Abstract
A number of studies have shown that observers can employ top-down control over attentional capture by abrupt onsets. For example, abrupt onsets that do not match an observer's attentional set (Folk et al, 1992) or those that occur outside the focus of attention (Theewes, 1991), do not capture attention. Likewise, when observers perform a search task high in perceptual load, the ability of abrupt onsets to capture attention is attenuated (Cosman & Vecera, 2008). However, in the real world our attention is regularly captured by items that are completely irrelevant to the task we are performing at a given time. Studies showing evidence of top-down control of capture by abrupt onsets have employed blocked designs in which reported reaction times are averaged across multiple blocks of trials. As a result, it is possible that onset capture in early trials is obscured by aggregating RTs across blocks of trials. The current set of experiments attempted to examine whether or not infrequent or strategically presented abrupt onsets that are completely irrelevant to task performance capture attention in the face of top-down manipulations known to attenuate such capture. Participants performed a contingent-capture task, a focused attention task, and a perceptual load task in which the abrupt onset was never relevant to performance of the task. Across all experiments, it was shown that during early trials abrupt onsets captured attention regardless of top-down attentional control settings. This effect eventually decreased and disappeared in later trials, indicating that top down control took time to “tune up” to the displays used in a given task before affecting attention capture. These results suggest that the configuration of top-down control takes time and that abrupt onsets can initially capture attention despite top-down control settings.