Abstract des Autors

In modeling complex behavioral phenomena, there is often a tension between choosing a theory in which the constituent elements have a restricted set of properties and choosing one in which there is a variety of elements with a range of contrasting properties. In the first case the range of behaviors that can be generated by the model would be much more limited than in the second case. However, if those bahaviors map on to what is actually abserved in real life, the explanatory power of the model with fewer assumptions is generally recognized as greater. Another point in favor of the simple approach is that, if the model's predictions and behavior diverge, the choice of what to change in the model is more constrained and, so, potentially easier. In this chapter we review a theoretical model - the parallel force unit model - that we have recently proposed as an account of force control (Ulrich & Wing, 1991). The model follows the route of choosing very simple elements; they are all identical, and no interactions are assumed. Even though this may sound unpromising as a structural basis for accounting for behavior, the model does capture a number of previously documented aspects of the mean force-time function of brief, isometric force impulses. However, the model also makes a prediction about the variability of the force-time function, and this has not previously been investigated. In the second part of this paper we describe a preliminary study of variability in isometric force impulses. Broadly speaking, the results support the theoretical predictions. However, there is an important discrepancy, and we discuss the implication of this for future developmet of the model. Verf.-Referat