Lightweight Nondescriptivism

Moral nondescriptivism holds that moral claims and moral judgments don’t, qua moral judgments, undertake any particular descriptive commitments. My intention here is to offer a lightweight version of nondescriptivism by offering two theses that I think collectively entail nondescriptivism, but neither of which involves hefty psychological assumptions.

These two theses are descriptive insufficiency and functional conceptualization. 

According to descriptive insufficiency, moral words or concepts fail to have descriptive content because they do not meet whatever the requirements are for having any particular content. It is extremely plausible that linguistic or conceptual content depends to some extent on the use of the word or concept. Other things (such as naturalness) may play a role, but a word or concept cannot have a determinate meaning without having a fairly stable or united use. Meaning is an achievement for a sound or mental token, and not free. Hence it should be possible for a word or concept to fail to have the kind of use necessary for having some meaning without being totally unlike the words or concepts that do have meanings.

Nondescriptivism only requires that moral contents, and the words or concepts that they are built out of, lack that kinds of uses that are sufficient for descriptive content. 

One very plausible route to descriptive inconsistency is through an externalist semantics. It is plausible that the meanings of at least some of our words or concepts depend upon the meaning attributed to them from the community that we live in. What I mean by the words I use is determined in part by an intention to use those words in the community-standard way.  What happens if there is no community-standard way to use a given word? Plausibly, the word will then lack descriptive content. If I intend to mean the same thing by ‘right’ as everyone else, and there is no fixed descriptive content that everyone else collectively can be plausibly interpreted as using, then my word lacks descriptive content. 

According to functional conceptualization, what it takes for a judgment to fall under the concept of a moral judgment is for it to have a certain functional role (which doesn’t by itself guarantee any particular representational content). That is, we characterize moral judgments by their motivational rather than their descriptive role.

There is nothing special about thinking that we might define a concept in this way. It needn’t imply anything too substantive about the essential nature of any mental state that falls under it.  We could also formulate the concept of a judgment formed-on-a-Tuesday that applied to all judgments formed on Tuesdays. Or the concept judgment-that-makes-me-think-about-France that applied to all judgments that made me think about France.  Neither of these concepts would cut psychology at its joints, but nondescriptivism needn’t undertake any commitment to do so. 

Nondescriptivism could simply state that our concept of a moral judgment collects together judgments that share a certain feature regardless of whether that feature forms part of the judgments psychological essence. The particular relevant features include the relevance (direct or indirect) to motivation. To make a moral judgment is to make some judgment or other that plays the moral-judgment-role. Moral judgments qua moral judgments have no descriptive content.

This thought makes good sense of the fact that motivation plays an important part in how we think about the judgments of foreign cultures. Why think that the Romans generally thought about right and wrong? It can’t be because we see that they made the same decisions we do, because they often didn’t. Plausibly, it is because they reacted to their judgments in the same fashion as we do. 

These two theses are independently plausible and come together to form a natural kind of nondescriptivism. The difference between ordinary descriptive beliefs and moral judgments is twofold. First, moral judgments are distinguished by their role, rather than their content (though they may also have content) and second, moral judgments need not have, and probably often lack, any particular descriptive content.

One nice thing about this lightweight form of nondesriptivism is that it needn’t imply that actual moral judgments lack descriptive contents. It may be that they do, or it may be that they don’t. Different cultures might employ concepts that have different descriptive contents but collectively count as the same moral concepts by virtue of their role. Or it may be that some people in our society have moral concepts with certain descriptive contents and other people have moral concepts with distinct descriptive contents, or no contents at all.

This might help explain why some people insist that noncognitivism is absurd because it tells them what they mean with their own concepts. If content externalism is true, it shouldn’t be obvious what we mean. But even so, we can agree that realists do have beliefs with genuine descriptive contents, but as a whole there is no particular descriptive content shared by everyone who makes moral judgments. (And, unlike hybrid versions of noncognitivism, we don’t need to hold that every moral judgment has some descriptive content or other.)