The motion-induced position shift depends on the perceived direction of bistable quartet motion.

Motion can influence the perceived position of nearby stationary objects (Nature Neuroscience 3 (2000) 954). To investigate the influence of high-level motion processes on the position shift while controlling for low-level motion signals, we measured the position shift as a function of the motion seen in a bistable quartet. In this stimulus, motion can be seen along either one or the other of two possible paths. An illusory position shift was observed only when the flashes were adjacent to the path where motion was perceived. If the flash was adjacent to the other path, where no motion was perceived, there was no illusory displacement. Thus for the same physical stimulus, a change in the perceived motion path determined the location where illusory position shifts would be seen. This result indicates that high-level motion processes alone are sufficient to produce the position shift of stationary objects. The effect of the timing of the test flash between the onset and offset of the motion was also examined. The position shifts were greatest at the onset of motion, then decreasing gradually, disappearing at the offset of motion. We propose an attentional repulsion explanation for the shift effect.