Can all muscular load sharing patterns be regarded as optimal in some sense?

Muscles crossing a joint usually outnumber its degrees of freedom, which renders the motor system underdetermined. Typically, optimization laws are postulated to cope with this redundancy. A natural question then arises whether all muscular load sharing patterns can be regarded as results of optimization. To answer it we propose a method of constructing an objective function whose minimization yields a given load sharing pattern. We give necessary conditions for this construction to be feasible and investigate its uniqueness. For linear load sharing patterns the Crowninshield and Brand (1981) objective function is reproduced – nonlinear ones require a more general formula.