Early saccade planning cannot override oculomotor interference elicited by gaze and arrow distractors

Humans tend to perform reflexive saccades according to the eye-gaze direction of other individuals. Here, in two experiments, we tested whether preparing a saccade before the onset of a task-irrelevant averted-gaze stimulus can abolish this form of gaze-following behavior. At the beginning of each trial, participants received the instruction to prepare for a leftward or a rightward saccade. This was provided either on a trial-by-trial basis (Experiment 1) or was maintained constant within a whole block of trials (Experiment 2). Then, a central fixation spot changed in color, acting as a "go" signal to perform the saccade. Simultaneously with the go-signal onset, a task-irrelevant distractor face looked either leftwards or rightwards. In so doing, no temporal overlapping was likely to occur between saccade preparation and the presentation of the distractor. Arrows were also employed as non-social control stimuli. In both experiments - and regardless of the distractor type - saccadic latencies were smaller when both the instruction and the distractor conveyed the same spatial vector, rather than the opposite. Taken together, these results suggest that an early preparation of saccades is not sufficient to override the oculomotor interference effects elicited by both social and non-social distractors.