Associate Professor Jonathan Brennan was recently awarded an NSF grant, in collaboration with Edith Kaan (University of Florida) and Teresa Bajo (University of Granada), for a project titled “Predictive processing in naturalistic language comprehension through EEG and computational modeling.”
This project investigates how humans understand language by predicting what a conversation partner might say next, in the context of proficient second-language comprehension. Dr. Brennan and his collaborators aim to study brain signals while bilinguals listen to an audiobook story in their first and second languages, and use computational modelling (with AI language systems) to test the kinds of predictions people make, and how those are affected by differences in language background. This offers insight into how AI can incorporate multiple languages in a realistic way and additionally increases awareness of bilingual language with programs targeting educators and the general public.
The project collects electroencephalography (EEG) signals from three groups of bilingual participants with different levels of experience, which reflect fast-changing brain responses that are highly sensitive to expectations in language. AI is used to capture linguistic features of the stories the participants listen to, with statistical analyses being used to test which of these features best capture brain activity and how this may be different for individuals with different language backgrounds.
The full details of the grant can be found on the NSF website.
