An agent’s belief is held arbitrarily when, from the agent’s own perspective, the belief has been influenced by some non-truth-related factor, such as the agent’s upbringing, her social circle, attitude towards epistemic risk or practical goals. For instance, if the agent thinks that she would not have believed a proposition, P, on the very same evidence had she received a different kind of upbringing or had she been more epistemically cautious, then her belief that P is formed and held arbitrarily. This project is an investigation of, what Skipper (2023) has called Rational Arbitrariness: the thesis that the agent can hold beliefs that by her own light has been influenced by non-truth-related factors. The project aims to defend a precise, well-motivated answer to the question of whether Rational Arbitrariness is true and to explore the connections and implications of Rational Arbitrariness to closely related debates in epistemology, like the debate between epistemic permissivists and impermissivists, whether rationality requires epistemic immodesty, and whether self-ascription of epistemic luck can be rational.
The key conceptual and formal tools for my project come from epistemic utility theory. Epistemic utility theory (or accuracy-fist/accuracy-oriented epistemology) enables us to measure the expected accuracy of entire belief systems (or credence functions) as well as the closeness or similarity of the two credence functions from each other.
Via epistemic utility theory, I shall investigate a strategy for harmonizing Rational Arbitrariness with the thesis of epistemic immodesty in a way that does not violate any other plausible norms. The strategy that I will explore differentiates two kinds of perspectives from which the expected accuracy of an agent’s belief system or credence function can be evaluated: (1) the perspective of the agent’s credence function and (2) a neutral perspective which does not assume the agent’s credence function whose expected accuracy we are calculating. Now, the key question of this approach is “What is this neutral perspective?”. This project will investigate whether such a neutral perspective can be identified or constructed and how to determine a set of different credence functions that are equally “good” from the point of view of this neutral perspective. 
The outcome of the project will be a detailed, precise evaluation of Rational Arbitrariness from the point of epistemic utility theory, a precise conceptualisation of an epistemically neutral perspective of evaluation, and a discussion of the role and implications of Rational Arbitrariness to closely related topics in epistemology.