Research

PUBLICATIONS

Interrogating the Linguistic Argument for KK 

The Journal of Philosophy (forthcoming)

The KK thesis says that if S knows that p, then S knows that S knows that p. A common argument for KK is that it can explain why certain assertions (of the form p & ~KKp) are infelicitous. I undermine that argument: only the less plausible version of KK ("Weak KK") can explain the data, and in any case, a "believe that you know" norm on assertion can explain the data without appealing to KK. Overall, KK has much less explanatory power than is often supposed.

PAPERS IN PROGRESS (drafts available)

[1] In Truth-Aptness Defeat and Credence, I argue that upon receiving a truth-aptness defeater against P, you should not lower your credence in P—even though you may end up with a high credence in P, and a low credence that that credence has a true content. The reason is simple: if the truth-aptness defeater is right, then a lower credence is no better than a higher one.

[2] In Truth-Aptness Defeat and Firmness of Hold, I investigate the animating question of my dissertation: how should you respond to truth-aptness defeat? I argue that none of the items on the epistemologist’s menu will do: even if truth-aptness defeat does require lowering credence (contrary to [1]), merely doing so isn’t a satisfactory response to it—and neither is suspending judgment, disbelieving, etc. Far from leaving us at an impasse, though, this makes for a surprising discovery: there must exist a second dimension along which our credences can vary, and it is along this dimension that we should adjust them upon receiving truth-aptness defeat. This dimension cannot be understood in terms of indeterminacy, credal resilience, or any other familiar properties of our attitudes. In-stead, modifying an attitude along this dimension involves weakening the dispositions typically associated with it. Belief, suspension of judgment, desire, fear, etc. can vary along this dimension, too, not just credences. I call this dimension how firmly an attitude is held.

[3] In A New Case for Epistemic Permissivism, I argue that in certain cases, agents will be epistemically permitted to believe any proposition in an inconsistent set, but not permitted to believe all of them. Hence, for some proposition P, the agent will be permitted to either believe P or suspend judgment on P. (These cases are importantly different from Preface Paradox-style cases.)

[4] There is disagreement about which norms govern assertion. Beneath that disagreement, though, are two widely held assumptions: first, that the best way to adjudicate it is by seeing which candidate norms can best explain why certain assertions are infelicitous; second, that a candidate norm can explain why an assertion is infelicitous simply by counting it as improper. In The Epistemic Significance of Infelicity (draft available), I challenge both of those assumptions. First, I point out that often, candidate norms will count felicitous assertions as improper, too—which means that simply pointing to an assertion’s impropriety is not enough to explain its infelicity. Second, I show how infelicitous assertions can provide (what I call) phenomenal support for candidate norms of assertion: hearing an infelicitous assertion can justify believing a certain norm in the same way that a visual experience can justify believing that there is a desk before me. The upshot is a new and improved scoreboard for candidate norms of assertion (and other speech acts, too): one which doesn’t just pay attention to how well different candidate norms can explain certain data, but which takes into account the degree of phenomenal support those candidate norms receive.

[5] According to perspectivism, whether I’m morally obligated to ϕ only depends on facts that are within my epistemic grasp. For example: if I have no reason to believe that offering my friend cantaloupe will cure their illness, that fact cannot make it morally obligatory for me to offer them cantaloupe. In Hyperintensional Obligations and Satisfaction (draft available), I argue that perspectivism has two surprising commitments. First, that moral obligations are hyperintensional: that just like beliefs, they’re constituted by modes of presentation. Second, that in certain cases, agents will have a moral obligation to ϕ, it will be true that they ϕ—though accidentally—and yet they will not satisfy that obligation, not even in a non-praiseworthy way. Perspectivism is often supposed to be more commonsensical than its rivals. Given these commitments, however, it may be less commonsensical than is commonly thought.

this page last updated 10/25/25