Paradoxical effects of practice of fast single-joint movements.

We studied the effects of extensive practice of fast, unidirectional, single-joint elbow flexions against a small extending torque bias upon the kinematic and electromyographic (EMG) characteristics of the movements as well as upon the reconstructed hypothetical control patterns (equilibrium trajectories). The subjects were tested at different distances, both with and without the bias torque prior to and after the practice sessions. The basic finding was paradoxical: The subjects did not improve their performance at the practiced task (against the bias) and at other distances in the same condition; however, they showed an increase in movement speed and a decrease in movement time at all distances in unpracticed conditions (without the bias). Changes in the EMG patterns were similar in both conditions. We hypothesize that the principle of learning the dynamics of interaction with the experimental setup in combination with a very steep learning curve form the basis for the observed paradoxical effects of practice. The equilibrium-point hypothesis of movement control provides the least controversial description of these effects as compared to the force-control and EMG-control approaches.