Abstract

My interactivism holds that content emerges from interactivity of agents and world, that agents entertain contents in virtue of their embodiment of skills which, when embedded in the right context, robustly tie them to objects of their attitudes. This rebels against entrenched Cartesian solipsism about the mental, and, particularly, a vestige of internalism: that there exist naturalistic counterparts of Fregean modes of presentation --reifiable, internalistically constituted entities which account for the ways contents seem . They ought not be coarse-grained , nor fine-grained . But we cannot have it both ways. ;In their place, cognitive significance is explained by context-sensitive ways skills tie agents to objects; and intentional generalizations quantify over skills. Skills are roughly dispositions with intentional identities conferred by function-bestowing causal reasons for their continued presence . Unlike MOPs, skills are world-directed--specified by reference to worldly things--and multiply realizable, since what it takes to be so directed is not fixed. There is no reificational account of what it takes to have skills. ;I apply this interactivist picture to: attitude attributions, arguing that, contrary to many philosophers, ascriptions do not refer to MOPs under which agents entertain contents; and attention to pragmatics explains intuitions upon which contrary views rest; demonstratives, arguing that Kaplan's semantics of character cannot invoke MOPs qualia, arguing that qualitative content is intentional--generated by the perceptual system's retuning of itself to the world, making contents of color experience only specifiable by reference to colors-in-the-world; the worry that intentional properties aren't what drive behavior, arguing that the facts in virtue of which intentional taxonomy applies screen off the efficacy of neurologically or even computationally specified properties; and theories of rationality, arguing that internalism about rationality is wrong; being or failing to be rational depends on factors outside agents' heads