Should the Equilibrium Point Hypothesis (EPH) Be Considered a Scientific Theory?

Abstract

The purpose of this commentary is to discuss factors that limit consideration of the equilibrium point hypothesis as a scientific theory. The EPH describes control of motor neuron threshold through the variable lambda, which corresponds to a unique referent configuration for a muscle, joint, or combination of joints. One of the most compelling features of the equilibrium point hypothesis is the integration of posture and movement control into a single mechanism. While the essential core of the hypothesis is based upon spinal circuitry interacting with peripheral mechanics, the proponents have extended the theory to include the higher-level processes that generate lambda, and in doing so, imposed an injunction against the supraspinal nervous system modeling, computing, or predicting dynamics. This limitation contradicts evidence that humans take account of body and environmental dynamics in motor selection, motor control, and motor adaptation processes. A number of unresolved limitations to the EPH have been debated in the literature for many years, including whether muscle resistance to displacement, measured during movement, is adequate to support this form of control, violations in equifinality predictions, spinal circuits that alter the proposed invariant characteristic for muscles, and limitations in the description of how the complexity of spinal circuitry might be integrated to yield a unique and stable equilibrium position for a given motor neuron threshold. In addition, an important empirical limitation of EPH is the measurement of the invariant characteristic, which needs to be done under a constant central state. While there is no question that the EPH is an elegant and generative hypothesis for motor control research, the claim that this hypothesis has reached the status of a scientific theory is premature.

The negotiated equilibrium model of spinal cord function

“…Shadmehr et al 2010;Ingram et al 2017). Among the latter, the equilibrium point hypothesis (Feldman 1966;Sainburg 2015) in particular warrants mention in order to avoid any potential confusion with the negotiated equilibrium model presented in this paper. In brief, the equilibrium point hypothesis addresses the question of how movement is initiated and controlled; it proposes that the CNS produces limb movements (or maintains limb position against changing external forces) by modifying the threshold lengths at which the muscles controlling the limb become active.…”

Neural Control of Stopping and Stabilizing the Arm

“…This type of control learns both reference trajectory and feedback gain schedules simultaneously, purely through experience and without the need for an a priori model of body and/or environmental dynamics. Model-free, trial-and-error learning may indeed be a sufficient mechanism for controlling both movement and posture, but in itself, it fails to account for evidence of internal representations of mechanical conditions that appear to allow both adaptation in—and generalization to—novel dynamic environments, such as applied force or inertial fields ( Sainburg, 2015 ). In addition, a large amount of evidence indicates that the CNS takes inertial dynamics of body segments into account when making point-to-point reaching movements ( Cooke and Virji-Babul, 1995 ; Sainburg et al, 1995 ; Ketcham et al, 2004 ), that sensory feedback is used to control evolving movement ( Flanders et al, 1986 ; Cordo, 1990 ), and that human-object interactions are planned based on information about the physical properties and mechanics of the object ( Dingwell et al, 2002 ; Cothros et al, 2006 ).…”

Estimation of Involuntary Components of Human Arm Impedance in Multi-Joint Movements via Feedback Jerk Isolation

“…As the static postural task does not include an underlying movement, we are able to apply the do-not-intervene-voluntarily paradigm. While the do-not-intervene-voluntarily paradigm does not guarantee exclusion of voluntary feedback or constant unperturbed states (Sainburg, 2015), it is widely used and accepted as a plausible approximation of both of these assumptions (Mussa-Ivaldi et al, 1985;Gomi and Kawato, 1997;Osu and Gomi, 1999). Thus, given the absence of an underlying movement, and the fact that the elements of the BIP vector π are all constant, we are able to use an estimation interval duration T est that is longer than the delay of voluntary feedback δ v in the static postural task.…”