Motor imagery: perception or action?

Motor imagery has been studied using subjective, behavioural and physiological methods and this paper reviews theoretical and practical issues from all three viewpoints. Attempts to measure motor imagery on a subjective scale have met with limited success but alternative methods are proposed. Research on mental practice suggests a number of different processes may be needed to explain the variety and variability of effects obtained. Recent studies of spatial and motor working memory signify the importance of a primarily visuo-spatial component in which actions are consciously represented together with a more properly motoric component which must be activated to generate either images or overt actions. Finally the question of whether motor imagery is primarily perceptual or motoric in character does not have a simple neurophysiological answer due to the highly distributed nature of motor control. Nevertheless some of the key mechanisms serving both spatial and motoric components have been provisionally identified.