Zoom Link: https://mit.zoom.us/j/94817043024

Speaker: Yibei Chen (Postdoc Associate)

Affiliation: Senseable Intteligence Group

Title: Contextual priming modulates brain state dynamics during story listening

Abstract: Narrative comprehension involves dynamic interactions between stable cognitive frameworks and immediate contextual information. However, how context shapes brain dynamics during real-time story processing remains unclear. We used Hidden Markov Modeling to examine how contextual priming influences brain state organization during narrative comprehension. Participants (N=38) listened to an identical story in the scanner after receiving either an affair or paranoia context. Results showed a stable primary brain state consistently activating auditory, default mode, and language networks across contexts. Complementary secondary states exhibited context-specific modulations, particularly involving attentional and control networks. These modulations were predominantly driven by character speech, with differential processing patterns emerging based on character relevance to the contextual framework. In the affair condition, a control network-focused state (State 3) showed suppression during husband-character speech but enhancement during potential-affair-partner speech. In contrast, an attention network-focused state (State 2) activated strongly during the girl character’s speech. Out-of-sample behavioral data (N=112) further indicated parallel context effects, confirming that context shapes narrative inference. These findings illustrate that narrative comprehension relies on stable core brain states complemented by flexible secondary states that adaptively reconfigure to process contextually relevant narrative elements.