Ilanit Hochmitz¹ (), Yaffa Yeshurun; ¹University of Haifa

Target identification is impaired when it is preceded and/or succeeded by task-irrelevant items. When such temporal interference occurs at relatively long intervals, it is termed temporal crowding. Temporal crowding has been observed at stimulus-onset-asynchronies (SOAs) that clearly exceed the limits of visual masking (SOA>150ms), emerging even when items are separated by almost half a second. Here, we aimed to advance our understanding of the temporal dynamics that give rise to temporal crowding by dissociating the roles of backward and forward interference. Across four experiments, we used an orientation-estimation task, in which observers viewed sequences of two or three oriented items separated by relatively long intervals (five SOAs: 175–475ms). In three-item sequences, the target appeared either first, second, or third, depending on the experiment. These different target positions introduced different types of interference: primarily backward, both backward and forward, or primarily forward, respectively. In two-item sequences, the target always appeared first, followed by a single distractor, providing a minimal case of backward interference. Mixture-model analyses revealed that temporal crowding was strongest for the second item, with significant SOA effects on target-encoding precision and substitution errors, but not on guessing rate. A similar, though weaker, pattern of effects was found when the target was first, whereas minimal interference was observed when the target was third. In two-item sequences, even a single succeeding distractor was sufficient to elicit interference: precision increased with SOA, and substitution errors decreased. Thus, temporal crowding arises from a combination of strong backward and weaker forward interference, and this long-lasting impairment can emerge even with a single distractor. Surprisingly, we found stronger forward interference in substitution errors, manifested as higher rates of reporting the preceding than succeeding distractor. Moreover, source confusion did not reflect a complete loss of temporal-order information but rather a coarser representation of items’ order.