Transformations in Film as Reality (Part One)

Questions Regarding the Genesis
of Nonfiction Film

Komatsu Hiroshi

As part of next year's commemoration of the 100th anniversary of the birth of film, Documentary Box will be running a four-part series of articles exploring the history of cinema's relation to reality. Each piece, written by a different film historian, will investigate how both documentary as a genre, as well as the "realistic feel" of cinema itself, have evolved over the last century. The Japanese film historian Komatsu Hiroshi leads off by taking up the problem of nonfiction in cinema's early days.

The Editors

Conceptually understanding that a certain film is nonfiction should presuppose knowledge of the period, place, cultural context, and ideology in which the film was made, as well as, most of all, the existence of film fiction

Most people today possess a vague image regarding "documentary cinema" (called, in Japanese, "kiroku eiga," or "record cinema"). That term was first widely used in Japan after the war; in the prewar, the expression "bunka eiga," a translation from the German "Kulturfilm," was the more popular term. Both expressions basically referred to the same film form and content: documentary film and bunka eiga were generally considered modes which, in opposition to the constructed narrative of dramatic film performed by actors, transmitted truth and reality - while still bearing the filmmaker's artistic originality, unlike newsreel cinema. To offer a definition strict to film history, it is well-known that the term "documentary film" in a narrow sense referred first to British documentary cinema and to the word's creator, John Grierson. Grierson initially used the expression in an essay in 1926 to refer to Flaherty's films; he later clarified its meaning and developed it, offering the famous thesis concerning "the creative treatment of actuality." The expression apparently became standard terminology in England and America by the end of the 1920s; when Paul Rotha, for instance, spoke of the various forms of cinema in his The Film Till Now, published in 1930, he was already setting apart "documentary film."1 In the way the term "avant-garde film" was once used to refer to a specific group of French 1920s films, to this extent, "documentary cinema" could be redefined to indicate that set of films produced first by Grierson for the EMB and the General Post Office, which appeared on stage during the British Documentary Movement. (Bunka eiga could also be brought back to its original historical position in the same way.)

If we are to replace the expressions "documentary film" and "bunka eiga," transformed as they have been from proper into general nouns, with a term that possesses a certain clarity, one that is less vague and more theoretically and historically meaningful, we might be able to rely on the concept of "nonfiction film." The concept to which "documentary film" currently refers is, in the end, subsumed in nonfiction. Focusing on the problem of nonfiction film should also allow us to make an issue of the period before Grierson.

However, in the same way people have maintained a vague image of documentary film as only that which is not dramatic film, what nonfiction truly is has largely not been considered due to the tendency in film studies to avoid nonfiction film. While there may be reasons behind this - this essay is in part directed at explaining them - it is true that in the last few years some scholars (albeit very few) have come forth with the will to tackle the problem of documentary and nonfiction film. Unfortunately, these studies have not presented a satisfactory theory and the reason lies in the fact that most lack a consideration of historical origins. This neglect of history repeatedly restores the concept of film to a tacit understanding, an unquestioned assumption about the essence of cinema in which, in this case, documentary or nonfiction film are only used as convenient words to designate a given group of films. One can, for example, consider a recent essay by Michael Renov as a case that seemingly falls into error because of a lack of full historical inquiry.2 As a metacritical paradigm, he presents the four functions of the documentary text: 1) To record, reveal, or preserve; 2) To persuade or promote; 3) To analyze or interrogate; and 4) To express. These four functions, however, are neither peculiar to documentary (or nonfiction) nor exclude that which is not documentary. On the contrary, since cinema in general expresses in its shape the time, place, cultural context, and ideology in which it was created, these functions can also be what people normally read into dramatic films. Documentary film cannot be differentiated through these four functions; they can be thought of only as arbitrarily selected functions which can be read into any film.

To continue from a different angle, one can say that the source of the fallacy in many discussions of the documentary text is related in part to film studies' tendency to avoid nonfiction. In here we find why exchanging the terms "documentary" and "bunka eiga" with "nonfiction film" produces both precision and, in seemingly reverse proportion, certain difficulties. This is directly connected to the issue of film history: what one can call the gap between the clarity of concept and the imprecision of representation - and/or the discrepancy between the idea and how it is manifested in actual forms of representation.

