Abstract
Visual illusions characterized by brightness enhancement, induced color, relative motion and shape distortion are not entirely determined by bottom-up stimulus parameters defining the pattern. Here, we demonstrate how top-down factors can modify the perceived strength of such phenomena. The Ehrenstein illusion requires freely moving eyes to perceive and sustain enhanced brightness; with fixation the illusion will be diminished and ultimately disappear. Freely moving eyes are also necessary for the perception of the Kanisza triangle and the watercolor effect.
A second extraretinal factor involves active motion. The two concentric rings of Pinna appear to rotate in opposite directions when the observer moves toward or away from the stimulus. When the same flow field defined by expansion or contraction is imaged onto the retina of a stationary observer, the illusory motion is reduced substantially. This is likely due to the difference in proprioceptive feedback and possibly vestibular signals in the two conditions.
A third important factor affecting the perceived strength of illusions is attention. Shape illusions due to geometric optical distortions, as well as induced brightness and color illusions, require the observer to spread attention evenly across the stimulus pattern. These illusions diminish with focal attention.
Future research may benefit from the quantification of these extraretinal factors and may shed light on our understanding of visual perception under ordinary conditions of viewing. Such studies may reveal the interaction between local and global neuronal processing involved in veridical and nonveridical perception. Extraretinal factors may be considered within the cognitive context of Einstellung (set) proposed by the early Gestaltists. Neurophysiologists aiming to understand these factors require alert behaving animals. Computational models simulating illusions must also take them into account.