Inability of rhesus monkey area V1 to discriminate between self-induced and externally induced retinal image slip.

Abstract

Retinal image slip can result from an eye movement across a stationary object or alternatively from motion of the object while the eyes are stationary. The ability to discriminate between these two kinds of retinal image slip is necessary for the perception of a stable visual world. In order to determine if this ability is already present in monkey visual area V1, we asked if single V1 units are able to differentiate between externally and self-induced retinal image slip. Externally induced retinal image slip was realized in the 'object motion' condition (OMC) by moving a behaviourally irrelevant visual stimulus ('object': a bar or a large random dot pattern) across the receptive field while the monkey fixated a small, stationary target. Conversely, self-induced retinal image slip of comparable size was evoked in the 'ego motion' condition (EMC) by asking the monkey to pursue the target, moving at the speed of the object in the OMC, while the object was kept stationary. We recorded 221 units from visual area V1, 51, (23%) of them directionally selective, and compared their responses to self-induced and externally induced retinal image slip. Many of them seemed to give some preference to externally induced retinal image slip. However, on closer examination it became clear that this seeming preference could be attributed to the fact that oculomotor performance was less precise in the EMC than in the OMC, causing a larger deviation from the optimal retinal image trajectory in the EMC. We show that the impact of eye position errors can be eliminated by the use of a position-invariant stimulus, such as large-field random dot patterns. We then show that the impact of both eye position errors and deviation of eye velocity from target velocity in the EMC can be eliminated by moving the stimulus in a given OMC trial according to an inverted replica of the eye movement trajectory in the preceding EMC trial, guaranteeing identical retinal stimulation in the OMC and the EMC. If identical retinal stimulation was ensured, none of the V1 units tested was able to differentiate between externally and self-induced retinal image slip. We conclude that V1 does not contribute to the perception of a world which is stable despite eye movements.