Abstract
Apparent motion is a robust perceptual phenomenon in which observers perceive a stimulus traversing the vacant visual space between two flashed sitmuli. Despite a wealth of research on apparent motion, it remains unknown whether the underlying perceptual mechanism is constrained by principles of geometry or physics. We tested the geometry hypothesis against Newtonian mechanics in two experiments. Experiment 1 used a modified “window” paradigm (McBeath & Shepard, 1989), in which participants adjusted the position of a gap to indicate their perception of apparent motion. A Pacman-shaped object was presented in succession across two positions for 550 ms, differing by 0° to 120° in steps of 30°. Participants adjusted gap position to allow the object to pass through it smoothly. We found that adjusted gap position increased linearly with angular disparity, consistent with a curved path of motion defined by the unique center of rotation, as predicted by kinematic geometry, rather than a straight path constrained by the object’s center of mass, as predicted by Newtonian mechanics. This finding suggests that the perception of apparent motion abides by kinematic geometry. To test this conclusion more directly, participants in Experiment 2 were given a target detection task in conjunction with concurrent apparent motion (Yantis & Nakama, 1998). Similar to Experiment 1, a Pacman-shaped object was briefly presented in alternation (90° differences). Participants were instructed to respond as soon as they detected the target. We found that participants’ RTs were significantly longer when a target appeared on a curved path, compared to a straight path, suggesting that apparent motion, as predicted by geometry, disrupted target detection. Taken together, our findings suggest that the “filling-in” perception of apparent motion is guided by principles constrained by geometry but not physics.