NEURONS IN MONKEY PARIETAL AREA LIP ARE TUNED FOR EYE-MOVEMENT PARAMETERS IN 3-DIMENSIONAL SPACE

1. A functional class of neurons in area LIP on the lateral bank of the intraparietal sulcus were shown previously (Gnadt and Andersen 1988) to be related to the metrics of saccadic eye move ments. In this study. we tested LIP neurons at different depths with respect to the plane of fixation. 2. Sixty-one neurons were identified for their increased activity before saccadic eye movements. While holding the location of the target constant at the center of the frontoparallel (saccadic) response field, the neurons were rested systematically during eye movements to target positions proximal (near) to the plane of fixation, at the plane of fixation, and distal (far) to the plane of fixation. By necessity, the movements to these targets required a combination of saccadic and vergence movements. 3. Seventy-two percent of the neurons were found to change their activity as a function of target depth relative to the plane of fixation. The neurons had broad tuning curves for depth. Some cells preferred ''near'' target positions, some preferred ''far'' positions. and others responded best in the frontoparallel plane of fixation. 4. The location of a neuron's response held in the frontoparellel plane remained constant regardless of target depth. However. the magnitude of the neuron's response increased when the target was positioned at the preferred depth and it decreased for targets positioned at nonpreferred depths. This indicated that the neurons always were related to the same frontoparallel coordinates, but responded more vigorously when the target was positioned at its preferred depth. 5. The visual display apparatus allowed independent presentation of two stimulus cues for depth: binocular disparity and accommodative demand whereas other cues were held constant. For many neurons, either cue was sufficient to tune the activity in depth, though most neurons responded best for the geometrically appropriate combination of the two cues. 6. Comparison of the binocular tuning for depth with the individual monocular responses showed that the tuning for depth was not produced by simple linear combination of two monocular response fields. 7. We tested a subset of the neurons in a double-movement task that dissociated the retinal coordinates of the visual stimuli from the eye-movement coordinates of the second movement. These tests confirmed earlier findings that this functional class of neurons are active when the eye-movement coordinates marched the neurons' response field. It was not necessary For a visual stimulus to fall within the neurons' response held for them to become active. 8. This study and previous investigations have shown that these neurons are not uniquely dependent on the stimulus parameters on the retina, but more directly to the motor spatial parameters representing the difference in current eye position and desired eye position. We suggest here that these neurons express a premotor signal for directing gaze that is encoded in three-dimensional space.