In complex tasks subjects occasionally look-ahead to objects of future manipulation (Pelz & Canosa, Vision Research, 2001). In a model-building task, we have previously found that the frequency of these fixations can be manipulated by task structure, suggesting that they reflect subjects' cognitive plans. They also resulted in earlier saccades to the previously fixated item when subjects subsequently reach to pick it up. We postulated that look-ahead fixations might play a role in planning the reach. We examined the eye, head and hand movements of subjects in the same task. We found, however, that look-ahead fixations did not influence the velocity, amplitude or duration of head or arm movements when subsequently reaching to an object. Thus these anticipatory eye movements did not overtly influence the object-related action. Secondly, we found that when reaching to an object following a look-ahead, the first fixation directed towards the target was significantly closer to the target (mean 11cm) than those that accompanied reaches without a preceding look-ahead. This increase in accuracy may be linked to the greater eye-hand latency and suggests that look-ahead fixations can influence subsequent eye movements without directly influencing the associated reach. Additionally, head movements that accompany look-aheads are significantly smaller in amplitude (by 10 degrees) than those that accompany reaches to the same locations, supporting other evidence that the head and hand may be closely linked.

Supported by NIH grants EY05729 and RR09283