A quantitative approach to understanding the formation and change of coordinated movement patterns.

This study shows how the identification of relevant collective variables for behavioral patterns and evaluation of their stability may lead to a better understanding of processes underlying pattern change in coordinated human motor behavior. Based upon observed transitions in behavioral patterns, relative phase was identified as an appropriate collective variable for characterizing different patterns of bimanual coordination. The stability of this collective variable was determined through perturbation techniques. Subjects performed one of two patterns of hand coordination (in phase and out of phase) as the frequency of movement was systematically increased. Torque perturbations were administered in order to disturb the ongoing coordinative pattern, and the time required to return to the same pattern (the relaxation time) was measured. A transition from an out-of-phase to an in-phase pattern of coordination occurred at a critical frequency in all subjects. Moreover, when subjects began a trial in the out-of-phase pattern, a strong and statistically significant increase in relaxation time was observed as the frequency increased prior to the transition, indicating a loss of pattern stability. No transitions occurred when subjects began in the in-phase pattern of coordination; nor was there a change in the measured relaxation time as a function of frequency. Together with earlier results on critical fluctuations, these findings indicate that instabilities play an important role in effecting change in patterns of motor coordination. The potential relevance of these concepts and quantitative tools for other motor activities and other levels of description, for example, neural patterns, is discussed.