Kevin Coffey

It is sometimes said there are two ways of formulating Newtonian gravitation theory. On the first, matter gives rise to a gravitational field deflecting bodies from inertial motion within flat spacetime. On the second, matter’s accelerative effects are encoded in dynamical spacetime structure exhibiting curvature and the field is ‘geometrized away’. Are these accounts of Newtonian gravitation theoretically equivalent? Conventional wisdom within philosophy of physics is that they are, and recently several philosophers have made this claim explicit. In this paper I develop an alternative approach to Newtonian gravitation on which the equivalence claim fails, and in the process identify an important but largely overlooked consideration for interpreting physical theories. I then apply this analysis to (a) put limits on the uses of Newtonian gravitation within the methodology of science, and (b) defend the interpretive approach to theoretical equivalence against formal approaches, including the recently popular criterion of categorical equivalence.

Many scientific theories purport to describe empirically inaccessible aspects of the world. The agenda of foundational physics is more ambitious still: to characterize the world at the level of 'primitive ontology'. In this pursuit it often posits new and peculiar physical features, a proclivity aided by the abstract, mathematical way in which foundational theories are framed. But how do we decipher their physical content, and in particular the accounts of primitive ontology they offer, and why think that content is true? This dissertation explores the relationship between these questions -- between the problem of ontological interpretation in foundational physics and the nature of scientific realism. I do not argue for a unified account. Rather, I examine three contexts in which their interactions are particularly interesting and philosophically significant. First, I consider the role scientific methodology plays in a prominent naturalistic defense of scientific realism: the no-miracles argument. The most sophisticated form of this argument rests on the broad claim that successful scientific methodology is irredeemably theory dependent. I examine the nature of this theory dependence within the context of Maxwell's development of electromagnetism, and argue that the presence of competing ontological interpretations undermines the realist's attempt to draw sweeping epistemic conclusions from methodological success. I next consider the concept of theoretical reformulation and how it applies to the claim that Lagrangian dynamics is a reformulation of (part of) Newtonian dynamics. How should this claim be understood in light of the fact that the world is non-classical? What can this tell us about the concept of theory reformulation itself, which plays a central role in arguments against scientific realism? I provide an analysis of theory reformulation in terms of counterfactual interpretative judgments, and then cast doubt on its justification in the classical case by developing a non-Newtonian interpretation of Lagrangian dynamics. Finally, I consider whether competing ontological interpretations raise underdetermination problems for realism about foundational physics. After first re-formulating the underdetermination argument to avoid recent objections that it fails to pose any distinctive threat, I suggest a formulation of realism that vitiates the underdetermination threats posed by competing ontological interpretations.