id_734. STRATEGIC MIND WANDERING AND THOUGHT CONTENT IN
NATURAL LISTENING
Michał Domagała1, Marcin Leszczyński1,2
1 Cognitive Science Center, Jagiellonian University, Ingardena 3, Kraków, Poland
2 Department of Psychiatry, Columbia University College of Physicians and Surgeons, 1051 Riverside Drive, New York City, USA
INTRODUCTION: Mind wandering (MW) refers to a shift in attention away from the task at hand toward internally generated thoughts. While traditional accounts conceptualize MW as a failure of executive control, resource-based theories propose it reflects strategic allocation of cognitive resources when external task demands are low. Evidence for such modulation has largely come from controlled, semantically and temporally impoverished studies.
AIM(S): We aimed to characterize how the content and temporal dynamics of reported experience change during MW in continuous, meaningful contexts. Specifically, we tested whether MW is strategically expressed during task-irrelevant (uncued) segments of a naturalistic narrative, and whether semantic coupling between verbal reports and the narrative decreases during these periods.
METHOD(S): Participants listened to a custom-designed auditory narrative alternating between two protagonists and were instructed to later recall information about one (cued) while ignoring the other (uncued). After each segment, they reported their attentional state using experience-sampling probes. In a second experiment, participants also provided verbal descriptions of their ongoing experience. Eye movements and pupil diameter were recorded. Using language processing methods (sBERT and word2vec embeddings) we quantified semantic coupling between verbal reports and the narrative.
RESULTS: We observed more frequent MW during uncued compared to cued segments. This pattern replicated across self-reports and independent analysis of their verbal reports. Semantic analyses revealed reduced coupling between the narrative and participants’ expressed content during uncued segments and MW episodes.
CONCLUSIONS: These findings demonstrate that MW can be strategically modulated in semantically rich environments. MW was also associated with reduced coupling to ongoing input, extending the laboratory-based accounts of strategic MW to continuous narrative comprehension.
FINANCIAL SUPPORT: This research was funded by a grant from National Science Center of Poland (Sonata 19: UMO-2023/51/D/HS6/02920).