Corroborating External Observation by Cognitive Data in the Description and Modelling of Traditional Music

Any repertoire of traditional music - i.e., considered by the members of a cultural community to be part of their collective heritage - may be said to behave like a formal system. In the great majority of civilizations with oral traditions, musical systems do not give rise to abstract speculation. The theory which underlies each of them is not formally expressed, thus remaining implicit. In such a context, to describe the systematic organization of a given repertoire amounts to unearthing the rationale of (1) the principles and rules by which it is governed, (2) the ways in which it functions. Modelisation consists in eliciting the minimal utterance to which each performer ultimately refers, when actualizing any one of the musical pieces in the system, or - in a multipart music - any of its component parts. The validity of both description and modelisation results from the interaction of Western conceptual tools and the criteria of relevance obtained from pronouncements by the recipients of the tradition. Formulating the theory of a musical system requires that at each stage of the investigative process the native's cognitive grasp of musical practice and its symbolic imports - whether this grasp is conceptual or metaphoric - be corroborated by the data gathered through observation. In this paper I describe three case studies of research among "living laboratories" of African culture bearers, each of which illustrates a principle of such collective cognitive systems.