Department of Philosophy
Département de Philosophie
University of Geneva
Université de Genève
I am currently finishing a PhD thesis defending a methods-infallibilist account of propositional knowledge:
One knows that p iff one believes that p on the basis of a method that could only yield true beliefs.
The thesis contains mainly (1) a motivation for the account, (2) a formal model for it, (3) a contextualist version of it. The following works in progress are part of it.
Introduces a new formal model of knowledge in terms of methods and based on the methods-infallibilist idea. The models are an extension of Scott-Montague neighbourhood semantics. We argue that methods models are more explanatory than both simple neighbourhood models and classical epistemic logic models, and more versatile (from E to S5) and natural than classical epistemic logic models.
A defence of the methods-infallibilist analysis of knowledge. Will bring together two drafts arguing, respectively: (1) that an infallibility condition is required to solve Gettier and explain lottery intuitions, (2) that safety and other impossibility-of-error conditons are best formulated in terms of methods.
Discusses the notion of "close possibility" that is widely used in safety-like conditions on knowledge. Two notions are distinguished: counterfactual closeness is a matter of what could in fact have happened given the circumstances at hand, and normalized closeness is a matter of what would normally happen in the type of circumstances at hand. We argue that intuitions are driven by the second but that if knowledge is the norm of action it should still be accounted in terms of the first.
Argues for a strong tie between knowledge attributions and circumstantial modal statements; consequently, contextualism about "know" should explain the context-sensitivity of know in terms of the same mechanisms as the context-sensitivity of circumstantial modality statements. In progress.
About Timothy Williamson's argument against epistemic luminosity. I argue that Williamson's Mr. Magoo-paradox relies on a wrong conception of inexact knowledge and the margin for error requirement. I put forward a new conception and a new requirement that solves the paradox. The new Margin for Error requirement is inspired by an infallibilist conception of knowledge, and makes room both for inexact knowledge and the KK principle. However, I also argue that the conditions at which KK is validated are not easy to fulfill.
(Paul Egré's work had me finally looking closely into this.)
I argue against Jason Stanley that the linguistic data does not show that "know" isn't gradable. And I argue that actually, it is to a limited extent. I discuss whether this implies that there are degrees of knowledge, and whether this helps contextualists. (So far the answer is: not obviously!)
Here are "Savoir est-il verbe gradable? (en français), the initial paper given at Jean Nicod (2/4/2006), the notes (en français) of my talk on this subject for the PhilEAs talks in Geneva (11 nov 2006).
Current version: "The Limited Gradability of Knows", in english, the text of the conference I gave at the Linguistics and epistemology conference in Aberdeen, May 12-13th. This version includes relevant linguistic data gathered from the web.
Last update: 15 Jan 2010 20:57:58.
Dernière mise à jour: 15 Jan 2010 20:57:58.
© 2006-2010 Julien Dutant unless otherwise indicated.
© 2006-2010 Julien Dutant sauf mention contraire.