Transformations in Film as Reality (Part 6)

On the Virtues and Limitations of Collage

Paul Arthur

We ran Elizabeth Cowie's article on documentary and spectacle in our last issue of Documentary Box as the first part of our revival of the popular "Transformations in Film as Reality" series, which has explored the history of film's relation to reality—how documentary as a genre, as well as the "realistic feel" of cinema, have evolved over the last century. As the next edition in that series, we are pleased to present Paul Arthur's discussion of the use of collage in documentary, especially in relation to the work of Emile de Antonio.

—The Editors

André Bazin was not the first but possibly the most influential proponent of the idea that cinema, under the aegis of photography, "freed Western painting, once and for all, from its obsession with realism and allowed it to recover its aesthetic autonomy." 1 What Bazin located as the autonomous properties of film were obviously quite different from those which his contemporary Clement Greenberg defined as the modernist imperatives of painting. It would be intriguing to know what Bazin, who made his ontological claim in 1945, would make of the current international art scene in which boundaries between painting and photography have been all but abolished. Moreover, given his celebrated endorsement of temporal continuity and the transparent immediacy of film recording as the basis of realism, how would he assess the paradigmatic use of found footage collage in post-sixties' nonfiction cinema? Bazin was of course well aware that the introduction of extrinsic materials into cubist painting in 1912, and the subsequent development of various collage styles, posed a powerful challenge precisely to the aesthetic "autonomy" of the painted surface, and to the principle of organic composition, the integral relationship between part and whole. His attacks on Soviet montage effects and on analytical editing in general, elaborated in the essay "The Virtues and Limitations of Montage" and elsewhere, offers a possible blueprint for the rejection of collage as a divisive, "authoritarian" mode of filmic discourse. 2 From a contemporary vantage, however, Bazin's approach is of little help in understanding either the historical legacy of found footage or its material-aesthetic status in the representation of reality.

By 1945, the deployment of archival images to reanimate or polemically reinterpret prior versions of events, figures, and social processes--a practice nearly as old as the cinema itself 3 --had become a standard procedure in various types of non-mainstream film (i.e., apart from its common functions in commercial fiction films of filling iconic gaps in logic or continuity). Indeed, the recent outpouring of wartime newsreel compilations and military training films had underscored the primacy of found footage in corporate and state-sponsored propaganda (Frank Capra's "Why We Fight" series is exemplary here). In terms of the larger domain of nonfiction film, the use of appropriation techniques had developed along parallel, at times overlapping, trajectories: a figurative or experimental foregrounding of aesthetic features contained in otherwise mundane artifacts (films by Hans Richter, Walther Ruttmann, and Charles DeKeukeleire belong in this tradition); and the politicized activation of "suppressed" ideas or the inversion of conventional meanings culled from newsreels and other documentary materials (the work of Esfir Shub and Dziga Vertov are indicative of this tendency). 4

Found footage continued to play a role, albeit less prominently, in postwar Anglo-European production but was pushed aside by the advent of cinéma vérité and direct cinema in the early Sixties. Documentary styles that favored analytical editing and voice-over narration--in particular, New Deal documentaries of the thirties--became principal targets of vérités cultural ideology of spontaneity and semantic ambiguity. The movement's didactic insistence on present-tense recording and a non-judgmental, detached first-person address, its cult of the sync-sound image as replete signifier of the Real, implicitly or, as in the body of interviews with vérité practitioners, explicitly proscribed associational or rhetorical montage effects derived from the rearrangement of found materials. 5 To put it another way, the ostensible reduction of editing to the barest necessities--and, the contemporaneous fixed-camera masterpieces of Warhol notwithstanding, it is always necessary--in the accretion of successive moments of immediate recording served as a shibboleth for the vérité camp. The vérité method, as style and ideology, was tailor-made for the anti-authoritarian ethos of sixties' counterculture, and its dominance was predicated on a disavowal of history as vehicle for the adjudication of conflicting truth claims. Not incidentally, it was an almost ideal fulfillment, in the precinct of nonfiction, of Bazin's theory of cinematic realism.

