Harman argued that the best explanation of our scientific observations involves objective facts whereas that is not true of the best explanation of our moral observations. He did not prove that this is true. That would have to involve a detailed comparison of different explanations. But there are good reasons for suspecting that he is right.
That would be disconcerting. As Plato points out, it’s very difficult to make sense of our moral thoughts if they are all subjective. Shafer-Landau adds a variety of ways that subjectivism about morality would apparently conflict with our normal ways of thinking about it.
Williams’s question is: what do we need to show in order to have confidence in our moral beliefs?
One answer is that we need to show that there is something inconsistent or incoherent about amoralism, the rejection of morality.
The answer that Williams pursues involves what he calls defusing subjectivism.
This has two parts: characterizing subjectivism in a way that rings true and asking whether subjectivism, so understood, poses the kinds of threats that Plato and Shafer-Landau identify.
The first, linguistic, formulation of subjectivism that we considered would do a reasonable job of handling Plato’s points. If it were true, our moral thinking would be no more inconsistent or incoherent than our reports of our own feelings.
But this formulation of subjectivism is inconsistent with our moral thinking; it involves more than merely expressing attitudes. This does not mean that the first formulation of subjectivism is itself inconsistent or internally contradictory. It just means that it would not help us to understand the way we think.
This is trickier than I made it seem because, as John pointed out, we do profitably discuss matters of taste. For instance, you may point out to me that mint ice cream is excellent after dinner. Even if I generally dislike it, I may be convinced by what you say and thereby come to enjoy having mint ice cream. It’s possible that this happens just because I’m suggestible or eager to please; if so, that has little bearing on what we’re talking about. But I might come to appreciate something about mint ice cream that I hadn’t previously noticed: that it’s good after dinner. If so, it seems that we had an argument leading to agreement which is not the sort of thing that I said could happen. I still think that there’s a difference between a mere expression of one’s feelings, like “I’m hungry” and a claim about what’s good or bad. I think that John’s example shows that what tastes good after dinner is not simply a subjective matter of how someone feels, though it is obviously related to matters of taste.