Abstract
A stationary point light source in an otherwise completely dark room is often perceived as moving, a phenomenon known as the autokinetic effect. Previous studies have linked autokinesis to eye movements. However, it has remained an open question whether the autokinetic effect originates from erroneously interpreted signals related to eye movements, or it results from the motion of the stimulus on the retina (retinal theories). According to extraretinal theories, the illusory motion of the stimulus is the result of suppressed motor commands that occur during prolonged fixation. According to retinal theories, the retinal motion caused by unmonitored fixational eye movements is responsible for autokinesis. Fixational eye movements always occur during fixation, but the changes in the retinal image they produce are normally not perceived. In this study, we used retinal stabilization, a procedure by which retinal image motion is eliminated, to sort between extraretinal and retinal theories. Retinal stabilization is a powerful method to test competing theories, as it decouples the motor signals related to eye movements from the visual changes they are normally associated with. In a forced-choice discrimination task, subjects reported whether a small, bright dot, briefly displayed on a CRT, was presented at fixed location or drifted with uniform motion at 30'/s. Trials randomly alternated between the two conditions of normal viewing and retinal stabilization. In the latter condition, the position of the dot was changed in real-time in order to eliminate the retinal motion caused by eye movements. We show that the autokinetic effect correlates with retinal image motion but not with the amplitude of eye movements. These results speak against extraretinal theories and suggest that autokinesis originates from a failure of the visual system in discarding the fixational motion of the retinal image during viewing of an isolated stationary stimulus.
Supported by NIH EY18363 and by NSF BCS-0719849.