CTDH Conference

Chair: Sara Sikes

Jim Cocola, "All the Feels: Affective Analysis for Literary Studies"
When evaluating matters of content, language, and style among writers, machine reading can help humans discern patterns too arduous to tabulate by hand. In this paper (which could also be remediated as a demonstration) I employ natural language processing through the use of regular expressions in the Python programming language in order to conduct affective analyses across a series of literary genres: modern drama, the modern novel, modern poetry, and the slave narrative. While a well-established tradition of stylometry has employed similar methods toward the end of authorial attribution, and recent work in the digital humanities has focused on large corpora (Jockers) or single texts (Moretti), my efforts turn on questions of comparative literature concerning specific aesthetic movements, national traditions, and subject positions. Whereas sentiment analysis employs text mining methods in pursuit of coarser distinctions (positive, neutral, negative), I aim at a broader and deeper parsing of affect, drawing on classifications by leading psychologists in the field (James, Lazarus and Lazarus). In this vein, machine reading enables questions that may not even occur to human readers alone. Is modern drama happier or sadder in America or Britain? Which modern novelists—in emphasizing pain over pleasure—have the most masochistic tendencies? Are Romantic poets more likely to emphasize singular over plural forms of feeling? Do slave narratives by men contain more anger, fear, and rage than slave narratives by women? Natural language processing promises to teach us much about the ways in which writers use language to express affect. I aim to offer an overview and exemplification of that potential in these remarks. In the process, I will also attend to classroom applications in a seminar I'll be offering concurrently at WPI titled "Textual Engineering," possibly including contributions from one or more students enrolled in that seminar.

Yohei Igarashi, "Philological Data"
This paper returns to the early decades of English as a professionalized, departmentalized university discipline, and recovers how American philologists undertook an early and significant datafication of literary works. Specifically, my paper considers The Concordance Society (founded at the 1906 MLA meeting by the philologist, Albert S. Cook), which went on in the ensuing decades to produce collaboratively a series of concordances, via paper slips and other pre-electronic media, primarily for canonical vernacular poets. We know already from disciplinary histories that philologists dominated the discipline in the later nineteenth century – wielding more professional power than belletristic, appreciationist instructors – due to their credentials and the then-reigning disciplinary ideal of scientificity. We know too from histories of humanities computing that a landmark moment in “literary data processing” was the 1964 conference of the same name, organized by English professors Jess B. Bessinger, Jr. and Stephen M. Parrish in conjunction with IBM computer scientists. But in between these two moments, The Concordance Society – founded when philology was beginning its decline, but before the electronic literary data processing projects that would follow in the Society’s footsteps – is particularly instructive for us today with regard to the notion of literary data. This paper demonstrates that The Concordance Society negotiated between two views of literature: literature viewed as aggregable documents capable of yielding data, information, and facts, and literature viewed belletristically as artworks. By relating Lane Cooper’s 1911 Concordance to the Poems of William Wordsworth to later 19th-century philological scholarly practice, and drawing on recent theories of data, this paper suggests that the original tension between philological datafication and belletristic appreciation runs through the history of literary study – and that this tension can grasped not as an irresolvable, enduring conflict but as pointing to the ontological flexibility of literary artifacts as a modern disciplinary object.