A Serious Lapse
The study and analysis of film is not something that is easily quantified. Film is, after all, an art, and even the most mathematical forms of artistic expression (not sure what that would be – Op Art, maybe?) do not yield all their richness of meaning and nuance when viewed through a strictly quantitative lens. Music, probably the most mathematical of arts, suffers immensely if the analysis of it is restricted to the numerical patterns that it expresses by shifting from note to note and chord to chord. The mathematical relationships that guide any piece of music are of course the basis for the piece’s structure and affective qualities, but they certainly are not the whole of it.
By and large, it’s hard to quantify film – far harder than it is to quantify music, even. There are just so many variables: shot duration, on-screen movement, color, the hugely complicated soundtracks of many films, camera movement – and then the meta-patterns that result when you map any one of these artistic systems atop another. You can count frames, if you like, but just try to come up with a quantitative notation system for charting the innumerable, minute differences from one to the next.
In analyzing visual style, one important metric is shot scale: how far is the camera from the subject that it’s photographing? Shot scale is a notoriously imprecise system of measurement: it’s based on the scale of the human figure, which of course is anything but constant. Terms like close-up, medium shot, and extreme long shot refer, really, to a fairly wide range of shot compositions. And yet, in studying the stylistic patterns that comprise a film, it’s very important to take note of such sliding-scale features as shot scale, camera placement, soundtrack density, and so forth, even if that entails a tacit understanding that such measurements are necessarily rough. There may be a master kilogram in the vaults of the International Bureau of Weights and Measures, but there is no single master medium shot, for instance. As I tell my students all the time, this is the humanities, people. There ain’t no absolutes.
This is not to say that some scholars have not taken a quantitative approach to film studies, and with some solid, telling results. The most prominent such effort is probably Yuri Tsivian’s Cinemetrics; other notable online efforts in this regard include Shot Logger and my fellow WordPresser Nick Redfern, whose blog Research Into Film is devoted to “an empirical approach to film studies.” I am fully in favor of these projects, which not only inject rigor into the field, but yield strong results that are essentially scientific in the true, original sense of that word: related to knowledge. The more closely a text or a phenomenon may be observed and measured, the more we may know about it, and the stronger will be any arguments we may wish to make about it.
Imprecise as the study of this medium may often be, certain cinematic features may nevertheless be measured. If you have a really good eye, like my friend Patrick, whose book you really ought to buy, you can tell, just by looking at a shot, the lens that was used to shoot it. (I cannot do this.) Knowing the precise measurements of the taking lens will indeed provide you with some mathematically solid, quantifiable information that may be charted over the course of a film.
One of the most common and useful metrics for the film scholar is that of the rate of editing, which can be counted (at least with celluloid film) in easy units: for how many frames does a given shot remain on the screen? Because film passes in front of the projecting lens at a constant speed, shot duration can be measured with accuracy. By counting the number of shots in a film, precisely measuring that film’s running time, and then doing a teeny bit of division, one may discern a film’s AVERAGE SHOT LENGTH, or ASL. Studying ASLs can reveal quite a bit about the style of a scene, of a film, or of a director’s overall style.
But I have a problem with the ASL as a tool. I take issue with its name.
Since it is such a useful metric, many film scholars perform ASLs for the films that they study. When I study a film closely, the ASL is one of the first, most basic operations I perform in analyzing it. But I hereby propose – by whispering into the vast chasm of the internets – that this metric be renamed ASD: Average Shot Duration, for this is what it is actually measuring.
In my Intro to Film classes, even the students who fully grasp the difference between shot length (another term for “shot scale,” e.g., medium shot, extreme close-up, etc.) and shot duration (how long a shot remains on the screen) often confuse the two terms. I even catch myself doing this sometimes, as well. I try to avoid the very term “shot length” – using “shot scale” instead – precisely because “length” is used in everyday English to refer to both distance and duration, the fact that is responsible for my students’ verbal slip-ups.
And so, since I like to teach my students that calculating a film’s ASL is a useful endeavor, and because I would like them to be as careful and rigorous in their scholarly pursuits as possible, I hereby and formally suggest that film scholars unite to henceforth use the term AVERAGE SHOT DURATION (ASD) instead of the scurrilous and imprecise term Average Shot Length (ASL). We can do this, people. Let the groundswell begin here!
Oh, and I checked Ze Google to see if “ASD” was an acronym for anything else. I am pretty sure that the Anchorage School District won’t mind if we use their letters. Ain’t a lot of overlap there.
Update (11 October 2011): A recent post includes some additional information related to this one. Please check it out!