Explanation and Dependence.

Abstract

The deductive-nomological account, various causal accounts, and various unificationist accounts of explanation have all taken explanations such as the one of the motion of the planets by Newton’s theory of gravity and mechanics to be a paradigmatic example of explanation and of explanatory advancement within the sciences. Newton’s theory of gravity increased our understanding of a wide range of phenomena and nonetheless many were troubled by the notion of action at a distance that the theory postulates. Newton himself can be seen to take an ambivalent attitude towards the explanatoriness of his theory. On the one hand, he claims to have explained a range of phenomena from the law of gravity, but nonetheless he acknowledges the lack of a causal explanation. I think that this kind of situation is neither incredibly rare nor limited to peripheral cases in the history of science. In addition to the example from Newtonian gravity we can find this attitude towards certain quantum mechanical explanations that seem to require the acceptance of non-locality, and, in a somewhat different way, towards the role of spacetime in the explanation of inertial motion in general relativity.

I argue that these cases pose a serious difficulty for the unificationist account and, in particular, for the causal account. Moreover, I take these cases to be illuminating as to the nature of explanation and I develop an account of explanation based on a notion of dependence that is broader than causal dependence that allows us to account for this attitude. Doing so opens up the possibility of rehabilitating the explanatory status of laws by providing a way of addressing the counter-examples to the deductive-nomological account that does not rely on replacing the role of laws as providing a relation that does explanatory work with causal relations or a relation of unification. Lastly, this allows us to understand the debates about explanatory status in Newtonian gravity, in quantum mechanics over EPR style cases, and over the role of spacetime in general relativity as arising from empirical issues rather than from conceptual disagreements about the nature of explanation.