Abstract
After prolonged viewing of a static, oriented grating (the inducer), illusory motion ensues when gaze is transferred to a blank field: moving particles appear to stream orthogonal to the orientation of the inducer. Explanations of the classical motion aftereffect (MAE, in which motion adaptation induces opposite directed motion) cannot directly apply to a static inducer. Other accounts based on competition between orthogonal orientation channels or based on motion from eye movements are heavily debated. We propose that the illusory motion arises from responses of complex cells to a negative afterimage of the inducer. We recorded extracellular single unit responses in V1 in anesthetized, paralyzed macaques. We presented static sinusoidal gratings at the optimal orientation, spatial frequency, and size on a mean gray field. Upon removal of a stimulus that lasted 20–40s, most complex cells began to fire at an elevated rate. This after-discharge (AD) decayed over periods varying from ∼10ms to several sec. Our critical finding is that direction selective (DS), not just non-DS, complex cells in V1 show AD. Thus, following the removal of a static oriented stimulus, complex DS cells in V1 tuned to motion in both directions orthogonal to the inducer generate a strong output. To account for the AD, we used a phase-independent, oriented complex cell model and assumed that a negative afterimage of the inducer emerged peripherally in the visual system. When an inducer of preferred orientation is removed, its afterimage drives the model complex cell because it has the same orientation as the inducer. The model predicts that the AD is orientation tuned, and preliminary neuronal data support this. We are using our model to explore how DS cells can produce such strong responses to static stimuli. The explanation of illusory motion offered by our model differs from previous accounts because it requires neither eye movements nor competition between orthogonal orientations.