Bayesian Unification in the History of Dark Matter: Confirmation by Restructuring Old Evidence

Abstract

The history of dark matter is often glossed as a linear progression of accumulating evidence confirming its existence. The actual history, however, is far more complex and philosophically nuanced. In this paper, I revisit key moments in the establishment and justification of dark matter as a viable hypothesis and offer a Bayesian representation of the core inferences made at those moments. I argue, contrary to the textbook narrative, that certain evidential reclassifications played a pivotal confirmatory role even in the absence of new data. In particular, I show how unifying previously separate anomalous phenomena under a single modeling framework provided novel support for the dark matter hypothesis.