Abstract
The abrupt appearance of new object in a display, an abrupt onset, has long been shown to capture attention in visual search tasks. Onsets are believed to capture attention involuntarily, regardless of the goals or intentions of the observer. In a series of experiments, we repeatedly fail to replicate this phenomenon with only minor changes to the standard paradigm. We then show that capture only occurs when prior to the start of the search, a spatial cue orients attention to the center of the display. Specifically, onsets do not capture attention in a visual search task unless the fixation circle in the center of the display appears at the start of each trial. If the fixation circle is always present (i.e., during the inter-trial interval), onsets may be prioritized in the search task but response times on trials where the onset appears as the target do not produce a shallow slope for response latency as a function of increasing set size, the standard measure used to infer capture. Further studies show that this effect is not limited to the appearance of the fixation circle; color changes and auditory beeps also provide the necessary impetus for capture. Capture is also eliminated when the fixation circle does not cue attention to the center of the search array. These results reveal a surprising limitation to the robustness of attention capture by onsets. Attention must be cued to the center of the display prior to the appearance of the search array for capture to occur, potentially weakening the strong claim that onsets capture attention involuntarily.