Abstract
This paper addresses the challenge posed to our ordinary conception of theoretical rationality by cases of so-called "epistemic acrasia", which can be defined prima facie as cases in which a subject finds himself believing what he otherwise knows to have good reasons ( theoretically sufficient reasons) not to believe. Indeed, how can we account for empirical phenomena as commonplace as lying to oneself or willful blindness, without admitting a partition within our reasons for believing, between theoretical and practical reasons, for example, which would be incommensurable with each other, or which, at the very least, would maintain conflicting relationships, undermining our representation of reason as a unitary, coherent and, in law, universal law-making faculty?
Focusing in particular on the case of lying to oneself, our aim will be to propose a descriptive and explanatory framework that not only allows us to account for the possibility of this type of theoretical irrationality, but also to evaluate it on a normative level: if we have genuine "epistemic duties" - and we will defend the thesis that we do indeed have them - in what sense is the person who derogates from them by lying to himself or herself responsible? From the perspective of an ethics of belief, can we go so far as to assert that he is committing a fault ? This is an opportunity to question the legitimacy and relevance of importing certain moral categories and norms of practical rationality into the sphere of epistemological reflection and the understanding of theoretical rationality.
