Decision theory, pragmatic encroachment, and Gettier cases

In recent years, some epistemologists have argued that practical factors can make the difference between knowledge and mere true belief. While proponents of this pragmatic thesis have proposed necessary and sufficient conditions for knowledge, it is striking that they have failed to address Gettier cases. As a result, the proposed analyses of knowledge are either lacking or susceptible to counterexamples. Furthermore, I show that Gettier cases pose two problems that are specific to pragmatic accounts of knowledge. By addressing these problems through the development of a decision-theoretic account of knowledge, we can gain insight into what pragmatists think knowledge is. The result is a unique lack of deliberative skill explanation of why we lack knowledge in Gettier cases. Central to epistemological orthodoxy is the belief that only truth-relevant factors, such as the reliability of belief-forming processes or the counterfactual sensitivity of belief, distinguish knowledge from mere true belief. In recent years, however, this orthodoxy has come under attack. A number of epistemologists have argued in favor of Pragmatic Encroachment, which is the view that even if we fix all the truth-relevant factors, varying pragmatic factors can make a difference in determining whether or not a subject’s true belief counts as knowledge. Let us call any account that embraces pragmatic encroachment, a Pragmatic Account of Knowledge. The most powerful arguments for pragmatic encroachment appeal to the role that knowledge plays in deliberation. And various proponents have converged on some variant of the following principle: S knows that P only if S is justified in taking P for granted in deliberation. This principle along with a carefully crafted pair of cases offers an intuitive argument for pragmatic encroachment. For example, consider Catherine who possesses very strong but fallible evidence for her true belief that she was born in New York. We may suppose that she has been told so by her honest and reliable parents. Catherine’s true belief appears to be a paradigmatic case of knowledge and if she were filling out an employment form, she would be justified in taking her birthplace for granted. But given Catherine’s strong epistemic state, is she always justified in taking it for granted that she was born in NY? What if Catherine were caught in a sinister scheme where death would the punishment for falsely answering a question about her birthplace? In this high stakes situation, it would be irrational for Catherine to take it for granted that she was born in New York. After all, given the opportunity, she should go and double-check, gathering as much additional evidence as she can. So the epistemic principle connecting knowledge and deliberation along with our intuitions in these cases entails that even though all the truth-relevant factors remain fixed, Catherine’s true belief counts as knowledge in the low stakes scenario but does not count as knowledge in the extreme high stakes scenario. Most pragmatists also claim that the practical situation of the subject is relevant in a particular way, by determining the epistemic standards that must be met in order to know. So the reason 1I am following (Fantl and McGrath 2009) in the narrow use of this term. However, as they note, the term is sometimes used to refer to any account that allows for pragmatic conditions on knowledge (c.f. Weatherson 2005). 2Pragmatic accounts like those given by (Stanley 2005), (Hawthorne 2006) (Hawthorne and Stanley 2008), (Fantl and McGrath 2009), and (Weatherson 2012) have been described under the labels ‘subject-sensitive invariantism’, ‘interest-relative invariantism’, ‘anti-purism’, and ‘anti-intellectualism’. 3(Fantl and McGrath 2002) comes closest to this formulation. For ease of explanation, ‘believing’ will be taken to be synonymous with ‘taking for granted’.