Legal experts—lawyers, judges, and academics—typically resist changing their beliefs about what the law is or requires when they encounter disagreement from those committed to different jurisprudential or interpretive theories. William Baude and Ryan Doerfler are among the most prominent proponents of this view, holding it because fundamental differences in methodological commitments severs epistemic peerhood. This dominant approach to disagreement, and Baude and Doerfler’s rationale, are both wrong. The latter is committed to an overly stringent account of epistemic peerhood that dogmatically excludes opponents. The former violates the conjunction of three plausible epistemic principles: Complete Evidence, considering all epistemically permissible evidence; Independence, in which only dispute-independent evidence is epistemically permissible; and Peer Support, which involves epistemically permissible evidence. Instead, I argue for jurisprudential humility—we ought to be more willing to admit we do not know what the law is or requires, and take seriously conflicting views.

The problem of unwelcome epistemic company refers to the problem of encountering agreement with your beliefs from an unwelcome source, such as someone who is known to form unreliable beliefs or have values you reject. Blanchard (2023) and Levy (2023) argue that when we encounter unwelcome agreement, we may have reason to reduce our confidence in our matching beliefs. I argue that unwelcome epistemic company rarely provides reasons to reduce our confidence, and apparent successes at improving our beliefs using unwelcome company are explained by extraneous factors. Seeing why unwelcome agents are rarely evidence our belief is false requires making a distinction between two kinds of agents who regularly form false beliefs: unreliable agents and anti-reliable agents. While unreliable agents are common, they are uninformative. While anti-reliable agents would be informative, they are incredibly rare. Unwelcome agents are also rarely evidence that we have formed our own beliefs via an unreliable process, unless we have independent evidence that we are relevantly similar to them. This is hard to obtain given that unwelcome agents, by definition, have values and methods of forming beliefs that we do not find appealing. Moreover, attempts to use unwelcome company to improve our beliefs are likely to make our beliefs worse off in a number of ways. I argue we should adopt a policy of ignoring unwelcome company, letting them have little impact on our confidence in our beliefs.

Many of us are unmoved when it is objected that some morally or intellectually suspect source agrees with our belief. While we may tend to find this kind of guilt by epistemic association unproblematic, I argue that this tendency is a mistake. We sometimes face what I call the problem of unwelcome epistemic company. This is the problem of encountering agreement about the content of your belief from a source whose faults give you reason to worry about the belief's truth, normative status, etiology, or implications. On the basis of an array of cases, I elaborate four distinct kinds of problems that unwelcome epistemic company poses. Two of these are distinctly epistemic, and two are moral. I canvass possible responses, ranging from stubbornness to an epistemic prudishness that avoids unwelcome company at all costs. Finally, I offer preliminary lessons of the problem and distinguish it from the problem of peer disagreement.