Muscle stress is plainly one of the physical variables that the central nervous system probably wishes to minimize. This criterion does not uniquely define the patterns of muscle activation. It fails to explain the degree of coactivation of muscle antagonists that is widely found, and it cannot explain why two movements or movement segments that follow an identical trajectory driven by identical joint torques can be driven by different patterns of muscle activation. Muscle contraction provides for both net joint torque and limb stability. The minimization of the sum of muscle stresses, raised to any power, is an insufficient rule.
The author is with the NeuroMuscular Research Center at Boston University, Boston, MA 02215.