Disagreement, Semantics, and Meta-Semantics

Moral Twin Earth thought experiments appear to show that practical language is highly stable, and that many possible users of practical language are capable of having genuine disagreements with each other. This chapter clarifies a tempting generalization of this idea, which is that the members of every pair of possible users of moral language are capable of having a genuine disagreement. This is the Universal Disagreement thesis. It then shows how this thesis can be adapted to a contextualist semantics for ‘ought’ and other practical terms. It concludes by arguing that, for the realist, the central explanatory target is a claim about the stability of practical language.

Failures of Stability

This chapter argues against the Universal Disagreement thesis. Some possible communities use moral language, but do not have substantive disagreements with others who use their moral language differently. These are cases where the parties both use their terms with a moral role, but instead of differing over which substantive theory they follow when applying their moral terms (as in the original Moral Twin Earth cases), they differ in which additional roles they use these terms with. This is consistent with intuitions about Moral Twin Earth cases, but shows that they can lead to overgeneralizations about the semantic effects of a moral role. Instead, what needs to be explained by a meta-semantics for moral language is a more limited claim. Realists will have to show that moral terms are highly stable, but that it is possible to use a term such as ‘right’ with a moral role without referring to moral rightness.

A Common Subject for Ethics

The purpose of this paper is to conceptualize and explore what I shall call the Common Subject Problem for ethics. The problem is that there seems to be no good answer to what property everyone who makes moral claims could be talking and thinking about. The Common Subject Problem is not a new problem; on the contrary, I will argue that it is one of the central animating concerns in the history of both metaethics and normative theory. But despite its importance, the Common Subject Problem is essentially invisible on many contemporary ways of carving up the problems of metaethics and normative ethical theory. My aim, therefore, is to make progress – in part by naming the problem, but also by beginning to sketch out the contours of what gives the problem its force, by distinguishing between different paths of response to the problem and assessing some of their chief merits, and finally, by distinguishing the Common Subject Problem from another problem with which it has come to be conflated. This nearby problem is the Moral Twin Earth Problem. Whereas the Common Subject Problem is a problem about what property ‘wrong’ could refer to, the Moral Twin Earth Problem is a problem about how ‘wrong’ could refer to it. The upshot of the paper, therefore, is to rescue one of the historically significant problems in normative ethics and metaethics – a problem that is essentially about normative semantics – from the illusion that has persisted over the last twenty years that it is really, somehow, a problem about metasemantics. Once we have reclaimed this problem, we can see that it could still be a problem even if there are no distinctively metasemantic problems in metaethics at all, that it is a problem faced by a wider variety of views, and that the space of possible solutions is much wider and more interesting for normative theory, moral psychology, and moral epistemology.

Radical Interpretation and the Referential Stability of Wrongness

This chapter is one of three that draws out the consequences of Radical Interpretation for how concepts represent the world. This chapter introduces a famous ‘moral twin earth’ puzzle about the normative concept wrongness. It appears to have a distinctive referential stability: individuals or whole communities can be very mistaken in what they think makes an act morally right or wrong, but somehow they remain locked onto the moral subject matter. This chapter derives this stability as a prediction of Radical Interpretation. Radical Interpretation predicts the result when combined with first-order normative premises and premises about the conceptual role of the concept of wrongness.

Rationalist Metaphysics, Semantics, and Metasemantics

Rationalism offers an account of moral properties as a subset of the properties which serve to rationalize right actions, and these properties are fit to be the referents of our moral terms. That fitness can be exploited in constructing an externalist theory of reference determination for these terms. The resulting externalist theory draws support from standard responses to Moral Twin-Earth scenarios. The relevance of these responses to moral semantics has recently been vigorously challenged by Dowell and by Schroeter and Schroeter. The social character of meaning relations, which can explain the openness of questions about an analysis, may thereby also make Twin-Earth judgements beside the point. But the resources available to translators go beyond semantic competence and it is these resources that nonetheless make the Moral Twin-Earth responses relevant.