Abstract
A salient distractor’s ability to capture attention is thought to depend on goals: in search for a known target shape, the presence of a salient-color distractor slows performance when the target is reliably the unique shape (and participants use singleton-detection mode) but not when it appears among heterogeneous shapes (and participants use feature-search mode). However, such control breaks down when the target and distractor colors randomly repeat or swap from trial to trial. In addition, distractor interference is sharply reduced when the colors repeat vs. swap. This color-priming effect is taken to show that selection history guides attention and can override goal-directed control: a color’s priority increases if it characterized the target on the previous trial, and decreases if it characterized the distractor. Here, we test an alternative account, according to which color priming reflects how difficult it is to disengage attention from a distractor sharing the previous target’s color rather than increased priority of that color. To dissociate priority from disengagement effects, target and distractor colors were selected from four possible colors. Thus, the target color could repeat, be new or swap, i.e., take on the previous distractor’s color. Likewise, the distractor could repeat, be new or swap, i.e., take on the previous target’s color. Target-color activation (reflected by target repeat vs. new and distractor new vs. swap) modulated distractor interference, supporting the priority account over the disengagement account. However, distractor-color repetition (reflected by distractor repeat vs. new) did not modulate distractor interference, suggesting that it does not underlie resistance to capture by constant distractors.