Viewing preferences of rhesus monkeys related to memory for complex pictures, colours and faces

In order to determine the preferences of rhesus monkeys for visual stimuli, their eye movements were measured in response to presentations of complex pictures, fields of uniform colour, and of faces using the scleral search coil technique. The monkeys (n = 4) controlled both the onset and offset of the stimuli by the direction of their gaze. Each stimulus was presented 4 times, with 0 or 2, and 36 or 38 trials between successive presentations. Several trends were apparent in their scanning behaviour: (1) all 4 monkeys spent more time looking at pictures and faces compared to colour fields. As individuals, they differed in their overall propensity in looking at visual stimuli: monkeys that spent the most (or least) time looking at pictures spent the most (or least) time looking at colour fields. (2) Although the monkeys appeared to prefer pictures and faces to colour fields as measured by gaze duration, preferences for individual pictures, faces and colour fields were not evident. (3) Memory for recently presented stimuli substantially affected gaze duration which was significantly longer for the first compared to the second presentation of the pictures and faces, and memory was estimated to influence gaze duration over as many as 38 intervening trials. These effects were not significant in the case of colour fields. (4) There were no significant differences either in the average latencies to initiate eye movements or the number of saccades on the first and second presentations of pictures, colors or faces for the 4 monkeys. However, the average latencies to the first eye movement within a trial were longer for colour fields than for pictures for all 4 monkeys. Individual monkeys differed substantially in their mean latencies for the initiation of the first eye movement within a trial, which ranged from 235 ms to 414 ms in the two extreme cases. (5) At the presentation of faces, the monkeys tended to make saccades to major facial features, and only occasionally to the perimeter of faces. We conclude, firstly, that patterns of eye movements in monkeys reflect their natural predilection for sampling novel stimuli and that monkeys are motivated to view visual arrays. Secondly, that the protracted viewing of novel stimuli and long latencies to initiate saccades contrasts with short latency neuronal responses recorded in inferior temporal cortex. These behavioural and neurophysiological data suggest that neuronal activity specific to novel stimuli and to faces is not a consequence of oculomotor responses to these stimuli.