Speaker
Description
Cognition critically relies on both working memory (WM) and temporal information. However, how our brain processes temporal information in WM remains largely unresolved. Previous work using a novel n-item delayed duration reproduction task found that durations can be stored as discrete items in WM (Herbst et al., 2025). Herein, participants were presented with sequences of temporal intervals (items), marked by brief tones. After a retention period, participants reproduced these items. Using this protocol, I will discuss two studies. In the first magnetoencephalography (MEG) study, we investigated the encoding and retention of time intervals in WM, in which we contrasted brain activity as a function of the number of durations in the sequence and the total duration of the sequence. We observed that during the encoding of the temporal sequence, the amplitude of the evoked response to the last tone increased with a smaller number of items and longer sequence duration. Importantly, this amplitude positively correlated with the participants' reproduction of single-item sequences. During the retention interval, we examined how WM load modulated oscillatory power. Our results showed that alpha (8-12 Hz) increased with the number of items, in agreement with unpublished work (Herbst et al., submitted). In the second behavioral study, we asked whether and how these durations in the sequence interfere with each other, investigating the effects of prospective and retrospective cueing on recall of durations. So we added a cue indicating whether one interval in the sequence (first or second) or the full sequence of intervals had to be reproduced. The cue could be presented before (pro-cueing) or after (retro-cueing) the retention period. We found that a primacy effect on reproduction precision only occurs when retro-cueing for a long duration: the reproduced long duration was more precise in the first position.
Overall, remembering a sequence of durations is to some extent similar to other WM content. Our MEG results suggest that evoked responses and oscillatory power are neural signatures of temporal information encoding and maintenance, respectively. Second, we also found a primacy effect of precision. We expect these two studies to help further clarify how the ‘when’ is processed in the human brain.
