Abstract
WHY: Oculomotor tracking has both reflexive and voluntary aspects in most cases. Previous studies of how effort and attention modify tracking responses have employed a single target with varying instruction. Here we report experiments with two, independently moving targets with an instruction to track one or the other. A reverse correlation technique was used to measure the tracking responses to both targets so that the effect of attention and effort could be compared more directly. HOW: Each eye saw two concentric rings, viewed haploscopically through the optics of an SRI dual-Purkinje binocular eye tracker. Integrated velocity noise (low pass to 10 Hz) was added to the targets to generate random 4-D walks. Eye motion was sampled binocularly at 120 Hz in synchrony with the monitor frame rate and used to calculate both version and vergence velocity. These velocities were correlated to the ring velocities to yield “eye movement correlograms”. WHAT: Both targets contributed significantly to vergence and version, despite the effort to track just one. Both attended and unattended targets produced short latency responses (onset at 70–80 ms for H and V version; 90–110 ms for H and V vergence). The attended target showed additional, longer latency responses (onset at 100–110 ms for H and V version; 150–200 ms for H vergence). This effect was most evident with H vergence and was generally absent with V vergence. SO WHAT?: Because different visuo-motor subsystems require more or less processing and involve more or fewer synapses, there are latency differences in their responses to target motion. The eye movement correlogram may reveal these various modules as distinct components and allow a dry dissection of the visual processes involved in oculomotor control. The current results show that voluntary, attentive tracking is associated with a relatively long latency response component in addition to a shorter latency, reflex component.