What should one believe about the unobserved? My thesis is a collection of four papers, each of which addresses this question. In the first paper, “Why Subjectivism?”, I consider the standing of a position called radical subjective Bayesianism, or subjectivism. The view is composed of two claims—that agents ought to be logically omniscient, and that there is no further norm of rationality—both of which are subject to seemingly conclusive objections. In this paper, I seek, if not to rehabilitate …

Read moreWhat should one believe about the unobserved? My thesis is a collection of four papers, each of which addresses this question. In the first paper, “Why Subjectivism?”, I consider the standing of a position called radical subjective Bayesianism, or subjectivism. The view is composed of two claims—that agents ought to be logically omniscient, and that there is no further norm of rationality—both of which are subject to seemingly conclusive objections. In this paper, I seek, if not to rehabilitate subjectivism, at least to show its critic what is attractive about the position. I show that the critics of subjectivism assume a particular view about justification, which I call the telic view, and that there exist an alternative view, the poric view, on which subjectivism is appealing. I conclude by noting that the tension between telic and poric conceptions of justification might not be an easy one to resolve. In the second paper, “Bayesianism and the Problem of Induction”, I examine and reject the two existing Bayesian takes on Hume’s problem of induction, and propose my own in their stead. In the third paper, “The Nature of Awareness Growth”, I consider the question of how to model an agent who comes to entertain a new proposition about the unobserved. I argue that, contrary to what is typically thought, awareness growth occurs by refinement of the algebra, both on the poric and the telic pictures of Bayesianism. Finally, in the fourth paper, “Objectivity and the Method of Arbitrary Functions”, I consider whether, as is widely believed, a mathematical theorem known as the method of arbitrary functions can establish that it is in virtue of systems’ dynamics that scientific probabilities are objective. I differentiate between three ways in which authors have claimed that dynamics objectivise probabilities ; and I argue that the method of arbitrary functions can establish no such claims, thus dampening the hope that constraints in what to believe about the unobserved can emerge from dynamical facts in the world.