The hold exerted by vérité on documentary discourse, especially in the United States, began to weaken in the early seventies with the emergence or acceleration of political struggles for racial, gender, and sexual equality, and their concomitant demands for the revision of individual and collective histories. A mandate for the revival of found footage strategies was in part spurred by academic historians intent on reinterpreting American Cold War policy in the midst of the Vietnam debacle, and received further reinforcement from the gradual acceptance of new European historiographic models of evidence and argument, including Foucault's notion of archeology. 6 Hence the widespread post-sixties' appetite for found footage coincides with two interdependent initiatives: the desire to reformulate tropes of historical narrative, and the micro-political critique of historical exclusion or distortion conducted on the terrain of mass cultural representation by disenfranchised groups as a precondition for self-determination.

It is possible to generalize the opposition of vérité and found footage-based documentaries in more narrowly discursive as well as political parameters. Although it frequently operates in tandem with present-tense interviews, the filmic structures in which recycled footage appears inevitably privilege the perception of conscious construction over "unmediated" presentation, relations of (dis)continuity between past and present rather than a reified semblance of temporal unity. Where the film image in vérité is registered as unique, temporally bounded, and phenomenologically embodied through camera movement and position, in the collage format it tends to be seen as dis-embodied, materially discrete, and iconographically interchangeable with other images. That is, the found footage aesthetic assumes not only the possibility that images are capable of eliciting multiple responses--an important tenet in vérité as well--but that the burden of meaning shifts according to context and surrounding articulation; it cannot be universalized or freed from historical determination. The organizing "voice" in collage films is decentered or split between an enunciative trace in the original footage--encompassing both stylistic features and material residues of production such as film stock, speed of shooting, and aspect ratio--and an agency of knowledge manifest in the overall text through editing, the application of sound, titles, and so on. In its framework of enunciation, as well as its thematic focus, collage constitutes a corrective to vérités predominantly individualist (and performative) encounter with social reality.

In the opening of an anthology of essays devoted to collage and filmic montage, Matthew Teitelbaum declares that the practice of recombining extant images "invokes the discontinuous and the ruptured as the talisman of our century." 7 This sentiment has been echoed repeatedly and in various guises in art-historical writings since the 1970s, or roughly the same period in which found footage renewed its claim on documentary practice. 8 In one of the few dissenting opinions of which I am aware, Theodor Adorno, who had praised the potential of earlier collage experiments as a "negation of synthesis" or false unity of meaning, asserts that after World War II collage effects had become neutralized through overuse and the unproblematic display of mass cultural detritus. 9 It is possible that both valuations are accurate in regard to different tangents of post-sixties documentary. Nonetheless, what is surprising is that given its centrality in art-historical discourse, collage has received so little attention in classical or contemporary film theory.

As might be expected, Soviet writers such as Vertov and Pudovkin are more attuned to questions of mixed materiality than are writers primarily concerned with narrative cinema, and some of their utopian pronouncements on the political efficacy of montage resurface in the claims of a current generation of film activists. 10 Perhaps the most interesting clues to the conjectural status of found footage are discovered in fleeting comments by Siegfried Kracauer, Parker Tyler, and Bela Balazs, aimed at the attributes of newsreels rather than collage per se. There is mutual agreement that newsreels and documentary reportage in general are "innocent" or "artless" due to their lack of aesthetic reconstruction. Kracauer, in a discussion of Housing Problems (1934), states that "it is precisely the snapshot quality of the pictures that makes them appear as authentic documents." The absence of "beauty" yields a greater quotient of "truth." Tyler makes a related point, suggesting that newsreels fail to meet requirements of historical narrative since the latter "implies not only recorded events but furthermore a narration of events having form--beginning, middle, end, and coherent outline." Newsreels, according to Tyler, have no "abstract perspective" and thus "make actuality seem (in terms of formal significance) fragmentary, superficial, and even trivial." 11 The assumption here is that the representation of "actuality" must be ordered by a coherent system of causality.

