The Moving Image 3.1 (2003) 110-131
[Access article in PDF]
Fragmention and Segmentation in the Lumière "Animated Views"
André Gaudreault with the Assistance of Jean-Marc Lamotte
[End Page 110]
Until quite recently, anyone reading the various histories of cinema would have supposed that the vast majority of "animated views" shot before 1900 conformed to the model known as that of "photogrammic continuity" and that any exception to this norm, the product of some sort of segmentation, fragmentation, or assemblage, would have been rare indeed. And yet this is not the case.
As our recent research into the body of Lumière and Edison films demonstrates, an impressive number of views shot before 1900 are made up of successive frames that constitute more than a single moment in time and thus contradict the canonical model.
These noncontinuous views—in which the sequence of their constituent frames is necessarily affected by at least one break—involve the attaching or adjunction of at least two distinct segments by means of a minimal form of "editing." The interruptions contained [End Page 111] in these views derive either from an explicit fragmentation of the strip of film, implying cutting and gluing, or a segmentation of the film "shoot" through "in-camera editing" that does not require any physical cut in the film itself.
This is the case for some 8.5 percent of the Lumière views produced in the years 1897-1899 1 (reaching a peak of 19.3 percent in 1899) and 34 percent of the Edison views produced during the same three years (with an astounding peak of 59.8 percent in 1899). 2 The results of our research thus contradict the long-dominant discourse on the supposed supremacy of the strictly continuous sequence of frames, a discourse for which the earliest films should, in principle, have provided evidence. Indeed, the very presence of such a large number of interruptions in the very earliest views might seem paradoxical. Such interruptions confound our expectations, given the context in which the Edison kinetograph and the Lumière cinematograph operated. For these were relatively rigid apparatuses that were, in their very conception, intended to capture homogeneous and continuous views. Their technological consummation fulfilled in a certain sense the more or less conscious dream of an entire generation to bring together, on a single, unified strip of film, a series of distinct frames that, despite their composite nature, would render the illusion of uninterrupted continuity. And doesn't this dream correspond, moreover, to the magic quality of the cinematic effect: that of a multitude of separate photographs that appear, to the eye of the observer, as if they are in fact a single photographic image, one, however, that had been brought to life?
The Lumière Company's output is confined to a ten-year period, from 1895 to 1905. During this time, the company produced at least two thousand views. Of these a certain number were never listed in their catalogs; these will henceforth be referred to as uncataloged views. The rest, the cataloged views, were listed in one edition or another of the Lumière catalog, which began to appear in 1897. The cataloged views have a more official status than the uncataloged views. 3 Their titles are relatively stable, for example, making it possible to follow their exhibition over time, in particular using advertisements in local newspapers. The uncataloged views are a sort of "secondary production," marketed without the benefit of a catalog listing. Their status remains somewhat ambiguous.
A survey of the different editions of the Lumière Company catalog makes it possible to identify 1,422 titles. Because of the existence of variants, these titles correspond to a total of 1,428 views. 4 In 1992, the Service des Archives du Film at the Centre National de la Cinématographie (CNC) in France began to catalog and restore the Lumière [End Page 112] archives, making it possible to reconstitute this body of work to the point that we now possess...