If an eccentric, stationary target is flashed while a subject is performing an eye movement in the dark, can this subject make a saccade to the location in space where the target briefly appeared? Different predictions result from alternative hypotheses regarding the manner in which saccade goals are determined. Retinal error being defined as the vector from the eye position at the time of the flash to the position of the target, the retinal-error hypothesis predicts that the saccade vector will be equal to the retinal-error vector. This hypothesis assumes that the oculomotor system ignores the eye displacement between target presentation and saccade. If so, the target will be missed. In contrast, the spatial-error hypothesis predicts that the eye displacement is taken into account by the brain to calculate the target's physical location to which, therefore, a correct saccade could be aimed. 2. At issue is the generality of a fundamental principle of ocular targeting. Previous studies have established that, if the movement is saccadic, eye displacement is used by the oculomotor system to calculate the target's physical location. In the case of pursuit, perceptual experiments on humans suggest that eye displacement is taken into account although its velocity is underestimated. However, in a recent study McKenzie and Lisberger reported that saccade trajectories starting during pursuit conform to the retinal error hypothesis. In other words, velocity underestimation is close to 100%. 3. Although McKenzie and Lisberger's results are very clear, they might have depended on particular experimental conditions. The issue was reinvestigated in a situation facilitating the discrimination of stimuli.(ABSTRACT TRUNCATED AT 250 WORDS)
