RESEARCH ARTICLE

Abstract The current studies explore the informational basis of the coupling in human rhythmic movement coordination tasks. Movement stability in these tasks is an asymmetric U-shaped function of mean relative phase; 0 is maximally stable, 90 is maximally unstable and 180 is intermediate. Bingham (2001, 2004a, 2004b) hypothesized that the information used to perform coordinated rhythmic movement is the relative direction of movement, the resolution of which is determined by relative speed. We used an experimental paradigm that entails using a circular movement to produce a linear motion of a dot on a screen, which must then be coor-dinated with a linearly moving computer controlled dot. This adds a component to the movement that is orthogonal to the display. Relative direction is not un-iquely defined between orthogonal components of mo-tion, but relative speed is; it was therefore predicted that the addition of the component would only introduce a symmetric noise component and not otherwise contrib-ute to the U-shape structure of movement stability. Results for experiment 1 supported the hypothesis; movement that involved the additional component was overall less stable than movement that involved only the parallel component along which relative direction can be defined. Two additional studies ruled out alternative explanations for the pattern of data in experiment 1. Overall, the results strongly implicate relative direction as the information underlying performance in rhythmic movement coordination tasks