A retrieved context account of episodic recall and event segmentation

Poster Session D - Monday, April 15, 2024, 8:00 – 10:00 am EDT, Sheraton Hall ABC

Lynn Lohnas¹ (ljlohnas@syr.edu); ¹Syracuse University

There is a rich history from cognitive psychology and cognitive neuroscience that humans segment their continuous experience into meaningful events or episodes (Radvanksy & Zacks, 2014; Clewett et al., 2019). Recently, computational cognitive modeling has sought to bridge findings from event segmentation with episodic memory and temporal perception (e.g., Franklin et al., 2020; Horner et al., 2016; Pu et al., 2022). However, work remains to bridge findings across different episodic memory paradigms. The present research generalizes the retrieved context model framework (Howard & Kahana, 2002; Lohnas et al., 2015; Polyn et al., 2009) to behavioral and neural findings from event segmentation, free recall, temporal perception and serial recall. I demonstrate how this model can account for the influence of event segmentation on episodic recall using consistent principles and minimal changes across paradigms. In brief, each studied or retrieved item within an event updates a slowly changing temporal context representation, and boundaries between events lead to a larger shift in temporal context. Temporal context also serves as the retrieval cue, underscoring the importance of temporal information to representations of episodic events. These results provide an important bridge between serial recall and free recall. Both tasks require participants to study lists of items, but the instruction to recall items in any order (free recall) or in their studied order (serial recall) has led to divergent theories between paradigms. Further implications for theories of episodic memory and event segmentation will be discussed.

Topic Area: LONG-TERM MEMORY: Episodic