Abstract
A shape that is briefly flashed adjacent to another shape of similar color and texture appears as a moving protrusion that extends and retracts from the static shape (Kaniza, 1979). In two experiments, we tested for neural correlates of “polarized gamma” motion in normal adults using an array version of Kaniza's gamma motion display. The array consisted of square grating patches (20% contrast) flashing at 1 Hz, each flanked by a static collinear patch. In Experiment 1, VEPs in four subjects to the gamma display were compared to a control condition where flashing elements were presented alone. Results showed that subjects presented with flashing patches alone had large pattern onset responses, but minimal offset responses. However, subjects' responses to gamma motion were more symmetrical, with reduced onset and increased offset responses.
In Experiment 2, changes in perceived gamma motion were measured by varying the similarity of flashing and static patches. This was done by varying the contrast of the static patch from 0 to 80% (flashing patch contrast was unchanged). Eleven subjects judged the strength of gamma motion on a 5-point scale for each contrast level. VEPs were then recorded for the same stimuli. A “symmetry power index” was calculated from VEPs by dividing the total even harmonic power into the total even and odd harmonic power. Results showed subject ratings of gamma motion were highest with patches of equal contrast. Perceived motion was reduced with an increase in contrast difference. Further, symmetry indices agreed closely with subjects' ratings. These results suggest that gamma motion is apparent in the symmetry of VEP onset and offset responses, and that there is good agreement between the perceptual ratings and response symmetry. Symmetrical responses appear to reflect equal salience of illusory extension and retraction that is reduced when there is little or no perceived motion.