The Motion-Induced Contour

Abstract

An outline projection of a rectangular solid object, rotating in depth, produces a moving rotating in depth illusory contour across the empty region where an edge would normally be located even though it is not physically present. Observers see a contour demarcating the junction of the two frontal surfaces of the object when it undergoes the figural transformations of rotation-in-depth. Two experiments, in which several types of temporal change were compared, demonstrated that translation, flicker, or no motion are ineffective in producing the contour compared to rotation-in-depth. Also the four lines constituting the two inducing vertices of the object were sufficient to produce the motion-induced contour during rotation-in-depth. Three demonstrations further clarifying the nature of the motion-induced contour are also reported.