S~veral studies have reported accurate perception of physical dynamics from kinematic displays (e.g., Bingham, 1987; Kozlowski & Cutting, 1977; Runeson & Frykholm, 1983). In these studies observers view videotapes of human actors executing behaviors such as walking, lifting, throwing, and so on. The tapes are created in such a way that the actors' bodies are not visible: All that can be seen are the motions of major body joints. These ''patch light displays" lead to accurate judgments of gender, weight, and other properties of the actor or the environment. It is often assumed, on the basis of these data, that observers are perceiving the relevant dimensions as such.