Figure 1.
The distractor paradigm and measures for quantifying saccade trajectory deviations.
(A) An illustration of the distractor paradigm with sample saccade trajectories deviating towards (red) and away from (blue) the distractor. (B) The “maximum curvature” measure of saccade trajectory deviation [45]. (C) The “directional deviation” measure of saccade trajectory deviation [46], see text for detail.
Figure 2.
The lateral interaction theory and supporting evidence from the literature.
(A) A schematic illustration of why the lateral interactions in the SC can produce saccades deviating away from the distractors (redrawn from [1]), see text for detail. (B) Simulation results of Wang et al. [1]. Positive and negative values on the y-axis denote deviation towards and away from distractors, respectively. Thick and thin lines denote simulations using kernels of small and large excitation zones, respectively. (C) A graphical meta-analysis of studies that reported directional deviations in saccade trajectory. In Theeuwes and Godijn [6], the saccade target and the distractor could appear at previously stimulated or unstimulated locations; only trials in which both the target and distractor appeared at unstimulated locations were considered. In Van der Stigchel and Theeuwes [52], saccade deviation was binned according to saccade latencies; the data reported here is the average over all latency bins. The same procedure was applied to the data from [6], [19], [49], [52]–[54]. The data from White et al. [9] was collected from two monkeys; the target-distractor SOA was manipulated in this study, only the 0-ms SOA condition was included here. (D) A graphical review of studies that reported saccade deviations quantified with trajectory curvature. Note that the curvature measures differed drastically across studies and thus only the direction of trajectory curvature was considered in this analysis, with “+” and “−” signs on the y-axis denoting trajectories curving towards and away from distractors, respectively. The size of the symbols represents the number of experiments contributing to each data point.
Figure 3.
The strength of fixation activity modulates saccade trajectory deviation.
(A) The motor map in the SC, drawn from the equations and parameters given by van Gisbergen, van Opstal and Tax [79]. The color gradient represents the strength of connections originating from the fixation zone (the rostral pole of the SC). T: Target; D: Distractor; F: Fixation. (B) Non-uniform inhibition from the fixation zone can cause the target-evoked activity to drift away from the fixation, see text for details. (C) The inhibition from the fixation zone can also cause the target-evoked activity to drift away from the distractor, see text for details. (D) Saccade deviations obtained from simulations in a neural field model of the SC [1], [35], assuming that the fixation activity was either weak or strong.
Figure 4.
Mean directional deviation (A), trajectory curvature (B), saccade amplitude (C) and SRT (D) for each target-distractor separation in the gap and overlap conditions in Experiment 1.
Error bars denote ±1 SE; for clarity, only one side of the error bars is drawn. In (C) and (D), thick and thin horizontal lines denote distractor-absent trials in the overlap and gap conditions, respectively.
Figure 5.
Mean directional deviation (A), trajectory curvature (B), and SRT (C) for each condition in Experiment 2.
Error bars denote ±1 SE; for clarity, only one side of the error bars is drawn in (C).