Abstract

It is widely assumed that G. E. Moore was either oblivious or indifferent to circularity worries surrounding his enigmatic proof of an external world. I argue that this assumption is false. Drawing on unpublished archival evidence and overlooked passages in Moore’s posthumously published lectures, I provide, for the first time, an account of Moore on circular proof. I show that, as early as 1928–29 and as late as 1938–39, Moore identified an “important” and “unimportant” sense of begging the question (epistemic circularity and premise circularity, respectively) and maintained that a genuine proof should avoid both forms. Curiously, however, this standard is not upheld in his 1939 “Proof,” where this “important” sense is conspicuously absent from Moore’s discussion. This discrepancy raises an interpretative puzzle: Why does Moore’s standard for proof change? I provide an answer and explore its philosophical implications, clarifying the paradoxical nature of the proof and re-diagnosing the sense of intellectual dissatisfaction experienced by many commentators.