Abstract
Express saccades are movements of extremely short-latency (80–110ms). Despite their reflexive nature, recent work in our lab has shown that express “averaging” saccades elicited by two closely-spaced stimuli can be influenced by an instruction issued prior to stimulus appearance (Edelman et al, 2007). Previous work has also shown that when two stimuli are presented with a large directional difference, the production of express saccades decreases (Chou et al, 1999). Here we examine whether a high-level instruction (visuomotor set) can rapidly resolve the competition of two far-apart stimuli, allowing an express saccade to proceed. Subjects performed a 150–200 ms gap task in which two bright squares (1° x 1°) appeared, one 8° eccentric and positioned at a random direction 0, or +/− 60° away from the right horizontal meridian, and the other at equivalent positions to the left. Subjects were instructed to make a saccade to the stimulus (“target”) in the hemifield indicated by a central fixation arrow and to ignore the contralateral “distractor.” Subjects also performed a task in which no distractor appeared. EMs were recorded (Eyelink II) in two subjects at 500 Hz. Both subjects were “express saccade makers.” Surprisingly, the presence of the distractor increased reaction times by only 11 ms. Saccades virtually always landed on the desired target. We also found that: 1) narrowing the focus of the visuomotor set by positioning stimuli closer to the horizontal meridian reduced the distractor effect, 2) express saccades could occur with high frequency even without a visuomotor set, and 3) the distractor effect was large only when two stimuli were presented unexpectedly in the uncued hemifield. This evidence suggests that high-level processes are sufficient, but not always necessary, to suppress distractor-related transient visual activity to an extent sufficient for express saccade generation.
NIGMS GM00816-27 (SCORE) NIH/NCRR 5G12 Rr03060 (RCMI).