In this view, newsreel footage, a primary source for documentary compilation and collage, is merely denotative or "evidentiary," a series of discrete, transparent, essentially mute elements requiring a narrativized, univocal fabric in order to demonstrate how social events are connected. The notion of found images as more-or-less neutral bearers of meaning is reiterated in contemporary documentary theory by Bill Nichols and Thomas Waugh, among others. Nichols advises us that the presence of a "perspective, and therefore a representation or argument, differentiates a text from mere' film or raw footage," and implies that decontextualized fragments act as "value-free reproductions of the historical world." The recycled fragment thus seems to resemble Eisenstein's "montage cell," a shot that acquires coherent meaning only in juxtaposition with other shots. Waugh, rehearsing the collage effects of Emile de Antonio's Point of Order (1963), concludes that the ability to invest new meanings in appropriated images "is partly due to the naive, functional, purely denotative orientation of the original recording." He adds that "the transfer of video from the television screen to the movie screen somehow serves to absolve it of whatever lack of credibility infects it in its original, functional context. It attains purity as a document." 12 Presumably, the newly-realized "purity" is primarily a function of excising the fragment from its original context of institutional rhetorics and economic interests.

The understanding of newsreel fragments as devoid of enunciative or rhetorical specificity recalls Hayden White's discussion of medieval annals. Traditional historians characterize this form of historical writing as rudimentary, its objectivity defined by the absence of all reference to a narrator: "real events should not speak, should not tell themselves . . . they can perfectly well serve as referents of a discourse . . . but they should not pose as the tellers of a narrative." White goes on to argue that historiography is so wedded to the operations of narrative that the sorts of gaps and discontinuities exhibited in annals disqualify them from being taken as "particular products of possible conceptions of historical reality, conceptions that are alternatives to, rather than failed anticipations of, the fully realized historical discourse that the modern history form is supposed to embody." 13 It is as if at least one line of documentary inquiry, extending from Kracauer and Tyler through Nichols and Waugh (odd bedfellows indeed), had situated the newsreel and televisual report as modern versions of "un-narrated" annals. One might question first whether any product of mass culture can be successfully stripped of its contextual signification and, second, whether the transformative process is a matter of subtraction or amplification.

Setting aside for the moment issues of what can be included in the class of "found footage" and how the material is incorporated in particular filmic structures, the neutrality stance might have clear theoretical sailing were it not for a body of statements that appear to take precisely the opposite tack; that is, that the horizon of possibility--and the claims to political efficacy--in appropriated images is underwritten by a palpable over-determination, by ideological or contextual biases already amply inscribed in the artifact. The critical problem then is not combining unaffiliated fragments in order to generate meanings imposed by a more conscious historical perspective, but rather the interrogation of ideologically-motivated signifiers embedded in and disseminated through the original material. In the latter, found footage is by definition a dialogical strategy, pitting two or more distinct voices against each other. British nonfiction filmmaker Pratibha Parmar ventures that "It is [the] politicized appropriations of dominant codes and signifying systems which give us powerful weapons in the struggle for empowerment." Kobena Mercer remarks that found footage tactics can "expose the heterogeneity" of social identity denied by dominant visual discourses and can "defeat monologism" by "creolizing" newsreel and other representations. 14 De Antonio himself seems at times to buy into the idea that newsreels do indeed "speak for themselves"; referring to a particularly ripe image from In the Year of the Pig (1968) of French officers shooing away a group of Vietnamese rickshaw drivers, he calls it "the equivalent of a couple of chapters of dense writing about the meaning of colonialism." 15

Discrepancies between the two ontologies of found footage may turn out to be more pragmatic than substantive. Both groups are involved in promoting particular documentary styles or bodies of work reflecting a particular set of historical and cultural conditions. When they discuss the status or use of found footage, they may be referring not only to different types of materials but different sites of articulation (e.g., mainstream documentaries versus experimental essays). There are no obvious disagreements, say, about the character of newsreel fragments in the films of de Antonio. Moreover, both camps believe that the goal of documentary collage should be the contestation of received historical perspectives. However, there are still considerable divergences on several fronts. Analysts of mainstream documentaries such as Nichols and Waugh imply that the subversive transformation of artifacts must be coupled with a demonstrated "respect" for the integrity of the material, that the credibility of critique is dependent on a formal restraint in the handling of images (and sound). The principal, if not the exclusive, interpretive resources are those of editing and sound/image juxtaposition. In addition, the re-envisioning of historical narrative is predicated on the perceived identity, or at least close compatibility, between visual depiction and verbal reference--in the form of voice-over narration, printed titles and texts, and/or diegetic speech. What is illustrated in collage fragments must be related analogically to its linguistic description. An alternative position might hold that the process of recontextualization will inevitably, and should, vitiate the integrity of the original footage, that fidelity to technical standards (e.g., exposure, speed, cropping) of prior representations is neither a possible nor desirable goal. Similarly, the prerogative of radically de-forming found footage is based on a recognition that the site of revision is less the semantic content of borrowed images than it is the material traces and dominant visual/aural codes which they embody. 16

