Is knowledge heterogeneous, or must we accept, following Gilbert Ryle, a dualism between theoretical knowledge and know-how? The question has an ancient origin, but is still topical, as witnessed by the recent defense of a sophisticated form of intellectualism, according to which know-how is only a special case of theoretical knowledge.
In this presentation, I propose to address this question in the context of the situated cognition program. I will focus on the relationship between know-how and the perception of affordances, and their role in anchoring an agent in a specific situation. I will defend the hypothesis that what an agent knows when he knows how to do something can be fully described in propositional terms, without allowing know-how to be reduced to theoretical knowledge. This hypothesis lies at an intermediate level between a strict dualism of knowledge and sophisticated intellectualism.