I thought of this anecdote when I read about an upcoming conference at the University of Chicago Film Studies Center. It's titled "A Numerate Film History? Cinemetrics Looks at Griffith, Sennet, and Chaplin (1909-1917)," and it takes place this Saturday from 1 to 6 PM. The focus is on "the encounter between century-old films and computational statistics"—in other words, calculating the average shot length in American movies across a definite period to consider how visual-narrative systems evolved during that time. "Cinemetrics" refers to an interactive website launched at the U. of C. Neubauer Collegium for Culture and Society in 2005. It aims to "collect, store, and process scholarly data about films . . . to facilitate the analysis of film editing." Speaking at the conference are two Neubauer Visiting Fellows, professors Michael Baxter and Daria Khitrova, as well as U. of C. professors Tom Gunning and Yuri Tsivian.
Can computational statistics lead us to insights we can't attain through instinct? Or does numerical data get in the way of aesthetic analysis? The press release for Saturday's conference says the speakers will address Cinemetrics in terms of "possible promises—or traps" they present for film studies, so expect to hear the speakers make cases for both.