It is possible that this dilemma can be reduced to a contrast in aesthetic allegiances, with one group advocating a more reflexive brand of realism while the other opts for a formalist or figurative program of nonfiction collage. There is no way to resolve in this space the contradictory accounts of the status of found footage. Nonetheless, a closer examination of its supposedly denotative functions and of the range of interactions among sequential fragments will help to test some of the discursive assumptions made by documentary theory. Perhaps the most appropriate field for such an inquiry is the well-known batch of mainstream documentaries produced in the seventies and early eighties in the wake de Antonio's pioneering work, including Union Maids (1976), The Day After Trinity (1980), and Atomic Cafe (1982). These films share certain tendencies in construction, politics, and historical focus--tendencies which, by the way, continue to inform contemporary nonfiction projects.

It should be apparent that both modes of inscription--the illustrative or analogical and the metaphoric--can be commingled in the same documentary text. The selection of archival images to be included in a given sequence is governed by a variety of criteria, not just accuracy of depiction but also formal qualities of composition, expressive movement, camera angle, and so on. It is frequently the case that there is no existing or available footage to accompany a verbal description and the filmmaker/editor must resort to generic images that approximate, allude to, or symbolically represent the event in question. For example, in numerous documentary recreations of the attack on Pearl Harbor, airborne shots of Japanese fighters are shown in advance of, or intercut with, footage of explosions and burning naval vessels. It seems unlikely that cameras were placed aboard the planes of kamikaze pilots and, even if they were, how the footage of their attack was retrieved. Hence as viewers, we may grasp the fact that the shots of sinking American ships are exactly what they purport to be while those of Japanese aircraft are generic rather than literal. Needless to say, discontinuities of this type do not pose a problem for the average film viewer, they are easily assimilated within a larger fabric of narration, sound, and editing syntax.

To take a more specific example, at the beginning of The Life and Times of Rosie the Riveter (1980), a group of women are shown huddled around a radio listening, it seems, to FDR's speech declaring war on Japan. It hardly matters whether the figures in this shot are actually listening to the words we hear on the soundtrack; nor does it matter that these women are unrelated to the women shown in present-tense interviews narrating their stories of wartime factory work. Indeed, with the exception of rephotographed still images and brief snippets of home movies, the footage adduced to illustrate the women's reminiscences is almost entirely of a generic nature: not "this" factory but "a" factory, not "these" workers but "some" workers. In the compact we as viewers make with the film, expectations of fidelity between image and verbal statement are limited to areas such as historical or geographic verisimilitude (we would presumably balk at shots of Indian factory work recognizably taken in the 1930s). The point is that many, perhaps even a majority, of illustrative instances in documentary collage are understood not as literal but figurative representations of their subjects. Something similar can be said for the commonplace superimposition of synchronized sound effects (e.g., explosions, air raid sirens, the whirr of descending aircraft) over footage that was almost certainly silent in its original form. These are stock items in the arsenal of documentary technique, even though they patently violate the "integrity" of the material.

More interesting, but only slightly more problematic, are montage sequences connecting a series of verbal assertions in which the denotative and expressive valences of successive shots shift in its relation to what is being described. Near the beginning of The Day After Trinity, a twelve-shot sequence is employed to illustrate the build-up of the Nazi war machine by 1937. Images of marching troops, a bomber squadron flying in formation, and rows of advancing tanks create an impressive, rhythmically edited, right-to-left on-screen movement. An inserted close-up of Hitler in a virtual eyeline match is distinctly analogical, but the overall impression made by the sequence is symbolic, cued by the voice-over of a scientist who refers to a "wave" of German aggression (not incidentally, the leftward movement matches the map-direction of forces invading Western Europe). Again, while certain archival shots or sequences create a direct, transparent fit with their verbal referents, others are chosen for their visual expressivity or vividness in order to convey an abstract idea. In these interludes, questions of "who-where-when" are bracketed in favor of an affective charge or latent meaning.

