cSCAN Rounds

Tracking the dynamics of referential activation with EEG

We report results from a project investigating whether the dynamics of neural signals can be used to track the moment-by-moment activation of referents during language processing. We trained classifiers on EEG data to predict which of two images a participant was viewing, and then applied these classifiers to data recorded while participants interpreted questions about one of these objects. Our results suggest some promise for this new approach.