Network Analysis of Literary Translation in the U.S.

Tom and I both attended HILT a few weeks ago, taking the week-long Introduction to Network Analysis course with Elijah Meeks. We wanted some quick data to work with, so I pulled the spreadsheets from Three Percent’s Translation Database, a data set of world literature translated into English in the U.S. market. The data set includes information about the titles, such as authors, translators, publishers, original language, country of origin, and so on. (Three Percent, at the U of Rochester, is also a great place to learn about new translated literature and read some great reviews.)

Once we began playing around with the data, we thought there might be something worth exploring here – we wondered if there were distinct communities within the languages? might we detect any gender biases? what, if any, are the significant differences in the communities around different publisher types (say, academic, big publishing house, or small press)? In any case, the data needed (still needs) lots of work to be useful in answering some of these questions, but by exploring the existing data and testing the network analysis waters, we now have an idea how we might proceed.

The slides below outline some of our steps thus far and immediate plans. There are also a few visualizations resulting from initial attempts at detecting communities and important nodes within the German translation network – no surprises so far!