Threading neural feedforward into a mechanical spring: how biology exploits physics in limb control.

A solution is proposed of the hitherto unsolved problem as to how neural feedforward through inverse modelling and negative feedback realised by a mechanical spring can be combined to achieve a highly effective control of limb movement. The revised spring approach that we suggest does not require forward modelling and produces simulated data which are as close as possible to experimental human data. Control models based on peripheral sensing with forward modelling, which are favoured in the current literature, fail to create such data. Our approach suggests that current views on motor control and learning should be revisited.