ABSTRACT
It is commonly assumed that everyday movements are constructed by hierarchically arranging elemental movements. The prototypical reflex is such an element. So is the neural oscillator or central pattern generator. The basic idea behind this assumption is that a movement is a complex thing put together from simpler things. This basic idea is challenged: Movements and their units are not things but relations. The pendular clocking mode of behavior is softly assembled through the nesting of lawful relations among properties, namely, action, frequency, energy, and monomials in mass and length. Borrowing an idea of Reed's (1982), the pendular clocking mode of organization is seen to be a specialized usage of "resources," roughly, gravitational, elastic, chemical, and electrical potentials supplemented by information in the specificational sense.
The foregoing does not completely exhaust the description of the petidular clocking mode, however. A person in the mode has a perspective on his or her various neuromuscular efforts. There is an 'aboutness' or 'directedness' to the activity that points to its intentional character. In very blunt terms, the content of a person's intention (approximately, the desired act) is tantamount to an exceptional boundary condition or constraint that harnesses the lawful relations and fetters the use of resources. This intimated equivalency between the philosophicallpsychological 406notion of intentional content and the physical notion of (exceptional) constraints is defensible. Closely parallel but independent arguments are offered by philosophers for the physically anomalous nature of intentional content and by physicists for the physically anomalous nature of nonholonomic constraints (as viewed from the conventional standard of physics). The most basic parallels are these: Both intentional content and nonholonomic constraints warrant a descriptive language distinct from that which describes a system's dynamics, and both warrant the status of "causal" despite their failure to instantiate (quasi) universal regularities. These parallels between intentional content and nonholonomic constraints suggest a number of directions for naturalizing intentionality. They also raise the possibility that the notion of causality may need to be "intentionalized" and separated from the notion of regularity. Finally, they suggest—and they do so rather strongly—that physical states with (symbolic) contents that are causally potent are a problem for science in the large and not, as some have implied, a problem that uniquely defines the science of cognition.
