Over a period of more than 30 years Hugh Mellor's writings have illuminated an enormous range of metaphysical issues to do with chance (Mellor 1971), dispositions, laws, properties (Mellor 1991) and time (Mellor 1981). His work has had a salutary influence in encouraging metaphysicians to think about these issues in clear-headed, realist ways.
His work on the metaphysics of causation (Mellor 1995), in particular, is distinguished by its rigour, cogency and originality. The main outlines of his theory of causation are well known. He has argued that causation relates facts primarily, with causation between events deriving from causation between facts; that causation comes in deterministic and probabilistic varieties; that each variety can be explained in terms of closest-world counterfactuals in which single-case chances play a crucial role; that the important connotations of causation are that causes precede, are contiguous with, are evidence for, explain and are means for bringing about effects; and that these connotations are consistent with, or imply, that causes increase the chances of their effects.
There is much in Mellor's theory of causation that I find congenial. Indeed, I hold many of the same views precisely because he has persuaded me of their truth. But we disagree on one issue that is central to the conceptual analysis and metaphysics of causation. The issue concerns whether causation is a genuine relation. Mellor believes that it is not, whereas I believe that it is. In this chapter I scrutinize his criticisms of the view that causation is a relation and, in passing, consider some related arguments for the same sceptical conclusion advanced by other philosophers. My conclusion that Mellor's scepticism on this matter is misplaced is not too surprising, but some of the arguments I rely on to reach this conclusion highlight a surprising and hitherto overlooked feature of the concept of causation.
Before detailing Mellor's criticisms of the conception of causation as a real relation, it is worthwhile considering what reasons there might be that favour