Bootstrap and rollback: Generalizing epistemic circularity

Reliabilists accept the possibility of basic knowledge-knowledge that p in virtue of the reliability of some belief-producing process r without antecedent knowledge that r is reliable. Cohen (Philos Phenomenol Res 65:309-329, 2002, Philos Phenomenol Res 70:417-430, 2005) and Vogel (J Philos 97:602-623, 2000, J Philos 105:518-539, 2008) have argued that one can bootstrap knowledge that r is reliable from basic knowledge. This paper provides a diagnosis of epistemic bootstrapping, and then shows that recent attempts at embracing bootstrapped knowledge are found wanting. Instead it is argued that such arguments are afflicted by a novel kind of generalized epistemic circularity. The ensuing view is defended against various objections, and an explanation of the source of that circularity is offered.