Abstract
During visual searches, we can search for a target more quickly by enhancing or suppressing features based on present goal through a top-down attentional template. Such suppression is known as a template-for-rejection effect. It is controversial whether negatively cued feature captures attention (the search-and-destroy hypothesis) or not (the active attentional suppression hypothesis). The present study investigated whether template-for-rejection effect relies on active attentional suppression by introducing a letter probe task. We conducted three experiments to examine the effects of spatial arrangement of items (segregated in Experiment 1; intermixed in Experiment 2) and the number of potential target features (one in Experiments 1 and 2; six in Experiment 3) on the feature-based-template-for-rejection effect and active attentional suppression. Participants completed search trials and occasional probe trials (1/3 of total trials). In the search trials, participants were pre-cued by one of three cue conditions (a positive, negative, or neutral cue condition representing a target, distractor, or irrelevant feature, respectively) and identified the orientation of a target line segment. In the probe trials, a search display was discontinued after 30, 100, 250, or 400 ms (between participants) from the onset and was replaced by a probe display. Participants were exempted from target identification but were to report as many letters as possible in the probe display. The results demonstrated the template-for-rejection effect and active attentional suppression. Specifically, the search reaction times following negative cues were faster than those following neutral cues. Importantly, there were no time points that the percentage of reported letters for negatively cued items exceeded that for potential targets, inconsistent with the search-and-destroy hypothesis. However, we found neither a rejection effect nor active suppression when the search array included six colors. These results support the active suppression with a constraint that a single potential target feature is involved.