Reseach
In Preparation
Probably Scientific, Definitely Political: I Do Not Have to Believe It: I will argue that the mere fact that a claim is politically salient, even though it is well-supported within a peer-reviewed literature, is enough to justify its exclusion from one's beliefs. I will model a bounded Bayesian who cannot engage with every relevant possibility and show why they should engage with the possibility of an article being biased when its claim is politically salient. Then, I will show why such engagement destabilizes their credence towards the article's claim. Finally, I will use a probabilistic account of belief to show why this destabilization is enough to justify excluding politically salient claims from their set of beliefs.
Credences and Demographic Profiling: When Not Using Total Evidence is Rational (with Lewis Ross): We explain why not integrating profiles into the formation of credences is rationally permissible. We also identify situations in which, plausibly, it is rationally impermissible to integrate that information. We begin by surveying two promising accounts of why profiling is impermissible: the moral encroachment view and the suspension of judgment view. While acknowledging the strengths of each, we raise problems for both accounts. We then outline a view based upon the literature on Bayesian bounded rationality which allows us to capture what these views got right while avoiding some of their problems. In essence, we think that which facts about a person one attends to, and hence which propositions one takes into account when forming inferences about that person, can be constrained in a way that avoids the sad conclusion while being morally, epistemically and instrumentally permissible. Explaining how and why is the task of our paper.
The Verifiability Criterion of Meaning and the Hard Limits of Science: I evaluate Carnap’s claim that a language in which metaphysical sentences are meaningless is more efficient for scientific purposes. Carnap justifies his claim by saying that metaphysical problems are irresolvable, and seeking to solve them is a waste of time. However, there are problems which are considered meaningful yet irresolvable, even in languages that Carnap endorses. Taking these problems into account, is Carnap justified when he claims that his language is more efficient than a language which admits metaphysical sentences as meaningful? I will suggest a probabilistic model for resolving this problem, which implies that a final answer to this question is impossible without empirical data. I will also show how this empirical data is useful for philosophers to decide between a solely scientific language and a metaphysical language.