Roy Ramati¹, Amit Yashar¹, Yaffa Yeshurun¹; ¹University of Haifa

Previous studies suggest that temporal individuation and temporal averaging are mediated by distinct processes and that individuation over long timescales is shaped by temporal interference—temporal crowding—which differs qualitatively from visual masking. To further elucidate the underlying mechanisms, we examined how prior knowledge modulates temporal individuation and averaging with central presentation. Participants (n=36) viewed three-items sequences with varying SOAs (200–400ms). Two conditions required individuation—reproducing the orientation of a single item. In one condition, the target position was known in advance (always the second item). In the other, the target position (first, second, or third) was cued only after the sequence. Thus, in the former condition only a single item was ever relevant, whereas in the latter condition, all items had to be maintained in working memory until the cue appeared. In a third condition, all items were task-relevant because participants reported the items’ average orientation. Mixture-modeling analysis of individuation errors, revealed the hallmark of temporal crowding: regardless of prior knowledge, SOA affected the precision of target encoding. A robust temporal-position effect emerged—precision increased for items appearing later in the sequence. Critically, prior knowledge eliminated this effect: when the target position was known, the second item’s precision matched the third item’s precision in the no-prior-knowledge condition. Prior knowledge also reduced substitution errors, with the magnitude of the reduction depending on SOA and on whether the distractor preceded or followed the target. In the averaging condition, unlike individuation, SOA had no effect, supporting the dissociation between averaging and individuation. This pattern resembles our findings in peripheral temporal averaging. However, unlike in peripheral averaging, the third item was assigned the highest weight. Together, these results suggest that advance knowledge can dynamically narrow the temporal window of interference, possibly by advance filtering of information based on its temporal position.

Acknowledgements: ISF, grant number 1780/19 to Yaffa Yeshurun