A Defence of Objective Values

After some decades in which it has been taken for granted, in many parts of academe, that norms and values are subjective - determined by each person, or perhaps by our culture - there is now a growing interest in theories that posit objective norms and values. This project develops and defends a view of that kind. One objection concerns the role of testimony. If there are objective moral facts, and the aim of moral judgment is simply to grasp those facts, then it seems like relying on testimony should be just as good a way of forming moral beliefs as any other. Compare with other objective areas: we are comfortable forming the belief, say, that whales are mammals, based on the testimony of others. Not so with moral beliefs. We feel uneasy about forming the belief that affirmative action is unfair, say, simply because someone we trust says so. But if morality is just a set of objective facts, why should there be this difference? The article "Moral Deference and Authentic Interaction" (Journal of Philosophy, 2016), offers the following explanation on behalf of Moral Realism. Moral beliefs lead to action. And - other things equal - we want people around us to act according to their own understanding of right and wrong. We want this for at least two reasons: first, it allows us to learn about who they are, what kind of person they are. Second, it allows us to have a certain kind of personal interaction with them, as opposed to an interaction with a third party that advises them about what they should do. Another problem arises in so-called "population ethics", the theory of what we ought to do when our choice involves alternatives with different numbers of people in them. Several researchers have presented so-called "impossibility proofs", which show that any consistent theory about which outcomes are better than which leads to implications that are intuitively unacceptable. They see this as a very serious threat to the view that there are objective truths about values. The article "Priority Now without Repugnance Later" (under review) shows that we can avoid these paradoxes, by abandoning the notion of which outcomes are (impartially) better than which. Instead, we can build our theory around the notion "better for". The theory ranks outcomes as better or worse for each person, and then, in a second step, explains how the interests of affected persons provides the agent with reasons for action. A third issue concerns the mode in which our objectivist view is defended. Is it common to categorise theories as either "descriptive" (meant to capture our present normative language), or else as "revisionary" (proposals for changing our normative language). Should we defend our objectivist theory as a descriptive or a revisionary view? The article "Metaethics as Conceptual Engineering" (under review) defends the following answer: "Neither", or more accurately: "It doesn't matter". We should instead approach moral philosophy as "conceptual engineering", that is, as an inquiry the goal of which is to decide what kinds of normative concepts to use, going forward. Whether the view happens to be descriptive or revisionary is irrelevant to its defence.

-

Nonnatural Realism is the view that there are objective facts about what is good, bad, right and wrong, and that they are not reducible to anything else (such as desires or social norms). It has prominent defenders (Plato, Moore, Nagel, Parfit, Scanlon), but its doctrines about the meanings of normative words have not yet been articulated using the tools of formal semantics, nor have its metaphysics been spelled out in a formal framework. I will attempt such articulation, in order to answer two of the main objections in the literature. The first is that the view cannot explain why it is incoherent to evaluate descriptively identical things differently. Nonnatural Realists say this is because it is "part of the concept" of value that things have value "in virtue of" their descriptive properties. But exactly how this is part of the concept remains obscure. I will develop the following account: normative predicates express different senses when they are applied to particular things and to kinds, respectively. The particular-applying senses are definable in terms of the kind-applying senses, in a way that explains why descriptive twins must have equal value. I will provide independent linguistic evidence for this hypothesis, including zeugmas and conjunctive predications with kind-selecting predicates. Secondly, many opponents find it mysterious how normative facts can be fundamental and at the same time depend on descriptive facts. I will resolve this apparent conflict, by showing that normative facts about particulars have two kinds of grounds: descriptive facts about the particular and general normative facts. Furthermore, I will support the core view by showing that it can help resolve a number of outstanding problems in metaethics, concerning the possibility of cross-cultural normative disagreement; the existence of exceptionless moral principles; the nature of normative explanation; and the nature of normative epistemology.