Viewed from a dynamical systems perspective, learning corresponds to a modification of the coordination dynamics in the direction of a to-be-learned pattern through the interplay of coopération-compétition between behavioral information (i.e., information specifying the to-be-learned pattern) and existing coordination tendencies. In the present experiment, we study the effect of a control parameter, namely the eigenfrequency difference between the coordinated limbs (Q ), on the acquisition of a bimanual coordination pattern. It has been shown that £2 ^ 0 results in symmetry breaking of the coordination dynamics by (1) a change in mean of <j> (i.e., the phase difference between the limbs) and (2) a change in the standard deviation of <j>. Thus, by manipulating Q before learning, it is possible to induce a shift of a pre-existing attractor toward a to-be-learned relative phase. As a result, the initial situation of competition becomes one of cooperation, prompting several important questions: Does learning in a symmetry breaking condition facilitate a stabilisation of the to-be-learned pattern? And, if so, does it help learning? And does learning in an asymmetric situation transfer to a symmetric situation ? To examine these questions, 16 subjects were instructed to perform a relative phase of 30° specified by an auditory metronome. They were assigned to 2 groups: one in which the moment of inertia (which is related to the eigenfrequency) of one of the limbs was manipulated in such a way that the relative phase difference of 30° constituted a stable fixed point for the coordination between the arm movements (`symmetry breaking` group) and one in which such a manipulation was absent (`symmetric` group). Throughout the learning process (which entailed 3 consecutive daily sessions), regular tests were conducted to compare the ability of the subjects in both groups to perform </)= 30° in the symmetric situation. In addition, the effects of learning on the underlying dynamics were monitored using a "scanning procedure" (for details, see Zanone & Kelso (1992). Journal of Experimental Psychology: Human Perception and Performance, 18 (2), 403-421). The outcome of the experiment will be presented in terms of means, constant errors and standard deviations of relative phase and its theoretical implications will be discussed.
© Copyright 2002 International Journal of Computer Science in Sport. Sciendo. Alle Rechte vorbehalten.