SOURCES AND TOPOGRAPHY OF SUPRAMODAL EFFECTS OF SPATIAL ATTENTION IN ERP

This study aimed to finding modality-independent event-related potenti als (ERP) modulations related to spatial attention by minimising sever al possible methodological constrains which may account for the rarene ss of this finding in past spatial attention ERP research. ERP were re corded while subjects performed a shifting spatial attention task of e ither auditory, visual or mixed (both auditory and visual) modality. T ask requirements and stimulation devices were identical for either mod ality. Stimuli could appear (80%, valid trials) or not (invalid trials ) at predicted locations, and subjects had to perform a speeded respon se to every stimulus. Reaction time benefits were found for valid tria ls. Instead of the usually found enhancement of sensory-evoked modalit y-specific components for valid trials, ERP showed modality-independen t modulations as a function of attention. They consisted of a fronto-c entral activity starting 240 ms after stimulus onset and originated in the anterior cingulate gyrus, followed 40 ms later by a bilateral par ietal cortex activation. These processes would be reflecting the activ ity, respectively, of the anterior and posterior attention systems pos tulated in Posner's model (Posner and Petersen 1990), which showed a h igher degree of participation during invalid trials. P300-like process es were also observed for invalid trials. These results demonstrate th at supramodal effects of spatial attention on ERP can be obtained. Fur thermore, they extend Posner's model by indicating the timings of impl ication of the attention subsystems.