero and Elements of Semiology, tr. by Annette Lavers and Colin Smith (Beacon Press: Boston, 1970), page 86. back
|
