Abstract
Spatial attention paradigms typically require observers to monitor an empty location in space for the appearance of a target. Results generally support some variant of a spotlight model in which the target, as well as irrelevant nearby shapes, receive enhanced processing relative to ignored locations. In contrast, object-based attention experiments show that observers are quite good at selecting one of two overlapping shapes, a situation that would pose difficulties for a diffuse spotlight mechanism. Are these two kinds of attention carried out by separate mechanisms or by a single system operating in different modes? For example, one version of a single mechanism account assumes that object-based attention corresponds to allocation of spatial attention to the surface or contours of the object. We tested this account using event-related brain potentials (ERPs). Previous research has shown that attention to location results in an enhancement of the P1 and N1 components of the visual ERP. This pattern is not observed when subjects attend to other attributes such as color or size, suggesting that P1/N1 enhancement might serve as a “signature” of the operation of the spatial attention system. We compared ERPs for attended and ignored overlapping objects and found differences at latencies as short as 80 ms but comparable P1/N1 components suggesting that object and space-based attention are mediated by different systems.