The negation signified by the prefix "non" in the term "nonfiction" is conceptually very clear: namely, that nonfiction is what is not fiction. However, in the case of actual forms of representation, this kind of accuracy through antonyms creates problems. Since nonfiction is a notion bearing negation in its prefix, conceptually it should naturally presuppose the object of its negation, fiction. We can ask, however, just who was it who held such a conceptual presumption, and at what point in time? Presumably, it was those who possessed knowledge of film fiction. If not, then the concept of nonfiction film could not exist. Consequently, in terms of conceptual origins, it is clear that fiction must have predated nonfiction. Yet as I have argued elsewhere, concepts are often produced in film history after forms of representation3; that is, it is usually the case that conceptual speculation occurs significantly after perception and cognition in film history. Conceptually understanding that a certain film is nonfiction should presuppose knowledge of the period, place, cultural context, and ideology in which the film was made, as well as, most of all, the existence of film fiction.

If one reasons from this position, one could construct the hypothesis that cinema at the earliest stages might have been fiction, not nonfiction, but this would have to be a fiction different from that developed later on. One could also alternatively assert that the first films were provisionally nonfiction, but since the prefix "non" should have already presupposed fiction, this falls into the temporal contradiction in which nonfiction negates a fiction which is supposedly produced at a later stage. If one declares that the earliest films are nonfiction opposed to a presupposed fictional form from outside of and predating the cinema, this bears the contradiction in which film produces its own form by negating a different media. In the end, it is the gap between concept and form of representation that must be problematized. Looked at in this way, the concept of nonfiction, which we believed to be useful, turns out to be purely a concept intended for managing perception and cognition. That is, this kind of concept does not at all guarantee the objective character of representation.

Nevertheless, when making an issue of cinema before Grierson, especially early film, a consideration of the characteristics of nonfiction which the discourse of film studies has attempted to avoid can be meaningful even if nonfiction is purely a concept of convenience. By closing in on its truth, we can perhaps obtain the key that unlocks the reason why most people hold a vague image of documentary cinema. As a start, we should look at an example of a discourse that does not presume the concept of fiction. The following quote is from a Japanese text on film history written in 1910.

The moving pictures at that time were still primitive, and admission was from several tens of sen up to one yen, but I think audiences were still very satisfied. I must admit, however, that I was in fact one of these spectators. I can remember it well even now. In that era, there were probably no pictures over one or two minutes in length. Since they were films in which the beginning and end were connected in a loop, the same image was repeated over and over again. As films, they were just one step ahead of magic lanterns, with many simple pictures of marching soldiers and children's fights - a length and assortment of images similar to that of the present-day Mutoscope. In only a few years, however, longer films gradually appeared, with many featuring fires, scenery, hunting, horse racing, and especially magic. But in those days one could largely not see the slapstick or dramatic pictures which are very much in fashion today, nor reverse photography, which can be accomplished only in the moving pictures. For better or for worse, the Russo-Japanese War then put an end to the existing variety of topics and the movies, as can be imagined, immediately shifted to war subjects. But after 1906 and the securing of peace between Japan and Russia, the war topics, which had prospered before, gradually fell into decline and slapstick, comedy, tragedy, fairy tale, and historical dramas took their place. It was at this time that the Pathé Frères of Paris, France, first came to the fore.4

Apparently, the author of this article does not possess a cinematic point of view presupposing the concept of fiction; accordingly, the idea of nonfiction does not exist for him either. What exists in its place is the difference between the "simplicity" of the earliest films and the contemporary fashion for probably longer and more complicated slapstick and dramatic films. While the Russo-Japanese War films were, without a doubt, either actual war footage or so-called "fake documentaries," they were all marked as "war subjects" along with war-related slapstick and dramatic films. I have already pointed out elsewhere that to the eyes of spectators in 1905-1906 fake documentaries definitely looked like counterfeits as opposed to genuine war scenes.5 But this does not necessarily mean that the majority of people in that era retained a conception of nonfiction film. Even in 1910, when the above quote was penned, discourse on cinema was already dependent on a classification system existing at the end of the 19th century. This form of film classification, printed from the start in sales catalogs for motion picture companies, had proved useful in managing theater programming. The average exhibition bill at movie theaters around 1910 - the period just before film length began to grow - was composed of a set of 6 to 10 films of one or two reels, in which comedy, tragedy, landscape, and travel subjects were mixed together. In France, Pathé Journal began systematically producing a form of newsreel between 1908 and 1909; a new trend thus appeared in which cinema began to take on the characteristics of a newspaper or magazine - that is, it became more linguistic. But in Japan, systematic news reporting through images remained unfeasible in practice even during the Russo-Japanese War, when film war reports were shown in droves; the increased interest in cinema produced by the war films did not necessarily develop into the genesis of a conception of film fiction. The old form of classifying film remained dominant even up to the first half of 1914, just before the First World War; a different way of conceptualizing film - that is, one featuring the binary consciousness of true or false, fiction or nonfiction - was apparently uncommon. Even though the appearance of full-length films gradually altered the composition of the exhibition bill, distinguishing clearly between the attraction - the so-called "feature" - and the "added" short films, the classification system cinema continued to borrow from the catalogs of magic lantern shows still did not qualitatively separate landscape, educational, and industrial films from those that were historical or artistic.6 In other words, it was a situation in which all films continued to be received as homogeneous.

