The time to execute a saccadic eye movement typically depends upon several factors including the complexity of the task. To investigate how on-line changes in instruction influence saccade initiation, we devised a paradigm that combined pro- and anti-saccades and the instruction of which saccade to generate was sometimes changed in the middle of the trial. Such a change of instruction requires subjects to alter their preparatory "set".

Nine adult subjects (ages 22 - 46) were recruited to participate. Each trial started with the appearance of a central fixation point (FP) and subjects were instructed to fixate upon it. After a variable delay, an eccentric stimulus appeared and subjects generated a saccade either toward the stimulus (pro-saccade) or away from the stimulus (anti-saccade). The FP remained illuminated for the duration of the trial and its colour instructed the subject which type of saccade to make. A red FP indicated a pro-saccade trial and a green FP indicated an anti-saccade trial. On a small percentage of trials (25%) the instruction changed in the middle of the trial when the colour of the FP changed colour (from red-to-green or green-to-red) prior to appearance of the eccentric stimulus. The time of the shift in FP colour (i.e., shift time) occurred randomly 0 - 400 ms (in 100 ms increments) before target appearance. On the remaining 75% of trials the FP colour and therefore the instruction did not change.

On control trials (i.e., FP colour did not change), saccadic reaction times were typically longer for anti-saccade trials than pro-saccade trials and subject generated more direction errors on anti-saccade trials (i.e., saccades toward the stimulus instead of away from it). However, on trials in which the instruction changed immediately prior to the appearance of the eccentric stimulus, both SRT and the frequency of direction errors increased dramatically. These increases in SRT and frequency of direction errors occurred for both red-to-green and green-to-red changes in the FP and were greatest for trials with the shortest shift time (0 ms), least for largest shift time (400 ms), and changed gradually for intermediate shift times (100 - 300 ms).

These results demonstrate that this paradigm may be very useful to investigate the neurophysiological mechanisms underlying preparatory set. In addition, such a paradigm may be a useful tool in clinical investigations of frontal lobe function and saccade initiation.

Supported by the EJLB Foundation and the Medical Research Council of Canada.