Lizzie Gao¹, Anna C Nobre^1,2; ¹Yale University, ²University of Oxford

The natural flux of the dynamic external environment contains embedded temporal regularities across multiple time scales. Previous studies have examined how local temporal regularities guide perception and behavior, showing that items appearing at predictable moments are processed better than unanticipated items. In parallel, studies have examined how segmenting ongoing stimulation into discrete events modulates psychological functions and temporal judgments of stimuli. In this study, we explored how temporal regularity at two timescales – intervals between stimuli and durations of events – impacted stimulus processing, aiming to uncover the mechanisms of nested hierarchical temporal structures and their putative interactions. 20 healthy participants viewed rapidly presented faces (100 ms) from the Chicago Face Database. Their task was to respond as quickly and accurately as possible every time a pre-designated target face appeared. Targets appeared sparsely, once per eight faces on average, among a stream of another repeated non-target face. The target face remained the same throughout the experiment, but the non-target face changed on occasion, demarcating separate temporal contexts. Temporal expectations were manipulated both locally and globally. Locally, the inter-stimulus interval between faces could be fixed (consistently short (500 ms) or long (1400 ms)) or variable (intermixed short and long). Globally, context durations were fixed (6 seconds) or variable (4–24 seconds). Reaction times (RT) and accuracy were measured. Consistent with previous findings, local temporal expectations significantly improved RTs for faces at predictable short intervals. Response times to targets were slower soon after the non-target face changed and decreased over time, showing an effective induction of contextual boundaries. Behavioral measures were not sensitive to the global temporal predictability of contexts. RTs to targets were similar before and after predictable or variable boundary changes. Ongoing neural studies will explore the levels of stimulus processing impacted by local and global temporal structures with greater granularity.