Abstract
When spatial attention is directed by a target towards one side of the visual field, electrodes on the scalp record an asymmetrical electrophysiological potential over parietal cortex called the N2pc. This EEG component has recently been shown to reflect the engagement of attention at a location, but not sustained attentional deployment (Tan & Wyble, In Press). A computational model has interpreted this component as the attentional system “locking-on” to the location of a target (Tan & Wyble, In Press). The Pd is another lateralized component, thought to reflect suppression of distracting information. In contrast to this interpretation, we have run an EEG experiment using four RSVP streams which reveals a Pd elicited by two simultaneously presented targets, when the targets are separated by one degree of visual angle, but not when separated by four degrees. To provide a neural intuition for the mechanisms underlying the Pd component, we have revised our previous neurocomputational model of attentional dynamics so that it simulates both the N2pc and the Pd as suggested by a variety of experimental findings. The new model simulates the Pd as a laterally asymmetric imbalance in the inhibition of attentional control neurons in posterior parietal cortex. This imbalance can have multiple causes, which explains how the Pd is elicited both with and without distractors. Critical to this account is strategic regulation of the attentional focus, such that it is tighter when stimuli are spatially proximal, whether they are targets or distractors. Furthermore, since the Pd can be elicited by distractors without a leading N2pc, the model includes a reactive form of inhibition that can, in some cases, suppress distracting information before it can elicit an attentional response.
Meeting abstract presented at VSS 2015