Saccadic “adaptation” at late target reappearance

When the oculomotor system faces a discrepancy between the saccade endpoint and the visual target, it can attribute this discrepancy to a change in the environment or to a visual saccadic error. In the laboratory, a progressive adjustment in saccade amplitude is observed in the presence of a systematic displacement of the target during the saccade. According to literature, saccadic adaptation is disrupted when the target is shortly blanked after the saccade, and disappears around a 600ms-blank. Because an increased blank leads to the perception of target displacements, saccade amplitude modifications are thought to be negligible. We however believe that in the presence of a systematically varying environment, we are still likely to adjust our saccadic amplitudes, but that the nature of this adjustment should differ. We examined, in a double-step adaptation paradigm, the effect of various temporal delays in the presentation of the post-saccadic target on the amount of saccadic adaptation. We hypothesize that an automatic and slow modification of saccadic amplitude (implicit learning) should take place for intrassacadic target displacements, whereas a more explicit learning (voluntary and fast) should dominate when target reappearance is delayed. Preliminary results show that saccades’ amplitude is modified up to a 1200ms delay (gain change: M=-14%, SD=10%), and still visible in large proportion up to 600ms (M=-20%, SD=8.9%). Our visual system seems to “adapt” eye movements even when target displacements can be attributed to a change in the environment and not to saccadic errors anymore. Interestingly, we observed an increase in saccadic endpoint variability (entropy, exploration) and saccades’ latency during adaptation, associated with a smaller amount of saccadic adaptation, when target reappearance was delayed and its displacement perceived. Our results indicate that adaptation profiles likely depend on how participants experienced their saccade landing errors.