| The human visual system can notice differences between memories of previous visual inputs and perceptions of new visual inputs, but the comparison process underlying this ability has not been well characterized. A series of behavioral, event-related potential (ERP) and eye-tracking experiments tested if this comparison process is limited or unlimited in capacity. Subjects performed a task in which they looked for a changed item among non-changing items (any-difference task) and a task in which they looked for a single non-changing item among changing items (any-sameness task). Responses became slower and less accurate as the number of items increased in both tasks, but this effect was much greater in the any-sameness task. In addition, increasing the number of items in the any-difference task had little or no impact on the latency of eye movements toward the changed item or on the latency of the N2pc component of the ERP waveform, an index of attention. Further experiments showed that the comparison process is prone to interference from pattern masks and can operate on a selected subset of features by voluntary control. The results together suggest that comparisons between visual working memory and perceptual inputs can be accomplished by a combination of unlimited and limited-capacity processes. |