It was the shortage of actual war films during the Russo-Japanese War era that gave birth to fake documentaries. Among these there were both fakes which could potentially be mistaken for true battle scenes, as well as trick films of sea battles which obviously utilized miniature models. However, authenticity on the level of film reception was not that important; all these works were part of a homogeneous filmic representation. When spectators got angry that the war images unfolding before their eyes were false, their anger was in no way based on a conceptual opposition of nonfiction to fiction, but was simply directed against the fact the representation was inaccurate. The case of early Taisho (1912-1926) film audiences judging French-made "Japan dramas" as lacking accuracy is similar. This perception of inaccuracy also resembles what American spectators of about the same period experienced upon viewing European-made Westerns (a feeling we Japanese could not normally share). There certainly existed a mode of judging truth and falsity that differed completely from the dichotomous discrimination between fiction and nonfiction. In this context, even sea battle footage utilizing miniature models could be accepted as a "war subject" when there was a lack of actual Russo-Japanese War films. Regardless of the issue of whether or not audiences got angry viewing them, these films in the end functioned as depictions of the war situation reported in the newspapers and magazines; to spectators, these moving illustrations - even as miniature models - became news reports possessing a certain reality, in addition to serving as affective objects which could heighten emotions.

Considering the above, it is clear that the concept of nonfiction film did not exist in early cinema (and/or in its audiences or texts), but was historically formed through numerous stages of film history, experiencing many transformations in its meaning, reception, and position in the exhibition structure. The issue of when the concept of nonfiction was determined must differ from nation to nation and cannot be located within a unified time frame. It is also certain that gaps concerning the meaning of nonfiction existed between producers, audiences, and journalists. One issue we can particularly consider is the fact that, as film became more linguistic (not semiotically, but on the level of history and consciousness), the analogy between film and language made the literary genres of fiction and nonfiction applicable to cinema.

To spectators of Japanese film, however, the separation of cinematic representation into fiction and nonfiction remained practically a non-issue for some time. Looked at from the perspective of American or European cinema, this must look like the preservation of primitivism. To offer the most famous example from the early period, we can say that Viewing Scarlet Maple Leaves ("Momijigari"7), shot by Shibata Tsunekichi, set the pattern for later Japanese cinema. This film, featuring appearances by Ichikawa Danjuro IX and Onoe Kikugoro V, is clearly a recording of a performance by two famous actors. In that sense, it bears a form of representational meaning analogous to that of later documentaries which have recorded stage performances on 16mm film or with a video camera. Yet Viewing Scarlet Maple Leaves is also evidently fiction in the context of Kabuki; it accords with what anyone understands to be dramatic film: a fabricated story performed by actors. The painted landscape backdrops as well as the costumes and gestures of the performers all maintain a diegetic function, to borrow a term from film studies. But although one can recognize the elements of fiction assembled here, it is also apparent that the properties people generally attribute to nonfiction cinema - the actual recording and preservation of the fact and details of a dance and performance - are still present in full force. Of course, from our viewpoint, the fact that Viewing Scarlet Maple Leaves exhibits both qualities is not too surprising a discovery, given that the development of the oppositional conception of fiction versus nonfiction itself occurred later in film history, and that the mode of representation in early cinema was defined by a form of absolute representation rejecting such a dualism. What is rather more surprising is the fact that this absolute representationalism became the pattern which continued to rule over later Japanese cinema. In fact, in most cases, Japanese film - in the genres of shinpa tragedy or of kyugeki8 - continued to deny the development of the notions of fiction and nonfiction. There of course existed several exceptions, ones which did not succeed to the Japanese theatrical tradition and instead followed the model of Western film, such as the shinpa action serials of around 1910 and some of the early Taisho "rensageki,"9 but these imitations in no way became the dominant pattern. This was because Japanese cinema, even into the late 1910s, opted to maintain an absolute representationalism that could not be regarded as either fiction nor nonfiction. It did this through continuing to produce films as moving illustrations of well-known stories, to use intertitles only as the titles of scenes composed at the screen writing stage, to show an aversion to American cinematic illusionism, and to make the story depend on the patterned acting of the performers and on the detailed narration of the benshi.10

