Is Practical Knowledge in Any Sense Based on Observation?

Knowledge of our own intentional actions is normally conceived of as knowledge that is not acquired by observing those actions. However, since actions are part of the observable world, it has been suggested that observation still has to play some kind of genuine epistemic role. In this paper I try to reject this claim. I try to block the suggestion that observation must be, at least in some cases, a necessary, even if not sufficient, component of an agent's knowledge of what she is intentionally doing. Any rejection of the theory that this species of knowledge is non-observational has to include the assumption that some ingredient of it is perceptually acquired. Thus, if my argument is right, it must follow that practical knowledge must indeed be non-observational.