Face-contingent motion aftereffect

Abstract

The demonstration of many types of contingent visual aftereffects suggests the existence of interactions between the two contingent properties somewhere in the visual processing stream. For example, the McCollough effect, or the orientation-contingent color aftereffect, implies that interaction between color and orientation. One view holds that contingent aftereffects are due to adaptation of neurons that are sensitive to the particular combination of the two visual features; another view maintains that contingent aftereffects are the result of associative learning. Although the exact neural mechanism of contingent visual aftereffects remains elusive, it is nevertheless informative to know what pairs of visual properties can induce contingent aftereffects. Here we examined whether motion aftereffect can be made contingent on human faces. Subjects (the two authors and two naïve observers) viewed alternating expanding faces and contracting scrambled faces for 5 minutes. After various amount of delay, static test images were presented to the observers. A weak, but reliable, face-contingent motion aftereffect was reported by both the authors as well as the two naïve observers. An static face was seen briefly moving in the opposite direction to the adapting face motion, whereas a static scrambled face was seen briefly moving in the opposite direction to the adapting scrambled face. Similar face-contingent motion aftereffect was seen with rotating motion. These contingent aftereffects were very weak immediately after the adaptation, but gained strength after minutes of delay, a signature of contingent visual aftereffects. This result suggests that, somehow and somewhere in the brain, motion of faces is represented differently than motion of non-faces.