Abstract
One important theoretical hypothesis on visual crowding suggests that peripheral visual processing may contain less precise spatial and temporal phase information than foveal processing. To test this hypothesis, we created displays for which the percept depends critically on the internal representations of the spatial or temporal phase in the input stimuli. Five illusions were created and viewed in both foveal and peripheral vision: 1. The peripheral escalator. A column of ovals drifts in front of a grating tilted at 20 deg; in the fovea, the ovals are perceived to drift horizontally; in the periphery, the ovals appear to move obliquely. 2. The giddy-up illusion. Six ovals move slowly across the screen. Each oval is filled with a grating that drifts at a higher velocity. Foveally, the ovals appear to move slowly; peripherally, the ovals appear to move quickly. 3. The moving column illusion. A column of black/white diamonds shifts back and forth. Flanking columns (up to 5 degrees away) make the center column appear to drift up and down in synchrony with the direction of movement. 4. The P.D.Q. Kanizsa illusion. A 16×12 array of Kanizsa pacmen rotate in opposite directions so as to continually reassemble arrays of Kanizsa squares. As the rotation rate increases, the peripheral range over which the squares can be seen decreases. 5. Foveal-winks/peripheral-blinks illusion. Identical 3 Hz modulating fields are surrounded by white and black rings. In the fovea, the lights appear to modulate asynchronously (i.e., wink); in the periphery, the lights appear to modulate synchronously (i.e., blink). The drastically different percepts in foveal and peripheral vision cannot be explained by simple differences in their spatial resolution because blurring did not alter the percept in foveal vision. Rather, they support the hypothesis that visual processing in the periphery may contain less precise phase information.