Televisual Time


In 2016-2017, I worked as a Research Assistant on Televisual Time (REJ), a DH project organized by Assistant Professor of English Joel Burges. As this was a brand new project, we began at the ground level and spent much of the year thinking through theoretical and methodological questions. For Burges, the magazine was compelling as a physical instantiation of television’s temporal structure that, as a result, might be seen as a physical instantiation of time past. At the same time, it gives a zoomed-out perspective on TV that makes it akin to the kind of “distant reading” common to data-based DH analysis. With these comparisons in mind, what might a large-scale digital analysis of TV Guide tell us about the medium of television and, particularly, its way of structuring time?

Technology

This project introduced me to the necessary—and often frustrating—preliminary steps involved in getting an archived-based DH project off the ground. The original plan was to use OCR to read digitized copies of the magazine. First question: are the materials already digitized? At what cost? Although a digitized archive was found to exist, it was beyond the budget of such a new project.

Our next attempt was to use microfilm, which we were able to procure via Interlibrary Loan. However, these materials turned out to be inadequate for our uses, since the magazine had been recorded at a very small scale: between 2 and 4 issues vertically per 35mm microfilm reel. Once scanned and viewed as PDFs, the text of the schedules was unintelligible to the computer, and often to us as well.

Next, we turned to the magazine itself, and thus to eBay, where we were able to purchase a selection of issues from each decade. After scanning the magazines, we were able to OCR most of the text, but many of the TV Guide-specific elements, such as times and channels, were not picked up. The evolution from a two-column layout to a grid and the corresponding shrinkage of text brought diminishing returns.


Televisual Genre

Given these setbacks, I decided for the rest of the semester to see what we could gather from the OCR files, rather than dwelling on what we couldn’t. I thus found myself turning away from time and toward more accessible questions of genre. From 1952, when the magazine was started, through the early 2000s, TV Guide included genres alongside many of its listings. These genres, however, were inconsistent, both in their identity and their distribution. Using text-searching software, I collected data about genre and published the preliminary results as a research report on the Mellon website. For me, the most interesting outcome was the idea of calculating genre distribution in minutes: how long, for example, one could spend watching all of the comedies on air in a given week (hypothetically or with the aid of recording technology). For 1998, this number—13,560 minutes, or 226 hours—far exceeds the total number of minutes in a week, being 10,080. An odd comparison, to be sure, but intriguing nonetheless.


TV in the Seventies