Abstract
Background: In the global effect, saccades are displaced in the direction of a distractor near to the target, an effect that may reflect weighted averaging of neural activity in a collicular map. However, the dynamics of the temporal relationship between distractor and target necessary to generate the global effect are not known. Objective: Our goal was to determine the impact of temporal dissociations between the distractor and target on the global effect. Methods: In the first experiment, 12 subjects performed saccades to targets at 8° eccentricity on the horizontal meridian, with and without distractors located at either 4° (near) or 12° (far) eccentricity. Both targets and distractors appeared for 10 ms only. Distractors appeared at 0 ms (simultaneous), 20, 30, 40, 50 or 60 ms after target onset. In the second experiment, 12 other subjects performed a similar experiment with a wider range of temporal offsets, of 0, 20, 40, 60, 70, 80, 90 and 100 ms. The global effect was reflected in the difference in saccade amplitude between conditions with near versus far distractors. These amplitude data were analyzed first as a function of target-distractor offset. Next we assessed them as a function of integration time, meaning the time between distractor onset and saccade onset. Results: Both experiments showed that robust global effect could still be obtained with large target-distractor asynchronies, even up to 100 ms. The integration time analysis showed that a global effect could be generated by integration times as low as 90 ms, and maximal for integration times of 100–160 ms. Conclusions: Simultaneous onset of targets and distractors are not essential for the global effect. Distractors can generate a global effect even if they appear only 90 ms before the saccade is made, which is shorter than the 130–140 ms ‘constant reaction time’ found in studies using double-step saccade paradigms.