Inhibition of Return without Visual Awareness

Abstract

When a cue and a target are successively presented at the same location, reaction times to discriminate the location of the target are longer than when they are at different locations (inhibition of return: IOR). We found that visual awareness of the cue was not necessary for IOR to occur. Both eyes dichoptically viewed 9 × 9 scattered arrays of vertical or horizontal line segments. To avoid effects of eye dominance and binocular rivalry, cue displays were presented briefly (33, 50, or 200 ms). Three types of cue displays were randomised: (i) no cue: horizontal segments for the left (right) eye and vertical segments for the right (left) eye; subjects perceived scattered binocularly-combined crosses, (ii) binocular (fusible) cue: displays for both eyes had cue elements (a horizontal or vertical segment popping out among orthogonal background segments) and identical interocularly; subjects easily perceived the cue; (iii) dichoptic cue: displays for both eyes had cues at the same location, but all the segments were interocularly orthogonal. Here, because of the brief presentation that horizontal and vertical segments were just combined binocularly, and subjects could see only scattered crosses. Thus, they could not be aware of the cue, which exists at the monocular level. After the cue display disappeared, the target displays [same as the cue display in (ii), but with an independent location of the pop-out target] were presented (ISI=400, 800, or 1200 ms). Reaction time to discriminate location of the target was measured for three subjects who fixated on a fixation point. In our results, IOR took place in conditions (ii) and (iii). This suggests that localisation of the cue occurs without visual awareness, which then leads to IOR.