Semantic anti-realism and the right thing to do

The most compelling aspect of the anti-realist argument is its attention to the publicity of meaning. From Wittgenstein, we have the idea that all evidence anyone has on which to improve her literacy is whatever linguistic behavior she can observe. From Dummett (via the mathematical intuitionists), we have the idea that the grasp of meanings of words can only be shown by demonstrating competence in their use. The anti-realists have shown us that when we take these concerns seriously, we have to question the merit of our concept of classical truth (via classical negation), at least in ordinary natural languages.

Those who give the anti-realist argument its due respect acknowledge that the realist position and the anti-realist position aren't in genuine conflict. Instead, the positions are really about different things. The realist position is about (classical) truth and schmeaning (i.e., intension), while the anti-realist position is about meaning and schtruth (i.e., intuitionistic truth).

Ultimately, the disagreement amounts to this: (Where X is a theorem of classical logic that isn't also a theorem of intuitionistic logic or where X is a conclusion of a classically sound inference that isn't also intuitionistically sound), the realist is willing to assert:

We can see how X is true, and we can assert that X is true.

while the anti-realist is only willing to assert

We can see how X is true (because we know that any refutation of X is an absurdity), but we can't (completely literally) assert that X is true. If we were to say that X is true, we would be speaking loosely or using metaphor, rather than using the meanings over which we have observed and shown competence. A completely literal assertion that X would require us to have a competence that it is in principle impossible to demonstrate.



Now, what I think needs to be explored is whether similar concerns arise with respect to morality or our concept "good". By now you should know that I think they do. I know of no one who has explored this issue. The closest anti-realists have come is to consider how the word "truth" operates in ethical discourse. This, of course, is a linguistic issue rather than a normative one. I, on the other hand, aim to show that the same concerns that the anti-realist has with respect to our norms of language/logic can be paralleled with similar concerns regarding our norms of behavior; we don't need to resort to determining whether the truth of moral statements is classical or intuitionistic, we can instead appeal directly to the ethical norms that make them true and reflect on whether those operate classically or intuitionistically.

The basic idea is that our ethical standards should have the same publicity requirements as Wittgenstein and Dummett have observed in our linguistic standards. We can only learn moral good and bad from observing behavior, and how morally good or bad our actions are must be demonstrable (in principle). The immediate realist objection, just like in the linguistic case, is that the right thing to do just is the right thing to do even if we could never in principle tell what it was. Sometimes, there is just no way to know what is the right thing to do.

But the anti-realist response is exactly paralleled also. The anti-realist can say "Fine, if you don't want me to call this concept 'morality', I'll just call it 'schmorality'. Now you need to demonstrate that 'schmorality' is a worthless concept. I don't think you can do that."