Abstract

This thesis examines a series of questions about possibility. The starting point is a neglected claim in James Woodward’s interventionist theory of causation: interventions are non-trivially possible. Noting that interventionism is a realist theory of causation, in the first part of this thesis I explore what notion of possibility could possibly satisfy the interventionist’s objectivity demands. In chapter 1, I challenge Woodward’s argument that interventions need not be physically possible, an argument which sets the ground for what he takes to be the relevant notion of possibility: conceptual possibility. In chapter 2, I present a charitable interpretation of conceptual possibility as an objective kind of possibility arrived at via conceivability. I note that rationalist approaches to modal knowledge may fit the bill. To this end, I explore whether David Chalmers’ modal rationalism is able to furnish interventionism with a suitable notion of possibility. In chapter 3, I turn to Woodward’s Invariance(-based) View of laws to argue that, in fact, possibility ought to be physical, where physical possibility is construed as nomological possibility plus information about initial, boundary, and background conditions. Chapter 3 puts the emphasis on scientific counterfactuals. In the second half of this thesis, I explore what sorts of counterfactuals feature in scientific explanation. More specifically, I investigate their explanatory functions in connection to scientific models and model-based reasoning. In chapter 3, I argue that scientists can sometimes make sense of counterfactuals which describe situations in which the laws had been different, and show how the Invariance View can be extended to accommodate these counterfactuals. In chapter 4, I introduce and address two challenges to the interventionist claim that counterfactuals are explanatory when they are ‘same-object’ counterfactuals, that is, counterfactuals which describe invariance within the same object. These are the 'individuation challenge' and the 'idealisation challenge.' Finally, in chapter 5 I propose two desiderata any naturalised account of counterfactuals and counterfactual reasoning in science ought to fulfil and, to this end, defend the 'make-believe view of scientific counterfactuals.' This account, which uses a notion of fiction connected to the imagination, has a high unificatory potential: it accommodates regular counterfactuals, counternomics, and counterpossibles alike. Hence, the make-believe view of scientific counterfactuals outstrips the interventionist view of counterfactuals.