About

I was trained in British literary history; I teach courses in that subject along with courses in data science and digital humanities.

My active research interest is—to put it broadly—exploring the relationship between human cultural history and machine learning. It’s a relationship can be imagined in lots of different ways.

Most of the things I’ve published so far use machine learning to study literary history. But in that sentence, the verb “use” is perhaps a little slippery, or misleading—if it makes us imagine ML simply as a tool like a microscope that gets applied to a cultural object. I’ve found that machine learning is often useful, rather, because it can absorb a particular time- and place-bound human perspective, and reproduce that perspective in a reliable way.

There is perhaps a blurry line between that way of using models to study the past, and using language models in a generative way to reproduce or pastiche cultural practices. I suspect it’s going to be an interestingly blurry line.