Visual perception at the time of successive saccades

Saccadic eye movements profoundly influence the perception of space: stimuli presented briefly around saccadic onset are perceived compressed towards the saccadic target. We studied perisaccadic mislocalization with a double-step saccade paradigm, an important technique in eye-movement research where the second saccade needs to be planned before the first has been executed, and must therefore take into account the displacement caused by the first. In our study the saccades were "memory-guided", with both saccadic targets extinguished before commencing the sequence. We measured perisaccadic localization of a small probe dot briefly flashed at various times during the sequence. At onset of the first saccade, probe dots were mislocalized towards the first and also the second saccade target. However, on onset of the second saccade, there was very little mislocalization. We reasoned that the lack of mislocalization could reflect failure to encode the location of the second saccadic target in an appropriate coordinate space (that takes into account the motion of the first saccade). Perhaps this encoding takes time? To test this idea, we increased the viewing duration of the saccade targets (before commencing the saccade sequence), and observed mislocalization at onset of both the first and second saccade. Our data suggest that construction of the spatiotopic representation requires at least 200 ms, a notion reinforced by a series of experiments on saccadic displacement of motion. We conclude that perisaccadic mislocalization towards the saccade target occurs only after neural map has been constructed in suitable coordinates, and the construction of this map requires time. Meeting abstract presented at VSS 2012