Abstract Large language models produce assertion-shaped outputs that often bring their users to true belief. I argue for two conclusions of different strengths: that on the best current evidence, these systems do not transmit testimonial knowledge and that hearers are not now positioned to accord their outputs testimonial uptake, which does not wait on the first. From Burge, Williamson, Goldberg, Faulkner, and Lackey, I derive a Tracking Minimum: a source transmits testimonial knowledge that p only if the truth of p , or the evidential support for p , figures in the explanation of its producing that assertion, and does so in virtue of the source’s own production process rather than of a proxy on which that process depends. It is this endogeneity requirement, not any modal profile, that the five accounts share and current systems fail. The argument engages Burge’s extension of his framework to computational sources, the interpretability literature, and the published case for AI testimony, whose dispute with mine turns on an empirically decidable question, for which an experiment is specified. The positive proposal places LLM outputs within an extended epistemology of instruments, with a result that has gone unremarked: a working barometer satisfies the condition these systems fail.
Abstract I argue that questions can be transmitted through asking just as beliefs are transmitted through testimony. Just as we rely on testifiers to assert only when their epistemic position is good enough, so (I argue) we rely on queriers to ask questions only if they are in a position to know that the question is sound, which guarantees that the asked question has an answer. Along the way, I argue against the view that we can only ask questions of people who (we know) know the answer to our questions. A recurring theme of the paper is that we learn about the norms for asking not just in those cases in which we straightforwardly comply with or violate them, but also by the ways we creatively deviate from the norms to accomplish our conversational goals.
Abstract What makes knowledge more valuable than mere true belief? This is the so-called value problem . Some epistemologists think that providing a solution to this problem is a key desideratum for theories of knowledge. One simple argument for this position starts with the claim that knowledge is intuitively more valuable than true belief. A more subtle and ambitious argument is rooted in the idea that theorists about knowledge must explain why knowledge is distinctively valuable, on pain of embarrassment; for without such an explanation, the thinking seems to go, epistemologists must confront the awkward possibility that they are spending time and energy on an unworthy subject. In what follows, I flesh out each of the arguments sketched above in a number of ways, and I argue that none of these versions succeed. The moral is that the value problem does not have the metaepistemological significance that is sometimes attributed to it.
Abstract Is there a necessary epistemic condition for ascriptions of faith that P to be appropriate? The primary aim of this article is to address this question by developing a contextualist account of ascriptions of faith. We argue that the ascription of faith that P to a subject S in a context C is semantically appropriate only if there is a salient relevant alternative to P for S in C. We also explore the implications of the contextualist account for the debate on the possession of faith. One group of philosophers argues that having faith that P is incompatible with possessing knowledge that P, while another group disagrees, maintaining that one can simultaneously have faith and knowledge that P. The second aim of the article is to argue that the linguistic facts (at least those discussed in this article) about ascriptions of faith are not sufficient to establish that either group of philosophers is on the right track.
Abstract The phrase ‘do your own research,’ popular among fringe theorists, has been criticized for encouraging audiences not to defer to the experts. However, it also encourages the listener not to defer to the speaker either. Why do fringe theorists make confident assertions and then tell their audience not to take their word for it? I argue that telling your audience to do their own research can serve several functions for speakers with low credibility. First, it can function as an epistemic virtue signal, raising the speaker’s credibility with the audience. Second, if the fringe theorist can direct their audience’s research so they appear to independently find evidence that confirms the fringe view, that evidence will have greater probative force than it would if the fringe theorist provided it directly. Fringe theorists may tell their audience to do their own research to foster the illusion of independence while, in fact, the fringe theorist has intentionally or unintentionally directed the audience’s inquiry to confirm the fringe view.
