Jonathan Ichikawa

The aim of this thesis is to develop and defend an epistemic contextualist theory of testimonial ‘knowledge’. Epistemic contextualism is a semantic thesis which holds that the truth-conditions of the word ‘knows’ partially depend on the context of attribution. They so depend because the strength of evidence required for a subject to ‘know’ a proposition depends on features of the context of attribution. As a second-order semantic thesis, however, epistemic contextualism doesn’t tell us what constitutes evidence within a context. That is, it is neutral regarding the first-order matter of what particular epistemic properties are sufficient for an attribution of ‘knowledge’ to be true in any particular context. Taking advantage of this neutrality, I combine epistemic contextualism with a first-order theory of testimonial knowledge. The result is a unique contextualist theory of testimonial ‘knowledge’ – Contextualized Pluralism. On this view, what it takes for a listener to ‘know’ some attested-to proposition depends on the context of attribution, and this because the standard of evidence required for the listener to satisfy this attribution depends on context. Motivation for Contextualized Pluralism comes from several sources. First, I will suggest that certain pairs of cases which are taken to motivate epistemic contextualism also motivate a specifically testimonial form of contextualism. Second, I will argue that Contextualized pluralism can also fruitfully contribute to debates both within and out with the epistemology of testimony. Regarding the epistemology of testimony, I will maintain that Contextualized Pluralism can shield testimonial transmission from recent counterexamples. Moreover, it can provide a compelling solution to testimonial versions of familiar sceptical puzzles. Outside of testimony, I will argue that my theory can make substantive contributions to contemporary research programs in both the epistemology of knowledge attributions and feminist philosophy of language. Regarding the former, I will argue that if we find a function-first approach to knowledge attributions plausible, then we have good reason to adopt Contextualized Pluralism. Regarding the latter, I argue that Contextualized Pluralism can model a form of epistemic injustice which occurs when agents use identity power to manipulate contextually-governed epistemic standards.

View record