Abstract
Purpose: :
Saccade amplitude is modified by the Muller Lyer (ML) illusion in the same manner and to the same extent as perception, even when pointing is not affected (Bernadis et al, 2005, Exp Brain Res 162:133). This is contrary to the predictions of the "two visual systems" hypothesis. However, the type of saccade, display characteristics and subject instructions, all play a role in the size of effects observed. Here, the effect of the ML illusion on saccade amplitude, in circumstances in which subjects had to suppress reflexive responses, was investigated with two display times (200ms and 1000ms).
Methods: :
Twelve subjects were exposed to ML displays with one vertex at fixation and the other displayed eccentrically to left or right (6° separation between vertices). "In" (><), "out" (< >) and control (xx) configurations, direction and display time (200ms vs 10000ms) were interleaved. Subjects were instructed not to saccade until the display was extinguished. Eye movements were recorded with infrared oculography and primary saccade amplitude and latency measured offline. Only saccades with a latency from target offset of >100ms and an amplitude >3° were included in the analysis.
Results: :
Two subjects' data were not analysed for technical reasons. For both display times, the remaining 10 subjects suppressed reflexive saccades, and responded to the target offset, at the cost of increased latency (time from offset, 200ms: latency = 381±45ms; 1000ms: 305±40ms; t=4, p<0.0001; Fig A). For both display times, saccade amplitudes increased with the "in" and decreased with the "out" configuration. Effect size calculations (In-Out/Control, expressed as %) showed a trend towards larger effects with the 1000ms display time (200ms: 18±5% vs 1000ms:23±9%; t=1.6, p=0.06; Fig B).
Conclusions: :
Saccade amplitude was modified by the ML illusion, in the same manner and to a similar extent, as perception in these experiments. The effect of an extended preview period was to decrease saccade latency, and increase illusion effect size. However, the effect sizes were similar to what we have reported previously for both reflexive and memory guided saccades (Knox & Bruno, 2007, Exp Brain Res 181:277). These results are not compatible with any strong version of the two visual systems hypothesis.
Keywords: eye movements • eye movements: saccades and pursuits • vision and action