Active Allocation of Attention and the Motion-Induction Effect

Abstract

Motion induction refers to the illusory motion perceived within an extended stimulus (eg a bar) presented shortly after the abrupt onset of an adjacent localised cue. Motion is perceived in the direction away from the cue. The effect may reflect transient attentional activation triggered reflexively by the cue onset. In this study we examined the time course of the illusion at a cue to which attention was allocated deliberately, rather than in a stimulus-driven manner. While maintaining fixation, observers searched for a target cue defined by a unique conjunction of colour and orientation among irrelevant equiluminant distractors sharing the same features. Observers reported the direction of perceived motion within a horizontal bar which interrupted search at different delays, appearing in the middle between two randomly chosen cues. Because equivalent cues primed the bar from both sides, the direction of motion was balanced unless one of the cues was selected. In a related experiment we measured the effects of attention through cost/benefit analysis of detection times to a low contrast spot probing the vicinity of the cue instead of the bar. The incidence of motion induction was above chance at the target cue where reaction times were also reduced relative to nontarget locations. The two effects developed similarly over time, peaking at delays optimal for target selection. Also, active inhibition at irrelevant cues preceded facilitation at the target cue, suggesting that these two processes jointly determine the direction of perceived motion. The converging results from different paradigms support the attentional origin of the illusion.