In this paper I argue that the explanationist argument in favour of moral realism fails. According to this argument, the ability of putative moral properties to feature in good explanations provides strong evidence for, or entails, the metaphysical claims of moral realism. Some have rejected this argument by denying that moral explanations are ever good explanations. My criticism is different. I will argue that even if we accept that moral explanations are (sometimes) good explanations the metaphysical claims of realism do not follow.
I The Explanationist Argument
According to moral realists, moral properties such as justice and goodness take their own unique place in nature's ontological roll-call. Although realists disagree about the nature of these moral properties — for example, whether they are reducible or otherwise constituted by non-moral or natural properties — they all agree that such properties are genuine constituents of the world that are sometimes instantiated by objects, events or states of affairs. (I discuss what 'genuine' might mean below, in §III.1.) It is these properties, realists hold, to which our moral predicates refer, whose instances are sometimes correctly represented by our moral judgements and whose distribution we can, in favourable circumstances, come to know. [End Page 1]
The ability of putative moral properties to feature in good explanations is one perennially attractive argument in favour of the metaphysical claims of realism.1 The initially attractive thought is that moral properties earn their ontological rights in the same way as the metaphysically unproblematic properties of the natural and social sciences, namely by figuring in good explanatory theories.2 So just as, for example, a physicist may explain why an oil droplet stays suspended in an electro-magnetic field by citing its charge, or a social scientist may explain high levels of mental illness by citing income inequality, a 'moral scientist' may explain the growth of political protest movements or social instability by citing injustice.3 Likewise, just as an observer of the physicist may explain why he believes that the oil droplet is charged by citing the charge itself, and an observer of the sociologist may explain why she believes that income inequality exists by citing the inequality itself, an observer of the 'moral scientist' may explain why they believe that a situation is unjust by citing the injustice itself. In such cases, it appears that the instantiation of a moral property — injustice — is causally relevant in producing an effect — a political protest movement or moral judgement.
The principal defender of the explanationist argument is Sturgeon. Although Sturgeon's initial interest in moral explanations was to defend against an objection to realism, he has subsequently put forward the following positive argument in favour of moral — or more broadly evaluative — realism:4
So far I have focused on the relevance of the debate about evaluative explanations to evaluative epistemology, but there appear to be implications for metaphysics too. Many evaluative explanations of non-evaluative facts look like causal explanations: decency prevents people from doing certain things; injustice, like poverty, can provoke rebellions. And it is hard to see how moral properties like decency and injustice could have these effects unless they were real features of [End Page 2] the world. Many philosophers also find it hard to see how they could have such effects in the natural world unless they were themselves natural properties ... . So the acceptability of these explanations, if they are acceptable, would seem to provide an argument against skeptical views that would deny the existence of such properties, and also an argument that the properties in question are natural ones.
Sturgeon is clear that this argument generates only a presumption in favour of realism. This is because it is based on two assumptions that may turn out to be false. First, that moral explanations are sometimes good explanations. Second, that there are no independent objections to the existence of moral properties. These assumptions are clearer in a recent statement of the argument from Majors:
First of all, I take it as obvious that we do often say things like 'Jane gave him the money because she is a good person.' In such cases, as perhaps in others...