This review contrasts two approaches to motor control that have dominated the field over the past years. One of them is built on ideas of the control theory; it assumes that neuronal structures perform computations and operates with notions such as motor programs and internal models. The alternative approach is based on physics and neurophysiology. It refutes the assumption of neural computations and operates with such notions as neuronal thresholds and equilibrium states. The two approaches have different goals. The former tries to produce a formal description of how any system, irrespective of its physics and physiology, can produce typical features of biological movement. This research may be very productive and important for such fields as robotics and prosthetics. The latter approach tries to produce a formal description of how neuromotor processes within the actual systems for movement production (for example, the human body) are organized to produce coordinated movements. Its goal is to turn motor control into a subfield of physics of living systems. Recent developments of the equilibrium-point hypothesis, referent configuration hypothesis, and the ideas of synergic control represent important steps towards achieving this goal.