Abstract
Prior evidence suggests sequential saccades are processed in parallel: secondary saccades occur with shorter latencies, and attention is allocated to secondary spatial locations prior to primary movement execution. However, it remains unclear whether such saccades are planned independently or as part of a sequence. We measured saccadic curvature in sequential saccades to address this question. In this study, curvature of vertical saccades was measured prior to or following horizontal movements, and trials with single vertical saccades served as controls. In Experiment 1, the offset of a fixation point and onset of three horizontally-aligned targets signaled subjects to initiate sequences of two saccades. The first was vertically directed towards the middle target; the second was cued to be either rightward or leftward by a line on the fixation point. Analyses revealed that primary saccades tended to curve away from the direction of the subsequent saccade. In Experiment 2, subjects made a vertical saccade to a target, followed by a second saccade to either a near or far rightward target. We found that the curvature of the initial saccade away from the subsequent saccade was more pronounced when the second saccade had a larger amplitude. In Experiment 3, the fixation point was presented on either the right- or left-hand side, and two vertically-aligned targets were shown on the screen following fixation point offset. Thus, sequences consisted of primary horizontal and secondary vertical movements. We found that the secondary saccades curved in the direction of the antecedent saccades. In all three experiments, there were significant correlations between saccadic curvature of the right and left eyes; however, no significant correlations between curvature and either inter-saccadic interval or latency were found. These findings indicate that saccadic curvature is modified by characteristics of both antecedent and subsequent saccades, suggesting that sequential saccades are not planned independently.
Meeting abstract presented at VSS 2016