Université de Montréal,

Recent work has shown that human subjects are able to predict the biomechanical ease and the cost of controlling the endpoint stability of potential reaching movements and to use these predictions to influence their choices between reaching movements. Here we review related experimental results in which human subjects made free choices between two potential reaching movements that varied in terms of path distance, biomechanical cost, aiming accuracy and stopping requirement, and offer a brief perspective on a potential modeling approach that would explain these results. Our main results demonstrates that prior to movement onset, control constraints such as stopping and aiming do participate in a remarkably adaptive and flexible action selection process that trades-off the advantage of moving along directions of low biomechanical cost for unconstrained movements against exploiting biomechanical anisotropies to facilitate control of endpoint stability whenever the movement constraints require it. This reveals that rather than making choices based on a minimal energy cost, the nervous system can predict additional factors, other than energy or purely abstract criteria, which may influence the decision-making process, and supports a highly context-dependent view of this process, in which the subjective desirability of potential actions may be influenced by their dynamical properties in relation to the intrinsic properties of the motor apparatus