Use of the uncontrolled manifold (UCM) approach to understand motor variability, motor equivalence, and self-motion.

Variability of motor output often has been considered a form of noise that interferes 2 with reliable performance. This assumption, however, depends on the level of the 3 motor system under consideration. For targeting tasks, variability of the end-effector 4 position will affect the consistency of targeting, depending on the task requirements 5 and the size of the target. Variability of coordination patterns used in artistic perfor6 mance may impact the aesthetics of performance. However, variability at the level of 7 the motor elements, including small variations in coordination patterns, often reflect 8 task flexibility that is only possible when the motor system exhibits sufficient mo9 tor abundance (Latash 2012). Consider, for example, performing cardiopulmonary 10 resuscitation (CPR) on an infant, with the index and middle fingers exerting the 11 necessary force to produce adequate chest compression. If fluctuation in the com12 pression force of one finger leads to a tendency for higher total force output, then 13 reduction in the forces exerted by the other fingers is necessary to compensate and 14 maintain a consistent total compression force. This is only possible because there 15 are two fingers contributing to a single total force output. Note that an alternative ap16 proach would be to attempt to control precisely, to the extent possible, the variability 17 of individual finger forces such that each finger generates approximately the same 18 force on each repetition. The presence of compensatory finger forces, however, has 19 been well documented (Latash et al. 2001, 2002a, 2002b), and is consistent with the 20 notion of a functional synergy among the motor elements (Latash et al. 2007). 21 Determining whether variability of motor output reflects motor noise or flexible 22 motor patterns is not trivial, however. Bernstein observed that when blacksmiths hit 23