Essays on ideal norms and non-ideal agents

My dissertation focuses on idealized epistemology and challenges that arise when we apply it to actual human agents. With the rise of formal epistemology, worries about the role that idealization plays in normative epistemology and the relevance of highly idealized norms to us non-ideal agents have become more pressing. I focus on two types of challenges to idealized norms. The first, which is the subject of the first two chapters, concerns constraints that our limited cognitive resources and time place on our epistemic lives. The second is a metaphysical challenge to theorizing about credences, on the basis that they poorly represent our actual graded belief states. The first chapter focuses on clutter avoidance worries. Gilbert Harman argues that a strong logical closure norm which would require us to believe all the consequences of our beliefs must be rejected because following such a norm would “clutter up our minds with trivialities.” Call this the Clutter Argument. The Clutter Argument has attracted broad sympathy in the literature, yet there has been surprisingly little attention paid to how the Clutter Argument is meant to work. In this paper I argue that the Clutter Argument fails. First, I argue for a novel account of clutter, the waste view, which identifies clutter with beliefs that waste cognitive resources. I motivate the acceptance of this new account in part by presenting several counterexamples to Jane Friedman’s (2018) account, which identifies clutter with beliefs about subject matters in which the agent has no interest. I argue that the only plausible clutter principle is one that rules out dynamic norms that would require we believe too much clutter. Finally, I show that even with such a clutter principle, the Clutter Argument fails because actual human agents typically believe consequences of their beliefs implicitly which does require them to clutter their minds. My second chapter takes as a starting point the recent claim from Jane Friedman that the norms of inquiry conflict with many norms of contemporary epistemology. In response to this conflict, we face a number of options. We can resign ourselves to the incoherence in the epistemic domain. We can radically revise the epistemic to eliminate the conflict. Finally, we can cleave the norms of inquiry from the epistemic, most naturally by insisting that they are, at bottom, practical. Call this final approach the separationist view. I offer a puzzle for the separationist gambit and canvas the possible responses to it. I argue that the best response is to reject a key premise: the claim that when you ought to come to believe P, and you cannot come to believe P and Q at the same time, then you are not permitted to come to believe that Q. However, I show that any satisfactory response the puzzle dissolves the original tension between the epistemic and the norms of inquiry, leaving us with no need to pursue the separationist strategy. The final chapter tackles the issue of whether comparativism about credences squares with probabilism. I offer a dilemma for comparativism about credences, the view that comparative beliefs, not real-valued credences, are psychologically real and primitive. If we require comparative beliefs to be probabilistically representable in a very strict way, then we should think that many rational agents' comparative beliefs are not representable by a probability function. On the other hand, if we require only a weaker form of probabilistic representability, then credences will fail to represent agents sufficiently well to epistemically evaluate agents.