Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Does linguistic diversity shape cross-modal event segmentation? ERP correlates of event boundaries in Shipibo-Konibo
Poster Session C, Thursday, October 1, 10:45 am - 12:45 pm, Wangari Maathai
This poster is part of the Sandbox Series.
Natalia Morozova1, Roberto Zariquiey2, Balthasar Bickel1; 1University of Zurich, 2Pontificia Universidad Catolica del Perú
Humans segment unfolding events at different levels of granularity, identifying both coarse (large) and fine (small) event boundaries. Event segmentation has been documented in both visual perception and narrative comprehension and relies on tracking changes across dimensions such as time, space, causality, actors, goals, and objects. Previous research suggests that event segmentation is shaped by external factors such as event familiarity and culture. However, less is known about whether linguistic diversity shapes which event dimensions are tracked across populations during speech comprehension. Here, we investigate whether grammatical marking of event dimensions influences which event boundaries listeners perceive as coarse vs. fine, and whether linguistic event segmentation modulates attention to specific event dimensions during visual event perception. Specifically, we intend to compare event segmentation patterns in two culturally similar but linguistically distinct populations from the Peruvian Amazon: monolingual speakers of Spanish and speakers of Shipibo-Konibo, an indigenous Panoan language. Unlike Spanish, Shipibo-Konibo has an obligatory switch-reference (SR) system that marks changes in subject reference, temporal continuity, and action transitivity across clauses. For this pilot study, we recruited N = 28 Shipibo-Konibo speakers who participated either in a behavioural segmentation task or an EEG experiment. In the behavioural task (N = 15), participants segmented four videos depicting everyday activities by pressing a button to indicate either fine-grained or coarse-grained event boundaries. The videos contained experimentally manipulated event changes, including (a.i) changes in actors, (a.ii) temporal discontinuities, and (a.iii) changes in manipulated objects. During the EEG session (N = 13), participants completed two passive tasks: (1) viewing the same videos used in the behavioural task, and (2) listening to short procedural narratives containing different types of clause-linking constructions, including SR markers indicating (b.i) subject change, (b.ii) temporal change, and (b.iii) shifts from intransitive to transitive action, as well as (b.iv) discourse connectives without overt SR marking. Building on previous studies, we infer coarse versus fine boundary segmentation indirectly through event-related potentials (ERPs) associated with semantic and discourse updating, including N400 and P600 components. EEG was recorded from 32 scalp electrodes. Epochs were time-locked to the onset of the SR or discourse marker, and mean ERP amplitudes were analysed in the 300–500 ms and 500–900 ms time windows following marker onset. Preliminary analyses suggest that Shipibo-Konibo speakers track event changes across similar dimensions in both visual and auditory modalities. Behavioural segmentation results indicate that actor changes are stronger predictors of coarse event boundaries than temporal changes, whereas object changes are more strongly associated with fine-grained boundaries. Initial visual inspection of ERP grand averages suggests increased late positivity (P600) for subject changes relative to temporal changes, consistent with a coarse-boundary response associated with discourse updating and event model revision. Ongoing analyses will further examine ERP responses to changes in action transitivity and discourse connectives, and compare these responses to neural signatures found in visual event segmentation. Together, these findings may reveal how cross-linguistic diversity, similarly to culture, might shape how humans allocate attention to specific dimensions of event change across modalities.
Topic Areas: Speech Perception, Morphology