Abstract
Recent studies have examined the temporal properties of distractor events leading to attentional capture. These studies seem to indicate that attentional capture is absent when the distractor event is either not temporally unique (Von Mühlenen et al., 2005) or when its occurrence is expected (Lamy, 2005).
Von Mühlenen used a critical range of intervals of −150 to 150 ms between target onset and distractor event (SOAs), which resulted in absence of capture at 0 ms SOA. However, this range of SOAs makes it impossible to distinguish between both theories. At a SOA of 0 ms the distractor event is not temporally unique and occurs at the expected moment.
In the current experiment, Von Mühlenen's paradigm was adapted to make this distinction. A range of −150 to 50 ms SOA (steps of 50 ms) was tested on attentional capture in 9 observers. Temporal uniqueness would predict absence of capture at 0 ms SOA, while an expected occurrence account would predict absence of capture at −50 ms SOA.
When a task-irrelevant colour change occurs in the target, at or after target onset (0–50 ms SOA), reaction time to the target is shorter in comparison to a colour change in non-targets at the same SOAs. This rules out temporal uniqueness. However, when the colour change precedes target onset (negative SOAs; including the expected SOA of −50 ms) no reaction time differences are found; this implies that the theory of expected occurrence is unlikely as well.
Attentional capture by task-irrelevant events appears to depend on the temporal characteristics of task-relevant events (target onset); before task-relevant events occur, task-irrelevant events can be ignored. However, once the task-relevant event has occurred, distractor events influence our percept. This suggests that attentional capture is not purely bottom-up, but is mediated by top-down processes as well.