Abstract
The location of stimuli briefly flashed around the time of the saccade is misperceived. For example, often participants report a shift of the flashes in saccade direction. The mechanisms underlying these perisaccadic misperceptions remain poorly understood. Possible explanations include an increased uncertainty about the temporal alignment of visual information and eye position information, influences of spatial shifts of attention, and inaccuracies in the spatial representation of saccade targets. Here we manipulated saccade onset times as well as attention and spatial encoding by presenting saccade targets either at a location that was highly predictable or at a location that occurred less frequently. Predictable and unpredictable locations could change either slowly, across blocks of trials (experiment 1), or quickly from trial to trial (experiment 2). We found that saccades were slower for unpredictable target locations and flashes were perceived as more shifted on that side. However, saccade latencies and shifts were only poorly correlated, and the difference in shift vanished when predictable and unpredictable locations changed from trial to trial. Our data are not supportive of transient effects. Instead, they indicate a dominant influence of longer-term mechanisms, perhaps related to learning processes in the oculomotor system.