Abstract
Object representations and spatial attention are deeply intertwined constructs that are often behaviorally confounded. However, recent work has illustrated that these two processes are neurally dissociable sub-components of the focus of attention (Hakim, et al., in press). Here, we use two “online” measures to more finely delineate how task-relevant and task-irrelevant interruption influence these two sub-processes of the focus of attention. We use lateralized alpha (8–12 Hz) power as an index of sustained spatial attention (Thut et al., 2006) and contralateral delay activity as an index of the number of object representations (Vogel & Machizawa, 2004). In Experiment 1 (n=20), participants performed lateral change detection with four midline interrupters (circles or squares) that appeared on 50% of trials. In one block, participants were told to ignore the midline interrupters. In the other block, they had to report the shape of the interrupters, in addition to performing the change detection task. Following task-irrelevant interruption, spatial attention was immediately captured, but recovered. Object representations persisted without spatial attention for a few hundred milliseconds, but were lost by the end of the trial. When interrupters were relevant, spatial attention was immediately captured and object representations were immediately lost. In Experiment 2 (n=20), participants performed the same task as Experiment 1, but the change detection task was on the midline and the interrupters were lateralized. This allowed us to determine how the interrupters were processed. We found that when interrupters were present, they were actively suppressed as indexed by the Pd component. The results from these two experiments further suggest a dissociation between spatial attention and object representations. Additionally, they provide positive evidence that object representations can temporarily persist without spatial attention, a topic of central importance in working memory models. Futhermore, top-down control is able to influence both spatial attention and object representations.