Teaching Humanistic Data Analysis

This paper advocates for a “data humanism” that is distinct from both computer science and data science. Seeking to simply teach data science skills to humanities students misses the unique contributions our fields and our students might make to broader conversations about data and culture. The most important reason to imagine a distinct data humanism is that doing so in turn imagines students who will go into the world outside the academy and change it, rather than simply slotting into existing frameworks. This data humanism should be exploratory, iterative, and dialogic, and takes as its goal directing scholars’ attention to unexpected places in and outside digitized collections, while raising new questions about those collections and their absences. The paper concludes by offering four concrete recommendations for fostering humanistic engagement with data in the undergraduate and graduate classroom. These comprise starting with creativity, teaching using humanistic rather than pre-packaged data, foregrounding the investigation of corpora over computational methods, and likewise foregrounding the inculcation of a mindset for approaching data over any particular method.