Abstract
Previously, we demonstrated that instructions to make a saccade to one end of a horiz bar that appeared suddenly in a random location can affect saccade vector with no sacrifice in RT (Soc for Neuro 2002). Subsequently, we have shown the effect of such an object-centered motor set holds true even for express saccades (SfN 2004). Here we test for similar effects of an instruction to make a saccade to the upper or lower end of a vertical target array, as well as effects of a more complex instruction to make saccades to one of 4 targets arranged in the shape of a square. RTs and spatial effect of instruction were measured in 4 tasks: 1) Horiz - 2 targets were spaced horizontally (6°) and subjects were instructed to make a saccade to the left or right target 2) Vert - 2 targets were spaced vertically (6°) and subjects were instructed to go to the top or bottom target 3) 4 target/ 2 instruction:, 4 targets were arranged as corners of a square (6°×6°) and Ss were instructed to go to the top right, bottom right, top left or bottom left target. 4) like 3), but Ss were instructed to go to either to the top, bottom, left or right pair of targets. All instructions were indicated by an arrow that served as the fixation point prior to target appearance. In all tasks, targets appeared a random location on the screen. A 150ms gap was used to facilitate express saccade generation. The eye movements of two Ss were recorded at 500Hz using a video eyetracker (Eyelink II). As in the case for previous studies using horiz instructions, both vert instructions and simultaneous vert and horiz instructions influenced saccade endpoint with no increase in RT. In addition, comparing the results of conditions 3 and 4, we found only a slight interference of following the horiz instructions with the simultaneous following of vert instructions. These results beg the question: at what level of motor set complexity does its implementation cause an increase in saccade latency?
Supported by SCORE (NIGMS), RCMI (NCRR)