An understanding of the biomechanics and synergistic muscle groups of human standing is important to the development of theories of motor control, and is also requisite to designing systems to help persons with spinal cord injury regain some functional use of their legs. Certain researchers report a relatively small number of stereotyped postural responses, or strategies, to a perturbation. Pure strategies activate muscles on the same side of the body, and mixed strategies are thought to be sequential combinations of pure strategies.

The body was modelled with foot, shank, thigh, and torso rigid-body segments, and with 8 independently activated muscle groups. The action of muscles to cause angular acceleration of both spanned and unspanned joints was determined. Biarticular muscles such as gastrocnemius, hamstrings, and rectus femoris often acted to accelerate one of their spanned joints in the direction opposite the applied joint moment. Finite muscle strengths defined a polyhedron of possible ankle, knee, and hip joint-torques, and also defined a polyhedron of possible ankle, knee, and hip joint angular accelerations. Because the acceleration polyhedron was nearly planar, independent control of the ankle, knee, and hip joints is difficult.

A biomechanical analysis was used to explain the formation of muscles into synergis- tic functional groups. It was hypothesized that synergies consist of muscles appropriate for maximal acceleration of the body, and thus the same synergy could be used for both maximal and sub-maximal responses. Maximal acceleration determined net joint torques, but did not directly determine individual muscle forces. Therefore, depending on whether optimal joint torques were a corner point, surface point, or interior point of the polyhedron, the activations of all, some, or none of the muscles were determined. A principle of continuity was invoked to uniquely define the muscle activations from the net joint torques;​ for continuous solutions, small changes in the net joint torques resulted in only small changes in the muscle activations.

This study considered the initial response to accelerate the body from a static disturbed position towards a goal postural position. Without specification of a goal position towards which to accelerate the body, the activations of muscles were calculated for which body ac- celeration causes the center of pressure (between the foot and the ground) to be at either the toe or at the heel. The activations of muscles were also calculated for three different maxi- mal acceleration goals, or objectives:​ maximal acceleration of the horizontal component of the mass center towards a neutral position (CM Objective), and maximal corrective acceleration of the whole body towards vertical stance (Upright and Constrained Objectives). For the Constrained Objective, an additional constraint is added to the optimization problem, thus requiring that the net ankle-hip angular acceleration vector be exactly in the desired direction, with no component of acceleration normal to the desired direction vector. For the Upright Objective, the direction of the net ankle-hip angular acceleration vector is not prescribed.

For most initial positions and objectives, maximal acceleration required activation of a mixed synergy, or a combination of muscles on both the front and back sides of the body. The ankle synergy, consisting of those muscles on only one side (front or back) of the body, was optimal for the Constrained Objective. However, activation of the muscles on the back side of the body was not consistent with the proposed rules for continuous muscle activations.

Others have observed experimentally that posture is corrected by initially accelerating the hip either further into flexion or extension;​ i.e., the "wrong way." The results of this study suggest that accelerating the hip the "wrong way" is biomechanically optimal. Other experimental results, however, do not appear to be explained strictly from the biomechanics. The activation of muscles on only one side of the body is at times optimal, but most of the time it is biomechanically advantageous to activate a mixed synergy, or a group of muscles on both sides of the body. Therefore, the activation of muscles on only one side of the body, as is sometimes observed experimentally, may reflect certain neural constraints.