Amplitudes of head movements during putative eye-only saccades.

The mechanisms allowing humans and other primates to dissociate head and eye movements during saccades are poorly understood. A more precise knowledge of head movement behavior during apparent eye-only saccades may provide insight into those mechanisms. We studied the distributions of head amplitude in normal humans. In half of the subjects, these distributions indicated the presence of a population of minor ("residual") head movements during eye-only saccades, distinct from the continuum of head movements generated during frank eye-head saccades. Like full-sized head movements, the residual movements grew in proportion to target eccentricity, indicating their drive is derived from the premotor command for the saccade. Furthermore, their amplitudes related most strongly to the head amplitudes obtained when subjects produced full-sized head movements and were reduced when subjects were instructed to perform exclusively eye-only saccades. Both observations suggest that the drive for residual head movements originates downstream of the point in which the head movement command diverges from the generalized gaze shift command. The results are consistent with a model of head control in which a neural gate prevents the common gaze shift command from reaching the head premotor circuitry whenever an eye-only saccade is desired. However, the gate is either imperfect or the multiple pathways that relay gaze shift signals to the head motor circuitry allow for the gate to be circumvented. The results underscore the need for physiological studies to probe neuronal activity related to neck activation during eye-only saccades.