Yet another class of familiar collage constructions display a more troubling ambiguity. The opening of Atomic Cafe revisits the nuclear attack on Hiroshima. Pieces of an interview with the captain of the Enola Gay, in which he describes the chronology of events leading up to the blast, are laid over shots of peaceful Japanese streets, the municipal tower at the center of Hiroshima, the bomber in flight, and other pre-blast scenes. There are also several low-angle medium shots of a well-dressed, clean-cut man standing against a gleaming white sky. He gazes upward, we hear airplane sounds. The man's inquisitive look is cross-cut with the bomber, which is followed by the explosion. The image of this innocent citizen is presumably generic; its presence defies the logic of direct recording. In context, the montage sequence makes a discursive leap that frays our intuitions of documentary protocol, adopting a narrative editing trope that both heightens dramatic anticipation and elicits pathos for a specific individual. Since we may reasonably doubt that this man was an actual victim of the bombing, his function within the sequence is confusing. The fact that he does not belong to the scene portrayed becomes important, and misleading, in ways that related substitutions do not. 17 It is uncertain to what degree this example illuminates responses to other narrativized uses of found footage but, regardless of its general significance, it hints at the wide latitude of non-denotative tasks performed by documentary collage.

A final pattern, extracted from Union Maids, concerns a level of outright contradiction between image and speech that is not cognitively disruptive, as is the Hiroshima sequence, but considerably more intriguing. Activist Sylvia Woods is telling the story of how a group called the "Unemployed Council" tried to halt the eviction of an unemployed worker. She says: "The police came out from everywhere, one of them, a detective with a sawed-off shotgun. He was so big and so mean and he said. . . ." One does not have to be a weapons buff to notice that the image supporting her description is of a uniformed officer with a tear gas launcher and vest full of canisters. It is a powerful shot, although it is neither generic enough nor sufficiently analogous to create a transparent match. There are several ways of assessing the split evidenced here, depending on the model one adopts for film spectatorship. Given an active, narrativist model of hypothesis testing of the type endorsed by David Bordwell for fiction films, 18 we can infer that, as with many other events described in the film, no recorded evidence exists and that the shot is representative of "all cops" mobilized in the suppression of labor struggles of the 1930s. It is also possible that the menacing cop shown in the sequence was present at the confrontation and Woods's memory, given the erosions of time and subjective storytelling, is faulty. It is far from unusual for found footage to occupy the same structural position and logic as a fictional flashback. The tension between Woods's description and what is seen could serve to modify or even displace her verbal account, substituting detached vision for personal recall.

Rather than reaffirming the transparency of visual depiction in historical memory, this common form of dissonance raises the specter of evidentiary blockage or partiality. Documentarists who would never dream of restaging an event with actors do not hesitate in creating collages which amount to metaphoric fabrications of reality. The guarantees of authenticity ostensibly secured by archival footage are largely a myth. In consequence, the binary opposition of unalloyed illustration--as the province of conventional documentary-- and figurative reshaping is hardly as solid as it initially seemed. Ironically, this opposition is more pertinently applied to collage-interview documentaries of the seventies and beyond than it is to the seminal early films of de Antonio, as a brief consideration of In the Year of the Pig will readily confirm.

De Antonio was no stranger to the avant-garde film scene of the sixties, and it is perhaps no accident that he began assembling found footage tracts just a few years after Bruce Conner issued his devastating critique of dominant media's exploitation of violent disaster in A Movie (1958). To be sure, de Antonio had an ample store of documentary compilations on which to draw, and he was always reluctant to acknowledge the affinities between his work and avant-garde collage methods. That said, however, In the Year of the Pig represents his most radically formal treatment of the possibilities of found footage in documentary discourse--an approach which, regrettably, has had few imitators among later generations of documentarists. The film develops a series of flowing yet semi-autonomous segments bound together by historical chronology, an ongoing commentary by a group of scholars and public officials, and a set of continuous themes: the life of Ho Chi Minh; the relationship between Vietnam's struggle for self-determination and America's wars of independence; contradictions between American political rhetoric and expanding military involvement.

