Moore and Expressivism — Epoche B2
Concepts Are Not Properties: The Open Question, Gibbard's Later View and Akan Moral Vocabulary An argument about words, a conclusion about the world G. E. Moore's open question argument is the reason many people believe that goodness cannot be a natural property [1] . The argument's evidence, though, is entirely linguistic. It asks what a competent speaker can intelligibly ask, finds that "is what produces pleasure good?" stays a live question, and concludes that goodness is not the property of producing pleasure. Between those two things — what speakers can ask, and what properties there are — sits a gap that the argument never crosses. This column argues two connected claims. First, the open question argument supports a conclusion about moral concepts and no conclusion at all about moral properties , so it cannot underwrite either Moore's non-natural quality or the non-cognitivist's denial that there are moral properties. Second, the concept-level conclusion it does support is a claim about a particular evaluative vocabulary, and there is reason from Kwasi Wiredu's work on Akan moral language to doubt that it generalises [2] . Someone can deny both, and the second is the sort of claim that fieldwork could overturn. Moore's step Moore's own statement of his position is disarmingly flat: If I am asked 'What is good?' my answer is that good is good, and that is the end of the matter. Or if I am asked 'How is good to be defined?' my answer is that it cannot be defined, and that is all I have to say about it. The open question argument is the support for the indefinability claim. Offer any definition, Moore says, and it will remain significant to ask of the thing so defined whether it is itself good; if the definition were correct the question would be as empty as asking whether an unmarried man is a bachelor. Grant all of this. What has been shown is that the concept expressed by the English word 'good' is not the concept expressed by 'productive of pleasure'. Moore then infers that goodness is a simple, non-natural quality. That inference has an unstated premise: that where two concepts differ, the properties they pick out differ too. The premise is false, and we have known it is false at least since the identity of water and a chemical compound was recognised as a discovery rather than a definition. Two concepts can present one property under different aspects. The separation, and where Gibbard makes it The clearest modern statement of the separation comes from a non-cognitivist, which is what makes it useful here. Allan Gibbard's Wise Choices [3] , Apt Feelings (1990) offered a norm-expressivism: to judge something rational is to express acceptance of norms permitting it. Read at the level of properties, that looks like a denial that there are any properties of rationality or goodness at all, and it was widely read that way. His Thinking How to Live (2003 [4] ) changes the machinery and, more importantly here, changes what is being claimed. Normative thinking, on the later view, is planning: to think what one ought to do is to think what to do, and normative concepts are plan-laden. Gibbard then argues that an expressivist about concepts need not be a nihilist about properties. Whatever property is in fact "the thing to do" may be a natural property; the expressivist thesis concerns what we are doing when we think in these terms, not what the terms answer to. The position that results is expressivist about the concept and open-minded about the property. It is worth being clear that this is a change and not a clarification. On the earlier view the natural question to ask an expressivist was what moral judgements are about, and the natural answer was: nothing, they are not in that line of work. On the later view the question is admitted and its answer is left open, because the expressive thesis has been relocated to the concept. A reader who takes Gibbard's position from the 1990 book and then argues about properties is arguing with a view he no longer holds. Once that separation is made, the open question argument stops delivering an ontology to anyone. It tests concepts. Whether the property picked out by an unanalysable concept is natural, non-natural, or absent is a further question requiring further argument. Anyone still using Moore to rule out natural moral properties is running an argument about vocabulary and collecting a conclusion about the world. Is the concept-level result even general? That leaves the narrower Moorean claim: that the concept of goodness resists analysis. Notice what kind of claim this is. It is a claim about a word in a language, supported by what speakers of that language find it intelligible to ask. Such a claim has a scope, and the scope is the vocabulary examined. Kwasi Wiredu spent much of his career on the moral vocabulary of Akan [5] , his own language, and reached a conclusion that bears directly on this. In 'The Moral Foundations of an African Culture' and in Cultural Universals and Particulars (1996) he argues that Akan ethics is humanistic in a strong sense: what makes conduct good is its bearing on human interests, and the operative principle is a requirement of sympathetic impartiality — that one bring one's own interests into adjustment with those of others. He argues explicitly against the common assumption that African moral thought is grounded in religion; on his account the sanction behind Akan moral rules is social and human, not the will of a deity. Kwame Gyekye [6] , who disagrees with Wiredu about the relation of person to community, agrees with him that Akan ethics is humanistic in this way. Now run Moore's test in that vocabulary. Does "this brings the interests of all concerned into harmony, but is it good?" stay open for a competent Akan speaker in the way "is what is pleasant good?" stays open in English? Wiredu's account suggests not — that the criterion is not a discovery about the term but part of what employing it involves. If