On the relativity of perceived motion.

Perceived stability of the visual world during eye movements is traditionally explained as due to the presence of extraretinal signals, equal in magnitude to retinal signals. Motion is perceived when the two signals differ. An experiment is reported in which motion thresholds were measured during smooth pursuit eye movements. The results show that the traditional view is incomplete. Motion is only perceived when the two signals differ by at least a just noticeable difference (JND), the magnitude of which depends on ocular velocity and is independent of the direction of stimulus motion relative to the eyes. The data lead to the rejection of theories according to which ocular velocity is under-represented in extraretinal signals. In addition they show that retinal image motion carries no information about stimulus motion. Perceived motion, direction and velocity are relative concepts. They depend on the JND and therefore they are relative to extraretinal signals. This principle explains the Filehne illusion and even predicts the Aubert-Fleischl phenomenon. A similar analysis can be applied to understand vestibular effects on motion perception.