摘要
In visual search for a target, attention must shift to the target location. This is typically followed by target perception before a decision leading to a motor response. These processes, attentional shifting, perception, and decision-making, often overlap in time. To disentangle them, we adapted a previous search task (Zhaoping & Guyader, 2007) to include four stimulus conditions: Asimple, Bsimple, A, and B. While fixating at the center of the display, observers covertly searched for a uniquely oriented bar and reported whether it appeared in the left or right half of the search array. In each trial, the target could appear near or far from the fixation, creating variations in target recognizability. The target bar was more salient in Asimple and Bsimple conditions, attracting attention more strongly. This was reflected in an earlier (by about 50 ms) and stronger N2pc component in the ERP, and followed by a stronger central-to-lateral potential gradient, with more positive central potentials. In contrast, the less salient targets in conditions A and B typically compell observers to recognize the target before making their final decision. Due to crowding, recognizing the target was more difficult when it appeared farther from fixation. Although equally salient in conditions A and B, the target bar in condition A, but not in condition B, was part of an object that was a rotated version of other objects in the search array, leading to confusion in decision-making once this object's shape was perceived. This confusion was reflected in ERP differences between conditions A and B, appearing earlier over posterior electrodes (~300 ms after stimulus onset) than temporal electrodes (~400 ms), and this ERP difference between A and B was weaker or absent when targets were less legible due to eccentricity. There is a dichotomy between central and peripheral vision behaviorally and ERP-wise.