Scattered among long stretches of historical exposition are interludes that create nearly abstract montage effects appealing less to our rational faculties than our emotional instincts. They function as momentary breaks from the onslaught of verbal argument and description, but they are also integral extensions and complications of ideas which transcend simple paraphrase. In addition, de Antonio layers over the image track both sound effects and music which rely on purely metaphoric sensations as well as intellectual linkages. For the opening montage, the director commissioned the composition of a concrete musical crescendo of distorted helicopter noise which registers at different moments as out-of-control machinery and human screams. In the same sequence, he filters in the sound of two--perfectly segued--versions of "The Charge of the Light Brigade," a melody evoking strong colonialist associations; one is a scratchy old record and the other a more full-bodied "modern" recording. The formal juxtaposition of these two sound artifacts suggests an historical continuity between earlier and current episodes of foreign intervention. The significance of very first shot, a statue of a Civil War soldier, remains somewhat obscure until the playing of a few bars from "The Battle Hymn of the Republic" in the film's final segment--not only will there be no glory or honor on the battlefield, we have no business interfering in a civil conflict.

Thomas Waugh finds that "de Antonio's rhetoric is cool, scholarly and articulate. Its aim was to convince, not to inflame." 19 His assertion is not wrong but it devalues both the film's extraordinary range of figurative usage and those passages which do indeed seem structured to inflame our emotional sensibilities. For instance, in the pre-credit montage, de Antonio separates almost a dozen heterogeneous glimpses of the war (including a canny reference to Saint-Gaudens's monument to a company of black Civil War soldiers) with fades and black leader--in a manner not dissimilar from the opening of A Movie. These pauses signify the passage of time; they also instate an almost funereal cadence that is cued by the second shot, a close-up of a decaying tombstone bearing the inscription, "As soon as I heard of American independence, my heart was enlisted."

Perhaps the most brilliant formal stroke in In the Year of the Pig occurs at the conclusion. Daniel Berrigan is heard in intermittent voice-over delivering poetic testimony to the strength and moral rightness of the North Vietnamese cause. His words, along with those of newspaper editor Harrison Salisbury, are accompanied by fructifying images of valiant peasants planting crops and harvesting an abundant catch from the sea. There is a close-up of an overflowing basket of fish with their mouths gasping open. It is a complex image, signifying on one level the health and ingenuity of the Vietnamese people and, on another, the gasping futility of America's military aggression (it also evokes Mao's famous dictum about guerrilla fighters as "fish in water"). The double valence of this image is amplified when, after several shots of Ho and cheering peasants, de Antonio introduces for the first and only time shots of dead and wounded American soldiers. The most powerful image is of a blindfolded GI being dragged from the jungle by his buddies. It is followed by litters being hastily loaded onto choppers. The metaphoric blindness--moral and political--of the American cause, in juxtaposition with the symbolic vigor of the North Vietnamese, can only lead to ultimate defeat. The final shot of the film is a split-screen image, one half in darkness and other featuring the initial shot of the Civil War soldier, this time on negative film stock. The affective impact of this sequence is stirring, mournful, angry, and its processes of signification have little do with the recruitment of found footage to illustrate historical events. De Antonio's example should serve as a lesson to documentary scholars to be more mindful of the expressive, figurative dimensions of collage.



1. Andre Bazin, "The Ontology of the Photographic Image," What Is Cinema?, trans. Hugh Gray, (Berkeley: University of California Press, 1967): p. 16.

2. Bazin, pp. 41-52.

3. Jay Leyda recounts the story of how in 1898 Lumière cameraman Francis Doublier constructed a précis of the Dreyfus affair out of a group of generic shots--a military parade, a government building, a departing boat--none of which bore any direct relation to the "depicted" event: Films Beget Films (New York: Hill and Wang, 1964): pp. 13-14.

4. The British propaganda films of Len Lye, in their avant-gardist denaturing of original images, can be said to bridge the two approaches.

5. I have avoided the contested arena of "subjective" versus "objective" point-of-view, although my allegiances are clearly with those who view the vérité method as a form of de facto or "naive" subjective inscription--a program whose contradictions are refracted and exposed by the reflexive subjectivity in breakthrough avant-garde works by Stan Brakhage, Bruce Baillie, and others during the same period. Limitations of space and thematic focus prohibit a more detailed critical account of the vérité philosophy: for further discussion, see for example, Thomas Waugh, "Beyond Vérité: Emile de Antonio and the New Documentary of the Seventies," reprinted in Movies and Methods, Volume II, ed. Bill Nichols (Berkeley: University of California Press, 1985): pp. 233-57; see also my "Jargons of Authenticity (Three American Moments)," Theorizing Documentary, ed. Michael Renov (New York: Routledge, 1993): especially pp. 118-26.

