Abstract
It takes time for the retinal signal of a flashed target to reach the brain. If during pursuit the brain matched the signal to the site of fixation at a given point in time, the perceived location of the target would be biased towards the direction of pursuit. We examined whether the “mismatch hypothesis” could explain errors in exocentric localization. Given that neural processing time depends on the stimulus, it is possible that localization of a target relative to a static reference (exocentric localization) during pursuit would produce errors, but these should not be affected by spatiotemporal properties of the static reference. To test this idea, a target rectangle surrounding a static reference was briefly flashed while subjects were required to pursue a dot moving rightward. The task was to localize the target relative to the reference. We manipulated the adjacency of the target to the reference, as well as the lifetime of the reference (infinite, appearing, and disappearing when the target was flashed). The amount of mislocalization dramatically decreased as the target became closer to the reference in both appearing and infinite conditions, but localization remained accurate in the disappearing condition. The last result may be explained by supposing that both target and the disappearing reference were mislocalized by the same amount in the direction of pursuit. Interestingly, as the interval between the flashed target and the disappearance of the reference increased (up to about 150 msec), the amount of mislocalization approached that of the infinite lifetime condition. These results imply that the “mismatch hypothesis”, along with the “latency difference” cannot explain the errors in exocentric localization.