Abstract
Imagine yourself in a car race waiting for the traffic light to go green. Impulsivity could push you to accelerate when the light is still red. In contrast, temporally guided anticipation could lead you to accelerate at the time the light goes green. Whether these two types of early responses rely on the same or different neural processes is an open question. This question was investigated using an oculomotor task where the delay between a warning and an imperative visual stimuli was predictable. The spatial uncertainty of the "go" signal was also varied. On average, 10% of experimental trials were associated with a response before the "go" signal ("early saccade"). After the offset of the warning stimulus, the latency distribution of early saccades was bimodal, with a first mode peaking after 200 ms (1st mode saccades) and a second one starting to build-up after 375 ms (2nd mode saccades). With increasing delay duration: the number of 1st mode responses decreased whereas the number of 2nd mode responses remained approximately constant; the latency and variance of 2nd mode saccades increased; the maximum velocity of 2nd mode responses decreased. In general, the amplitude of 2nd mode responses was larger. These results show that there are probably two independent processes taking place before an expected event: an unintentional release of inhibition evoking an impulsive 1st mode saccade and an anticipatory process leading to a 2nd mode saccade.