Nathan Lauffer

I argue for the possibility of suboptimal knowledge: knowledge that's defective given that the knower ought to inquire. I do this by uniting two independently plausible claims from the literature on inquiry that, at first glance, appear incompatible: there's the Ignorance Norm, according to which one ought not know that p and inquire into whether p; and there's the idea that, even if one knows that p, they may still be obligated to inquire into whether p. As I explain, these principles entail the possibility of suboptimal knowledge. Once this argument for suboptimal knowledge is in place, I go on to defend it. Along the way, I identify the relevance and utility of suboptimal knowledge for debates about the epistemic significance of unpossessed evidence, the relationship between knowledge and stakes, and zetetic normativity more generally.

It is often thought that the phenomenon of rational inquiry for the sake of confirmation poses a challenge for Friedman’s “Don’t Believe and Inquire” norm (DBI). It is also often thought that, insofar as that phenomenon really does pose this challenge, it also challenges the claim that inquiry is essentially question-directed and driven by interrogative attitudes. This paper develops a new argument against DBI grounded in the phenomenon of confirmation-driven inquiry. This argument can presuppose that interrogative attitudes are essential to inquiry.

It is widely agreed that there are temporal asymmetries in the availability of evidence: no proposition that is entirely about a future eventuality is supported by evidence that is causally downstream from that eventuality. This paper explores the consequences of that asymmetry for norms of inquiry. We propose a temporal symmetry principle: holding fixed an agent’s total evidence, one may inquire about a proposition when it concerns the future if and only if one may inquire about it when it concerns the past. This is in conflict with suggestions in the literature to the effect that a rational agent who does not lose evidence may lose knowledge simply by going through time. However, we suggest there may be asymmetries at the level of epistemic reasons to inquire. Among the consequences of this asymmetry, we note a distinctive problem for monists about epistemic value.

Recent epistemological literature features compelling and novel arguments for thinking that an agent can rationally believe each member of a set of propositions while knowing that one of the members is false. Perhaps more provocatively, these proponents of “Rational Inconsistency,” as it were, claim that it’s also possible to know each true member of the set while knowing that one of the members is false. Call this “Knowledgeable Inconsistency.” In this article, I explain why, if Knowledgeable Inconsistency is true, then, on pain of an absurd implication about how confirmation works, we should reject non-skeptical Infallibilism about knowledge. For that same reason, the recent proponents of Rational Inconsistency should also reject non-skeptical Infallibilism.

Recent epistemological literature features compelling and novel arguments for thinking that an agent can rationally believe each member of a set of propositions while knowing that one of the members is false. Perhaps more provocatively, these proponents of "Rational Inconsistency," as it were, claim that it’s also possible to know each true member of the set while knowing that one of the members is false. Call this "Knowledgeable Inconsistency." In this article, I explain why, if Knowledgeable Inconsistency is true, then, on pain of an absurd implication about how confirmation works, we should reject non-skeptical Infallibilism about knowledge. For that same reason, the recent proponents of Rational Inconsistency should also reject non-skeptical Infallibilism.

According to the Doctrine of Normative Defeat (‘the DND’), you may lose justification to believe that p if you fail to possess negatively relevant evidence that you ought to possess. This paper presents an objection to the DND as it’s standardly developed: it carries with it an absurd implication regarding how one’s knowledge can be restored once one’s associated epistemic justification is presumed to be normatively defeated. I defend the force of this objection before closing with a note about what my argument means for the DND's motivation.

Moeser suggested participants default to linear ordering elements but they can be primed to impose either linear or partial ordering. This study seems problematic insofar as ‘greater than’ might be understood to incline participants to favor linear orderings. Recent follow-up studies strongly suggest participants do not default to linear ordering. It seems plausible, moreover, that the observed priming effect is far more pervasive than Moeser countenanced. The present work explores the extent to which priming for linear or partial orders conflicts with the construction of linear or partial ordering lists. Our two experiments suggest friction between priming and task, and also show once the priming tasks are cleared of potentially confounding semantic content, participants perform equally with respect to constructing linear or partial orders under similar priming conditions.

In this paper, I argue that the following claims are jointly inconsistent: (1) that an agent’s justification for belief, if it’s constituted by evidence, depends on the profile of her background beliefs, (2) that whether or not a group believes a proposition is solely dependent on whether the proposition is jointly accepted by its members, and (3) that prototypical group beliefs are justified. I also raise objections to attempts to resolve the tension by retaining (2) and (3). The upshot is a novel objection to the Joint Acceptance Account of group belief since it seems to be accompanied by a kind of skepticism.