Scott Weingart (Indiana University) wades into the conversation about Syuzhet, distant reading, and “the great unread” in a recent blog post, “Not Enough Perspectives, Pt.1.”
Syuzhet is an R package created by Matthew Jockers (University of Nebraska, Lincoln) and used to extract sentiment in text analysis, which became the subject of much discussion, in part spurred by Annie Swafford’s (SUNY, New Paltz) critique of the package last month.
Citing Andrew Piper’s statement that “Right now DH is all texts, but not enough perspectives,” Weingart asserts that the digital humanities “suffers from a lack of perspectives in two ways: we need to focus more on the perspectives of those who interact with the cultural objects we study, and we need more outside academic perspectives.”
The important point is that reading at scale is not clear-cut. This isn’t a neglected topic, but nor have we laid much groundwork for formal, shared notions of “corpus”, “collection”, “sample”, and so forth in the realm of large-scale cultural analysis. We need to, if we want to get into serious discussions of validity. Valid with respect to what?
Those looking to catch up on the Syuzhet conversation will find Eileen Clancy’s (The Graduate Center, CUNY) two-part storify (I and II) immensely helpful, and Ted Underwood (University of Illinois) has just added a further response in his latest post, “Free Research Question About Plot.”