Abstract
This article discusses the training of actors, which entails assisting the performer in the development of such abstract competencies as are appropriate to that activity; maintaining a character, establishing circles of attention, delivery of text, stage ‘presence’, and conveying the illusion of naturalism. Some of these activities are incredibly subtle and detailed, and involve the shaping of multiple simultaneous actions. The successful delivery of such actions is dependent upon the smooth functioning of both the actor's body and also their body of knowledge. This particular learning activity is interesting because it involves a complex mix of practical and intellectual knowledge which the competent actor needs to integrate into a unified competency. Drawing initially on the language of artificial intelligence research, specifically approaches to the development of artificial intelligence (AI), which have been referred to as ‘top-down’ and ‘bottom-up’, it is suggested that certain aspects of actor training aimed at the construction of such a unified competency correspond to the top-down approach. It is argued that the equivalent of this approach is found in techniques which allow the transfer of appropriate metaphors which give cognitive structure to the imparted knowledge. These structuring metaphors influence and organize action, and are therefore necessary for the optimization of performance and the development of excellence. It is also suggested that in cases where new knowledge is less well supported by metaphor than older knowledge then the older knowledge will tend to persist in language and practice.