Dark matters in contemporary astrophysics: A case study in theory choice and evidential reasoning

This dissertation examines the dynamical dark matter problem in twentieth century astrophysics from the point of view of History and Philosophy of Science. The dynamical dark matter problem (which should be distinguished from the cosmological dark matter problem) describes the situation astronomers find themselves in with regard to the dynamics of large scale astrophysical systems such as galaxies and galaxy clusters: The observed motions are incompatible with the visible distribution matter given the accepted law of gravitation. This discrepancy has two classes of possible solutions: either there exists copious amounts of some kind of matter that neither emits nor absorbs radiation (hence “dark”), or the law of gravitation must be revised.; Chapter 2 describes the physical and philosophical foundations of dynamical inferences—inferences from discrepancies between well-founded theoretical expectations and reliable observations to the characteristics of candidate solutions. Chapter 3 discusses the history of dark matter (beginning around 1930). Chapter 4 reviews the present evidence bearing on the dark matter problem. Chapter 5 evaluates the important candidate matter solutions in light of the available evidence. Chapter 6 evaluates two candidate gravitational solutions on evidential and methodological grounds, and addresses the problem of theory choice. I do not try to solve the dark matter problem, but to uncover and evaluate patterns of inference involved in evidential arguments for and against candidate solutions.; I show that Newton's “Reasoning from Phenomena” is a good framework from which to understand what is going on in this field. I argue that “higher order” and especially non-dynamical evidence is the best hope for solving this problem in dynamics. This is so in part because of “the dark matter double bind”: the very existence of the dark matter problem means that we cannot be sure of the overall matter distribution in astrophysical systems, and this in turn means that the observed motions by themselves cannot provide relative confirmation of any theory of gravitational interactions taking place at these scales. I use Newton's Rules of Reasoning to argue that we should retain General Relativity as our theory of gravity at galactic and greater scales, despite the lack of positive evidence to confirm it over its rivals at these scales.