Emory University / ENG 790 & QTM 490 / Spring 2023
One of the primary strands of digital humanities scholarship is quantitative literary analysis: the use of quantitative (and other digital) methods to study literary texts. These practices raise many major questions: What does large-scale analysis reveal about literature that cannot be discerned by reading alone? What happens when literary texts are converted into numbers, as they must be for any quantitative analysis to take place? What does it mean, both ethically and intellectually, to borrow methods developed in the sciences and/or from industry for literary and cultural studies scholarship? And what of the core concerns of literary and cultural studies? Can ideas about language, metaphor, style, labor, and power, among many others, be quantified, modeled, and/or otherwise explored at scale?
This course will take on these questions in both theory and practice, focusing our inquiry around a current state-of-the-art approach to quantitative literary analysis (and quantitative text analysis more broadly) involving what are known as large language models (LLMs). We will begin by contextualizing the advent of LLMs within the history of language modeling and predictive text generation, primarily in the area of electronic literature, before delving into the conceptual and technical details of LLMs as a class. This shared inquiry will enable us to think through possible projects in the area of quantitative literary studies that might engage the strengths–or, alternately, probe the limits–of LLMs as a method of analysis. The course will culminate in final projects undertaken in small groups, consisting of students skilled in both literary analysis and quantitative methods, which demonstrate how quantitative analysis can contribute to current scholarly conversations in the field of literary studies, and vice versa.
- Syllabus
- Course Calendar
- Course on Canvas (Emory students only)
- Email to Lauren or Ben!