ABSTRACT
Roland Stout is concerned, as I am in chapter 4 of Making It Explicit, to offer an account of rational agency that entitles one to endorse Kant’s understanding of the difference between agency and the causal processes characteristic of the rest of the natural world as the difference between acting according to a conception, representation, or idea (Vorstellung) of a rule and acting according to a rule. In Sellars’s sophisticated treatment of this idea in “Some Reflections on Language Games”,1 he satisfies this desideratum by requiring that rule-guided (as opposed to merely regular) activity involve an actual representation of the rule, and be sensitive to that representation in the sense that changing the representation (while holding other factors constant) changes the behavior. It was clear already to Sellars that such a definition cannot be appropriate to the normativity that articulates discursive intentionality, since responding appropriately to representations of rules already requires understanding the concepts in terms of which the rule-representation is couched. Stout is principally concerned to introduce quite a different way of understanding the Kantian dictum, in terms of second-order rules enjoining the following of firstorder rules. He recommends a shift from describing agentive processes as using rules to describing them as referring to rules – second-order rules.
