Perception of illusory occlusion in apparent motion

Abstract

We began with a random matrix of 8 dots displayed briefly on the CRT screen. The pattern was switched off and replaced by an identical array shifted horizontally and the procedure was repeated in a continuous cycle. One of the dots in the second frame was then masked off by an opaque white piece of cardboard. The dots in the surround continued to oscillate as expected but we found that the single unpaired dot also continued to oscillate behind the occluder even though it has no "partner" in the second frame. We found that the magnitude of this illusion ("entrained motion"); increased as we reduced the distance between the inducing dots and the test dot; was unaffected by increasing the number of inducing dots; was enhanced by using slow speeds of alternation; was reduced considerably if there was no visible occluder and the dot in the second frame was simply deleted; was specific to the distance moved by the inducing dots in the surround; the strongest effect was seen when the distance between the occluder and the unpaired dot was approximately the jump-size of the inducing dots. The results suggest that the motion signal derived from the whole dot cluster or "blob" (low spatial frequencies) is spontaneously attributed to the unpaired dot. However the motion of this dot is more readily accepted by the visual system when an opaque occluder is simultaneously visible.