Rather than ask “What is cinema?” Deleuze asks “What does cinema do?” and “How does cinema construct concepts?” Implicitly taking up a position bordering on Bazin’s “Total Cinema,” Deleuze avers that “the cinema is always as perfect as it can be”; thus to determine what is “good” or “bad” cinema needlessly complicates the issue of determining how cinema provokes thought through image-formation. For Deleuze the filmic image, be it a “digital” or an “analog” one, never descends into mere representation — the cinematic image is not an imitation or mimesis, “there is nothing behind the image,” because the image, thing-like, already inhabits space-time as only itself and nothing else. Hence, the controversy of a photographic cinema versus a post-photographic cinema does not for Deleuze necessarily exist, if we consider that both cinemas, practically any cinema, whether in the boldly reality-infused films of directors like Abbas Kiarostami or Jia Zhangke, or conversely in the Pixar films rendered completely through CGI, the cinematic image constructs a movement, an unfolding in time, that doubles as a way of seeing or thinking in the world. For Deleuze, the image inhabits the world, and in the image we inhabit the world. But there is not just one image, but multiple: “perception-images,” “action-images,” “relation-images,” “crystal-images,” “affection-images” (Rushton offers the possibility of a “digital-image” or “silicon-image” for the post-Deleuze age). In any case, these images line up under the general rubric of the movement-image and/or the time-image: what then differentiates these from each other? Rushton succinctly, and ably, categorizes Deleuze’s movement/time image dichotomy for us:

Since for Deleuze (who acquires his principal concept of the movement-image from Henri Bergson’s “three theses of movement”) time is “change or it is nothing at all,” the movement-image fails in some respects to embody time in all its fluctuant multiplicity since it is “defined in terms of actions and reactions, which aim to solve problems….” The time-image, on the other hand, gives off “a little time in the pure state” because it is fundamentally open to change: “The solutions to the problems established by time-image films are not solved in determinate ways — rather, their solutions are left open. This openness… means that, at a first level, the future remains open to change. That is to say, what has happened in the past need not determine ways of conceiving of the future. In short, a direct image of time is not just one in which the future is left open, but is one in which the past, the present and the future are all open to change.”

Deleuze argues that the movement-image came of age, and into ubiquity, during the early cinema industries that established themselves, respectively, in France, Germany, the Soviet Union, and the United States, before World War II. The movement-image would be exemplified by the films of directors as foundational as Eisenstein, D.W. Griffith, Abel Gance, Fritz Lang, John Ford, Elia Kazan, Howard Hawks, and Alfred Hitchcock. The films of each of these directors (albeit in different ways and on divergent scales) are modeled on a general movement-image system that through shot, frame, and montage techniques attempts to rectify or bring into rhythmic order a series of actions and reactions that revolve around a key problem, or array of problems, in the film scenario. The “past” in such films stands forth as that against which the present has to measure or reform itself, or rather, the past is presented as that which has to be severed or distinguished as apart from the present — the present and the past are never fully in accord, nor truly latent in each other, since the past in the movement-image film is always in some need of being resolved and/or “canceled out.” (This is why genre-cinema is typically, but not exclusively, a cinema of the movement-image.)