"What is documentary?" The question has been posed by innumerable scholars, and perhaps it can never conclusively be answered. Rather than yielding a simple response, it merely produces more questions: Does bias disqualify a text as "documentary"? Do documentary productions employ such tools as "actors" and "casting"? If a scene is staged by a non-actor, does it qualify as documentary? Is documentary defined by the maker's intent, or by the final product?
In his historical survey of the documentary, Jack C. Ellis defines documentary using the criteria of production, including the intent of the producer or director. Accepting impartiality as a myth, Ellis declares that "most makers" intend to "record and interpret" their subjects, as well as "inform and/or persuade" the audience (3). He cites the need for "nonactors ('real people' who 'play themselves')" (3). He identifies the makers' intentions regarding the experience of watching the film (as opposed to the lesson conveyed by the film): "the audience experience documentary filmmakers seek to provide is generally twofold: an aesthetic experience of some sort, on the one hand, and an effect on attitudes, possibly leading to action, on the other." (3) In Ellis's perception of documentary, the genre is defined by the intentions of the filmmakers; if they are unsuccessful in fulfilling these intentions, they may have produced a bad documentary, but it is a documentary nonetheless.
William Stott, on the other hand, defines documentary according to the audience experience:
"The essence of documentary is not information, as Grierson first thought. If it were, the classics of documentary cinema would be tweedle-dum 'industrials,' the worker-education films turned out by the hundreds. The essence, rather, is the same power to move that Grierson had all along sought in Hollywood's films; in them, he early wrote, 'I look to register what actually moves: what hits the spectator at the midriff: what yanks him up by the hair of the head or the plain bootstraps to the plane of decent seeing.' For he belived that emotion, properly felt and understood, does engender decent seeing; is intelligence." (11-12)
Stott admits that this definition of documentary is not popular, and likely never will be (12); after all, audience response is extremely difficult to gauge at any given historical moment, and much more so when one takes into account the factors of time (for public opinion changes over the years) and demographic (for an heiress in New York may or may not have the same tastes and opinions as an academic in San Francisco).
For the purposes of this project, I have adopted a definition of documentary somewhat closer to Ellis's than to Stott's. In choosing films, I have tried to consider the implied intention of the authors, while acknowledging, of course, the biases those intentions necessarily entail; nevertheless, I have not been overly concerned with the use of scripts or casting. While the government films are clearly meant to show the New Deal in the best light possible, the film produced by New York University seems, from this retrospective perspective, overtly driven by politics; the films commissioned by General Motors obviously serve commercial purposes, but I have not considered this to disqualify them as documentaries.
One thing is certain: the documentary is a phenomenon of the 20th century, and by the 1930s it was in popular use.