Attempting to explain the maximum number of phenomena using a minimum num-ber of independent principles is a long-standing aim of the scientific endeavour. The formation of general theories has always been accompanied by the art of modelling the phenomena of interest within the general framework of theories, the aim being to corroborate or falsify their inner theoretical structure by testing the model's predictions. In other words, models may be seen as applications of general theoretical principles to specific Systems and phenomena. The general principle of antagonism or competition, ubiquitous in natural non-living and living Systems, inspired Professor Perl to construct his pioneering models (i. e. PerPot and DyCoN) of specific Systems and phenomena such as the development of sport performance and motor learning (Perl, 2001, 2004). As he argued (Perl, 2004), modelling can help to analyse and understand the System under study and to predict its future behaviour. On the other hand, the use of inadequate, i. e. ex-cessively simplified or detailedi models, may limit enormously the development of knowledge about the System and contribute to a misinterpretation of research findings. Computer science, in general, and information theories of old cybernetics, in par-ticular, have served as a profound influence and Inspiration for models that seek to understand the physiological functions of the human organism during exercise. Concepts such äs homeostasis, feedback loops, information processing, memory storage, inputs and Outputs are commonly used concepts in current exercise physiology (Wilmore, Costill & Kenney, 2008). However, the introduction of complexity theories into human biology has begun to question the old Computer metaphor and has highlighted the need to develop new models that are able to capture the real nature of the human organism as a com-plex adaptive System (Beuter, Glass, Mackey & Titcombe, 2003; West, 2006). At all events, the brute-force simulations enabled by Computers of ever-increasing power are a weak substitute for capturing the basic principles that generate physiological phenomena. Hence, Computer science must develop in such a way as to offer alternative modelling techniques (e. g. biologically-motivated paradigms) for dealing with biological complexity. The change of focus brought about by complexity theories over the past few decades may be seen as a shift in world view away from looking at the brain-body äs a computer-machine System to view of the brain-body as a natural System, which it undeniably is. In other words, this shift may be construed äs a transition from treat-ing the brain-body as a direct product of an intelligent designer working on phylogenetic or ontogenetic time scales to the view of brain-body as a product of natural self-organising processes working on the same time scales. This paper outlines a modelling approach to fatigue physiology and compares the non-linear and linear modelling of such phenomena. It ends by discussing some preliminary experimental results which favour the use of non-linear explanatory models over linear ones.
© Copyright 2010 Sportinformatik gestern, heute morgen. Festschrift zu Ehren von Prof. Jürgen Perl. Veröffentlicht von Feldhaus, Ed. Czwalina. Alle Rechte vorbehalten.