Assertion, Sincerity, and Knowledge

Lottery cases (“Your ticket’s a loser,” asserted with evidence only of its low probability of winning) and Moore’s paradox (e.g. “It’s raining but I don’t believe it”) appear to support the knowledge account of assertion, according to which one should assert only what one knows.1 Since Susan does not know that Andrew’s ticket has lost, she should not assert that it has lost. Since asserting that it’s raining commits one to knowing that it’s raining, Sam should not go on to assert that he fails to believe it, since if that is true then he fails to satisfy a necessary condition on knowing it.2 The knowledge account thus seems well positioned to explain the error or incoherence in these assertions. This paper preserves an emphasis on knowledge but nonetheless presents grounds for an alternative explanation. My alternative approach divides the explanandum, explaining the error in lottery and Moorean assertions with one move (sections II and V) and the deeper incoherence in lottery assertions with another (sections III and IV). By ‘error,’ I mean the species of speech-active infelicity that J. L. Austin called an illocutionary ‘abuse.’3 By ‘incoherence,’ I mean not a directly normative but an expressive feature of the speech act: when an assertion is incoherent, in this respect, it necessarily gives prima facie evidence of confusion. In sections III and IV we’ll explore how lottery assertions distinctively represent the speaker as manifesting a confused state of mind. Of course, the speaker may not be actually confused, and the assertion may not, given special background conditions, actually manifest confusion. The point is that lottery assertions necessarily give prima facie evidence that the speaker is confused in a way that bears on her understanding of what she’s doing in making the assertion, even if that evidence is defeated by special background conditions. As we’ll see in section V, Moorean assertions are not necessarily incoherent in this respect: they do not necessarily give prima facie evidence of such confusion. To the good methodological question why work on assertion should pay special attention to lottery cases and Moore’s paradox, this is my answer: the cases highlight different aspects of assertion’s epistemic nature. Moorean assertions distinctively reveal the epistemic nature of the illocutionary norm informing assertion, whereas lottery assertions distinctively reveal a fundamentally epistemic species of incoherence that goes beyond mere illocutionary error. Here, in outline, is how the two sides of my argument fit together. The species of error at issue derives from a respect in which lottery and Moorean assertions are