Abstract
Recent work has shown that brief exposure to static images depicting an agent actively gazing at an object produces a motion aftereffect in the direction of the gaze. Such findings suggest that encoding of an implied agent-to-object motion takes place. Here, we adapted a previously used random dot motion paradigm, quantifying changes in perceptual decision thresholds for detecting left vs. right motion, to include probes in different spatial locations following the agent-and-object image. As with motion aftereffects from real motion, the aftereffects following exposure to the image were constrained to a limited portion of the visual field, specifically on and immediately above the line of sight of the agent in the image. The effect disappeared in control experiments featuring a blindfolded agent. These findings strongly support the notion that our brains encode others' attentive gaze as an implied agent-to-object motion, and for the first time reveal the spatial extension of this internally generated motion signal.