Causal Realism

Causal realism—in its most general form—is the view that causation is an objective feature of the world and not only a feature of our thinking (see, for instance, Bunge 1979: sect. 1.1.2; Costa 1989; Price 2001: 106; Esfeld 2011). Many enough consider themselves to be causal realists in this sense—they believe in real causal connections—and yet are happy to make do with neo-Humean characterisations of causation, for the sake of convenience, because they think it is at least empirically adequate. Although this book is meant to reveal some of the dangers of that choice of convenience, then in this particular chapter I am concerned more narrowly with more substantial accounts of the nature of causal relations. In fact, I will in the end focus specifically on causal realist accounts that in one way or another appeal to causal powers. This limitation excludes very little since most causal realist accounts involve an appeal to powers. Indeed, Michael Costa divides causal realism into causal objectivism and power realism (1989: 173), of which the former is merely a commitment to the idea that causal relations would obtain in the world even in the absence of minds perceiving them, while the latter commits to the idea that such relations are to be explained by appeal to powers.

Having identified the kind of causal realist accounts that are in focus here, which are meant to stand in contrast to neo-Humean accounts that are coupled to neo-Humean metaphysics, I will also introduce an intermediary class of accounts: those that believe causation is a substantial connection but don’t want to make a clean break with the traditional neo-Humean framework. I call them objectivist accounts (not to be identified with ‘causal objectivism’, which is just a commitment to the reality of causation). We have then (i) neo-Humean accounts, (ii) objectivist accounts, and (iii) powers-based accounts. I don’t mean this division to be exhaustive, but merely good enough for the purposes of this discussion. The neo-Humean and powers-based accounts are here understood as metaphysical contraries, but objectivist accounts have sympathies in both directions. According to the neo-Humean accounts, there is no substantial connection between causes and effects; causation is a relation that supervenes on the distribution of Kim-style events that obtain as a matter of brute fact. According to powers-based accounts, causation is a substantial connection that can be explained in terms of interactions between powerful particulars. Objectivist accounts are built on the assumption that causation is a substantial connection (contra neo-Humeanism), but their proponents nevertheless either think that the neo-Humean framework (or something very close to it) is the best we have by way of understanding causation, or they take themselves to be developing an alternative explanatory model of causation that is better than power realism. As I will argue, they end up not really being explanatory because they stay true to the neo-Humean restrictions on rational speculation.

Another way to explain the difference between neo-Humean and objectivist accounts is that the neo-Humean accounts that are of relevance here are not the ones that merely endorse neo-Huineanism because of epistemic constraints; they believe we can know—if we can know anything—that in objective reality there are no substantial causal connections. Objectivists might on the other hand endorse the neo-Humean characterisations of causation—or settle for them—because they believe an explanation of causation is beyond our reach because of some epistemic constraint. I venture to suggest that they share, or come very close to sharing, the attitude of J.S. Mill with respect to philosophical method and its limits.