I’m a recent PhD graduate from the Interactive School of Computing (in the Human-Centered Computing program) at the Georgia Institute of the Technology. I was advised by Jacob Eisenstein and Diyi Yang, and my research explored how language variation reflects communication goals in online discussions.

Language reflects and constructs society, and nowhere is this more obvious than the social spaces of the internet. The fast speed of communication in online communities leads to frequent changes in communication style. In general, the volume, velocity and variety of data available from social media can help address complex open questions in sociolingustics. My dissertation explored three such open questions: (1) How do people vary language when their listeners are only partly known? (2) How do speakers choose new words to adopt and which to ignore? (3) How do speakers’ social attitudes explain their language choices? Through natural language processing and statistical analysis of social media data, I found that (1) people adapt to the expected norms of their communities and, even in a crisis, to the expected norms of their ever-changing audience; (2) linguistic utility, more so than social spread, explains the adoption and abandonment of words; and (3) social attitudes related to politics are strongly tied to multilingual language choices while cultural attitudes have a less clear connection (work in progress). This research revealed the rich opportunities available through social media for sociolinguistic analysis, particularly for under-studied areas.

You can reach me at istewart6 at gatech dot edu. In my free time, you’ll find me making music and baking cookies.