In walking there is a preferred frequency at which metabolic costis minimized (Holt, Hamill & Andres, 1991). This frequency is predictable as the resonant frequency of a hybrid pendulum mass-spring system (Holt, Hamill & Andres, 1990; Kugler & Turvey, 1988; Turvey, Schmidt, Rosenblum, & Kugler, 1988). The system's timing is not a fixed entity, however. For example, it has been shown that when a change in walking speed is required humans adopt a frequency-stride length relationship which minimizes metabolic cost relative to any other imposed by an experimenter (Zarrugh & Radcliff, 1978). Frequency tends to increase linearly as a function of speed. Is there a principled means by which these changes in motion coordinates defining comfort states may be understood and predicted?