Looking for synergies between the equilibrium point hypothesis and internal models.

The nervous system needs to control movement of a complex body in a complex world subject to uncertainty due to sensory and motor noise. In this issue, Mark Latash argues in favor of a framework that addresses these problems – the Hierarchical Equilibrium Point Hypothesis (as we call it here, HEPH). According to this hypothesis, movement is created by having a central controller specify a shift in the reference body configuration. This time-varying reference body configuration is passed on to lower level (e.g., limb) controllers, and finally to the lowest level controllers which set the thresholds of the tonic stretch reflex of individual muscles as assumed in the lambda model. In the context of this paper we find it useful to discuss other approaches, in particular approaches that focus on estimation (e.g. Internal models, IMs) and approaches that focus on control given estimations (e.g. Optimal control, OC)1. Specific proposals have been put forward of how the nervous system may implement estimation (neural IM) and proposals of how the nervous system may implement control (neural Control, nC). The EPH is a version of nC which suggests that the motor system simplifies the control problem by centrally adjusting parameters of the spinal proprioceptive feedback. The idea of distributed hierarchical control seems to have emerged in many communities to solve the problem of control complexity. We will briefly review the differences across approaches and highlight possible synergies.