Abstract
Visual object recognition benefits from learning via self-generated rotations of 3D computer generated objects versus learning objects by watching the same rotations generated by another, unseen, participant. Here we show that during visual object recognition, the motor cortex is active only when viewing objects that have been learned with self-generated action. Thirteen subjects studied computer generated novel and familiar objects in two ways. Half of the objects were studied ‘actively’ by rotating them with a trackball on a computer monitor, whereas the other half were shown rotating without active engagement from the participant. The rotations of these ‘passively’ studied objects were recorded from a previous participant's self-generated rotations. In a subsequent test phase, participants were scanned, using fMRI, while they performed an old/new recognition task on the active and passively studied objects. Active study resulted in greater activation in motor regions of the brain when compared with passive study. This effect surfaced only when learning the appearance of novel 3D objects, and not during active and passive study of known, or familiar 3D objects. In addition, learning effects (old vs new items) occurred for both active and passive study (and for novel and familiar objects) in the visual association cortex. These results demonstrate that a) motor experience is stored in the brain and re-activated during visual recognition and b) both types of study result in the recruitment of visual association regions during recognition and c) that the motor activation seen here may facilitate visual recognition. This is the first demonstration of empirically manipulated active learning of objects resulting in motor system activation during a visual task. Such a conclusion suggests that the automatic activation of motor cortex during viewing of some objects is due to motor learning and not to the perception of implicit affordances.