Abstract

I examine the “omniscient interpreter” (OI) argument against scepticism that Donald Davidson published in 1977 only to retract it twenty-two years later. I argue that the argument's persuasiveness has been underestimated. I defend it against the charges that Davidson assumes the actual existence of an OI and that Davidson's other philosophical commitments are incompatible with the very conceivability of an OI. The argument's surface implausibility derives from Davidson's suggestion that an OI would attribute beliefs using the same methods as a fallible human interpreter. But this problem can be remedied via the adoption of an ambiguity theory of belief.