Is Morality Semantically Uniform?

Metaethical positions are often characterized on the basis of their answers to a set of ‘yes or no’-questions. One of these questions is: do moral judgements express beliefs? Those who answer ‘yes’ are cognitivists, those who answer ‘no’ are non-cognitivists. Non-cognitivists are then presented with a further question: if moral judgements do not express beliefs, then what do they express? Some think that they express emotions, others that they express prescriptions, states of norm acceptance or dispositions to form sentiments of (dis)approval. This way of characterizing the debate between cognitivists and non-cognitivists seems to presuppose that morality is semantically uniform: either all moral judgements express beliefs, or they all express emotions, or prescriptions, and so on. In that respect, the term ‘non-cognitivism’ is misleading, as it suggests that everyone who is not a cognitivist, that is, everyone who does not agree with the claim that all moral judgements express beliefs, is a non-cognitivist. That is not how non-cognivitism is commonly understood in contemporary metaethics, where non-cognitivism is not just the position that not all moral judgements express beliefs, but the position that all moral judgements express some non-belief state. If the distinction between cognitivism and non-cognitivism is understood in this way, there is some logical space between them: it is possible both to deny that all moral judgements express beliefs and that all moral judgements express emotions or prescriptions. What if some judgements express beliefs and others do not? I will try to show that this may well be the case and that the question ‘Cognitivist or non-cognitivist?’ poses a false dilemma, as is suggested by Michael Gill and Michael Ridge. If that is true, important consequences follow for the way in which metaethical debates function. Many cognitivists argue for cognitivism by arguing against non-cognivitism (and the other way round). The reasoning is: if non-cognitivism can be proven wrong, then cognitivism is right (and the other way round). But if there is some logical space between cognitivism and non-cognitivism, if they are contraries rather than contradictories, this way of arguing for a position by arguing against its contender will no longer work. I will proceed as follows. First, I will explain what I mean by the assumption that morality is semantically uniform and I will give some examples in order to illustrate that the semantic uniformity of morality cannot be taken for granted. Second, I will try to understand, using some ideas from Wittgenstein, why it has nevertheless so often been taken for granted. Third, I will discuss some authors in contemporary metaethics who have noted the false dilemma between cognivitism and non-cognivitism. I will evaluate the solutions they propose to overcome it. Fourth, I will indicate, again with some help of Wittgenstein, how metaethical research is possible without the assumption that morality is semantically uniform.status: publishe