Abstract
Just before a saccade certain cells (found for instance in LIP, SC and FEF in monkeys) are activated by stimuli presented at the location that will fall in their receptive field after the saccade is made. It has been proposed that this remapping contributes to a prediction of the postsaccadic visual detail that is compared with the actual one. Based on the difference between the prediction and the input, the visual system realigns the relation between retinal and world coordinates (Deubel & Bridgeman). We examined the visibility of this predicted transsaccadic visual image by removing the saccadic target before the saccade and asking participants to report if they had looked directly at the target. The participants (n = 19) first fixated a cross presented at the center of the screen. Aater an unpredictable interval, a flash presented in the periphery, indicated where they should move their eyes. At various times after the saccadic cue, a vertical bar was presented at the target location for 30ms. The participants’ task was to report if they looked directly at the bar. We found that even when the bar disappeared as long as 170 ms before the saccade – and was therefore never presented in the fovea – participants reported that they did look at it directly on average 85% of the time (±8.3%). Those results suggest that when observers saccade to a target that disappears before the end of the saccade, they see the target as if it was still there when their eyes land.