Japanese cinema continued in this unique state up until the 1910s, leaving the field of what was regarded as nonfiction cinema, while not absent, at least inactive. At the same time in Europe and America, the concept of nonfiction was clarified entering the 1910s as non-dramatic films were shot in numbers equal to or even surpassing that of dramatic cinema. The First World War, in particular, lent energy to reportage films. Yet despite this, it is still scholarly difficult to separate the concept of nonfiction from fiction given that what guarantees the nonfictionality of nonfiction is, in the end, not representation itself, but the articulated linguistic supplement. Conditions of exhibition can, for instance, make a certain film nonfiction: in that situation, various articulated languages are appended to the textual representation which mark the work as not fiction, such as people's word of mouth, references in newspapers, magazines, and advertising, or even the title given to the film itself. If the representational form itself also contributes to defining the film, it is probably because this representation exhibits the typical unusualness of nonfiction cinema: the representational signs such as people's iconography, subjects looking at the camera, camera slips, or the lighting quality, which are normal to nonfiction. Film history, however, proves that these signs do not fully guarantee nonfictionality.

The fact that U.S. and European nonfiction cinema throughout the 1910s constructed an illusionism that closely resembled fiction film in technique was confirmed at a workshop I recently attended held at the Nederlands Filmmuseum. If it was in this period (from about 1909-1918) that nonfiction film was fully conceptualized and its nonfictionality molded, then this process largely coincided with the formation of the classical Hollywood system based on illusionism. Ironically, this system, intended to increase the credibility of fiction, has now come to be distinctly recognized for its artificial and constructed nature as its characteristics have become clear. Is it not possible to say the same of fiction's conceptual opposite, nonfiction? Long after the distinct nonfictionality of nonfiction was molded in the 1910s, this nonfictionality has at present turned into an extremely idiosyncratic set of signs, estranged from truth and reality.

- Translated by A. A. Gerow



1. Paul Rotha, The Film Till Now (London: Jonathan Cape, 1930): pp. 65-66.

2. Michael Renov, "Toward a Poetics of Documentary," Theorizing Documentary, ed. Renov (New York: Routledge, 1993): pp. 25-35.

3. Komatsu Hiroshi, Kigen no eiga ("Cinema of Origins") (Tokyo: Seidosha, 1991).

4. Katsudo Shashinkai ("Moving Picture World") (July 1910): p. 21.

5. See Komatsu, Kigen no eiga, or Komatsu Hiroshi, "Some Characteristics of Japanese Cinema Before World War I," Reframing Japanese Cinema, ed. David Desser and Arthur Nolletti, Jr. (Bloomington: Indiana University Press, 1992): p. 238.

6. See "Kinema Record Overview of Films: From the First Week of January, 1914, to the Last Week of May, 1914," Kinema Record (July 1914).

7. Translator's note: One of the earliest films shot in Japan by Japanese, Viewing Scarlet Maple Leaves was made in 1898 to record the performances of two of Japan's greatest Kabuki actors.

8. Translator's note: Early Japanese cinema was divided into two broad genres: kyugeki, or "old dramas," based on oral tales or Kabuki plays about feudal Japan; and shinpa, or "new school," composed of often melodramatic stories of modern Japan.

9. Translator's note: Rensageki, or "chain drama," was a form of stage presentation popular in Japan in the 1910s which featured an alternation of scenes performed on stage (usually indoor episodes) and ones shown on film (mostly outdoor scenes).

10. Translator's note: The benshi was the "narrator" in Japan who stood next to the screen and explained every film to the audience during the silent era.


Komatsu Hiroshi

Film historian, born in 1956. While enrolled at Tokyo University of Fine Arts and Music, studied abroad at the Film Studies Institute at Copenhagen University under a Danish government scholarship. Received a doctors degree from the Graduate School of Tokyo University of Fine Arts and Music. Author of Cinema of Origins (Seidosha). Presently teaching film history, aesthetics, and German cultural history at the Literature Faculty of Meiji Gakuin University, the Literature Faculty of Waseda University, the Literature Faculty of Kokugakuin University, and the Liberal Arts Faculty of Saitama University.