I’m thrilled to have been awarded one of the four inaugural faculty fellowships in AI and Knowledge at the new Center for Humanistic Inquiry into AI and Uncertainty, housed within UW-Madison’s Institute for Research in the Humanities. My proposal “Linguistic Expert Knowledge / Knowledge Experts in the Digital Age,” is — no surprises here — related to my book project “Bad Grammar in the Digital Age.”
I will compare different versions of the identity of a “language expert,” beginning with the gulf between the descriptive, science-based approach taken by most linguists and a prescriptive, skill-based approach taken by language professionals (such as writers, editors, translators, and journalists). Both identities are threatened by the easy availability of tools that produce language that is grammatically correct and situationally appropriate without being anchored in a speaker’s identity or communicative intent. As AI is democratizing linguistic performance, the role of people whose linguistic knowledge and judgment we trust is receding. What does this mean for the public discourse between experts and laypersons — in linguistics, but also beyond? What happens to linguistic theory when the data on which it relies — authentic human utterances — becomes increasingly intermingled with synthetically produced text, text that is known to have been produced with algorithmic bias towards the regular, the formulaic, the stereotypical?
I look forward to discussing these and other questions with my fellow fellows and colleagues in the Uncertainty and AI group in the course of the next two semesters.