The interaction of conflicting retinal motion stimuli in oculomotor control.

Oculomotor response has been assessed in humans during the presentation of conflicting retinal motion stimuli. In the majority of experiments a background stimulus was made to move with a constant velocity ramp in one direction followed by rapid resets at regular intervals. In the absence of an adequate fixation target this ramp-reset stimulus induced a nystagmus with a slow-phase velocity and saccadic frequency which remained almost constant as reset frequency was increased from 2 to 5 Hz. Moreover, the induced eye velocity could be considerably increased if the subject attempted 'active' matching of display velocity. During both 'active' and 'passive' responses eye velocity gain reached a peak when display velocity was between 2 degrees/s and 5 degrees/s. The presence of small stationary targets induced a suppression of the passive ramp-reset response which was modified by target eccentricity and by tachistoscopic target illumination. When subjects pursued a sinusoidally oscillating target against a stationary structured background, eye velocity gain was significantly less than for pursuit against a blank background. The degree of interaction between conflicting stimuli was found to be dependent on their relative size, peripheral location and velocity. However, it appears that the human observer is able selectively to enhance feedback gain from one particular source in order to dominate stimuli from other unwanted sources.