The main approaches in sport sciences are based on conventional scientific deterministic paradigm. These studies appear not situated, out of context, parcelled (Schon, 1983) and, consequently, very far from coaches practical knowledge (Saury & Durand, 1998). Constructivist paradigm and complex sciences framework seem to be closer to practitioner representations and better suited to practical training situations (Hellard et al., 2002). These situations are characterised as they are by the specificities proper to each athlete, and by the intervention, at any one time, of a series of factors in complex interaction together, not to mention the transformation of the athlete through time. Our processes draws on a constructivist paradigmatic framework, characterised by the will to sustain a permanent reconstruction of the knowledge arrived, in terms of the finality which is being pursued, the context which prevails, and the precise nature of the problems arising out of the actual practice of training. The results thus arrived at can be deemed to constitute a body of operational knowledge, whose elaboration takes into account a criterion of pragmatic validity (the knowledge produced must be of practical applicability), and also its epistemological pertinence (the manner in which the findings have been arrived at, relative to the nature of the practice in question) (Le Moigne, 1995). Recognition of these two epistemological requirements implies the search for a series of indicators liable to be of practical significance for the coach, and which will enable the latter to objectify (Avalos et al., 2003; Hellard et al., 2002). The second step is to link local scientific knowledge to systemic meta-models. These models will be deemed a satisfactory one, if they comply with the criteria of pragmatic validity, and if they give rise to a body of thinking liable to contribute to a fuller understanding of the phenomenon involved.
© Copyright 2002 International Journal of Computer Science in Sport. Sciendo. All rights reserved.