6. Michael Renov ably discusses the impact of European historiography in "Early Newsreel: The Construction of a Political Imaginary for the New Left," Afterimage 14.7 (1987), pp. 12-15.

7. Matthew Teitelbaum, ed., Montage and Modern Life (Cambridge, MA: MIT Press, 1992): p. 7.

8. To cite but two examples, Peter Burger makes collage the centerpiece of his Theory of the Avant-Garde, trans. Michael Shaw (Minneapolis: University of Minnesota Press, 1984), and Thomas Crow discusses the integral relation between modernism and the importation of devalued artifacts in "Modernism and Mass Culture in the Visual Arts," Modernism and Modernity, ed. Benjamin H. D. Buchloh, Serge Guilbaut, and David Solkin (Halifax: Press of the Nova Scotia College of Art and Design, 1983): pp. 215-64.

9. Theodor Adorno, Aesthetic Theory, ed. Gretel Adorno and Rolf Tiedemann, trans. C. Lenhardt (New York: Routledge and Kegan Paul, 1984): pp. 233-4.

10. Vertov, who treated immediate footage gathered from disparate sources as if it were, in the usual sense, "found," analyzes the potential of heterogeneous montage in a number of essays: see Kino-Eye, ed. Annette Michelson, trans. Kevin O'Brien (Berkeley: University of California Press, 1984)..

11. See, respectively: Siegfred Kracauer, Theory of Film (New York: Oxford University Press, 1960): pp. 194, 202; Parker Tyler, "Documentary Technique in Fiction Film," The Documentary Tradition, ed. Lewis Jacobs (New York: Hopkinson and Blake, 1971): p. 254; Bela Balazs, Theory of the Film, trans. Edith Bone (New York: Dover, 1970): p. 165.

12. See, respectively: Bill Nichols, Representing Reality (Bloomington, IN: Indiana University Press, 1991): p. 127; Waugh, p. 243.

13. Hayden White, "The Value of Narrativity in the Representation of Reality," Critical Inquiry 7.1 (Autumn 1980): pp. 8, 10.

14. See, respectively: Pratibha Parmar, "The Moment of Emergence," Queer Looks: Perspectives on Lesbian and Gay Film and Video, ed. Martha Gever, John Greyson, and Parmar (New York: Routledge, 1993): p. 11; Kobena Mercer, "Diaspora Culture and the Dialogic Imagination," Blackframes, ed. Mbye B. Cham and Claire Andrade-Watkins (Cambridge, MA: MIT Press, 1988): p. 59. The critical discourse on found footage in avant-garde cinema largely complements the position expressed by Parmar and Mercer. See for example, William Wees, Recycled Images (New York: Anthology Film Archives, 1993).

15. Gary Crowdus and Dan Georgakas, "History is the Theme of All My Films: An Interview with Emile de Antonio," Cineaste 12.2 (1982), p. 22.

16. I am extrapolating this stance from the films and statements of American and British experimentalists, recognizing that the line between avant-garde and conventional documentary has become increasingly blurred. The formalist approach to found footage is exemplified in the work of Craig Baldwin, Keith Sanborn, Leslie Thornton, Tony Cokes, Martin Arnold, and Isaac Julien, among others. For an instructive collection of critical statements by filmmakers, see the appendix to Recycled Images, pp. 65-100.

17. The insertion of the phantom Japanese man-on-the-street is reminiscent of a virtual leitmotif in the "Why We Fight" series in which the exact same piece of footage of a woman in a kerchief rushing anxiously along bombed-out street in used in three different episodes supposedly set in three different countries.

18. See David Bordwell, Narration in the Fiction Film (Madison, WI: University of Wisconsin Press, 1985): especially pp. 31-48.

19. Waugh, p. 250.

Paul Arthur

Paul Arthur has been writing about nonfiction film since 1972. He has published in journals, magazines, catalogues, and book anthologies. He has also made more than a dozen experimental films and is a former President of the Board of the Film-Maker's Cooperative in New York.