Abstract
Studies have shown that the perceived value of some feature of a target stimulus, such as its color or orientation, can be modulated by stimuli around it. We wondered if such interactions were limited to modification of features, or if there are conditions under which these interactions can cause one to see a stimulus when none, in fact, occurs. Two disks were presented in two separate locations directly below fixation—a “central” location closer to fixation and a “peripheral” location farther away. Each disk was flashed a variable number (0–4) of times on a given trial. In separate blocks of trials, observer reported either the number of central or peripheral targets perceived. The near-foveal and peripheral visual signals influenced one another: when the target flashed once and the distracter more than once, observers often perceived two or more target flashes. The perceptual bias was asymmetric: A single flash in the center accompanied by multiple flashes in the periphery was perceived as multiple flashes, but multiple flashes in the center accompanied by a single flash in the periphery was not perceived as a single flash in the center. The asymmetry argues against an explanation based on cue integration, a framework often used to explain the audio-visual flash illusion that inspired this study. Moreover, whereas an auditory stimulus affects visual perception only when the numbers of flashes and beeps differ by one, here the peripheral flashes affected the perception of visual stimuli near the center even when the numbers of flashes in the two locations differed by two. The findings indicate that a) different locations in visual space interact in complex ways to radically alter visual perception, and b) the interaction between visual stimuli in different locations is of an inherently different nature than that between vision and other modalities.