How to Play the Lottery Safely?

Abstract

According to the safety principle, if one knows that p, one's belief that p could not easily have been false. One problem besetting this principle is the lottery problem – that of explaining why one does not seem to know that one will lose the lottery purely based on probabilistic considerations, prior to the announcement of the lottery result. As Greco points out, it is difficult for a safety theorist to solve this problem, without paying a heavy price. In this paper, I first reject three existing safety-based solutions to the lottery problem, due to Pritchard, Sosa, and Broncano-Berrocal. Failure of these accounts reveals that there is something crucial missing in the safety principle. By way of remedying this, I propose to integrate a safety principle with a condition regarding one's own awareness of (nearby) error-possibilities. The resulting account, as I argue, enjoys a number of theoretical advantages, including its capacity to handle the lottery problem.

Inductive knowledge and lotteries: Could one explain both ‘safely’?

“…That said, we believe enough has been said to show that the present way of developing a safety-based account holds some serious promise (cf. Zhao, 2020aZhao, , 2020b In sum, the overall lesson for the safety theorist is that knowledge is not just a matter of a belief obtaining external modal stabilities. Whether the subject herself is aware of such stabilities is also relevant for our epistemic evaluation.…”

“…As other versions of safety, Safety(weak) is relativized to close worlds where one forms the belief with the same method (/basis) as in the actual world. Elsewhere one of us argues that not just any method can give rise to a safe belief and knowledge (Zhao, 2020b). If one forms a belief by way of wishful thinking, then even if that belief is stably true in all close possible worlds, intuitively the belief is not knowledge (cf.…”