I am inclined to think that we know with assurance part of the analysis of ethical terms; and I wish to describe briefly two elements which I think we know belong to an analysis of "right" in its normative sense. What I wish to propose is that there are two truth criteria for ethical statements, two material conditions under which we reject ethical statements as incorrect. The first of them has been discussed by many philosophers. It is expressed in Kant's categorical imperative, in the views of Hume and Adam Smith that revolve about their idea of the impartial spectator, in Sidgwick's axiom of justice, and in the writings, I think, of Mr. Ayer and Mr. Stevenson. Roughly, the point is that, if a person says, "This action A is wrong, but those other actions, B, C, D, which are essentially similar to it, are not wrong," we refuse to accept his statement. If A is to be wrong and B is to be right, there must be some difference between the natures of A and B. One way of putting it is that we demand ethical consistency. It is not easy to state just what ethical consistency is. We might put it tentatively thus: it is always mistaken for a person to say that A is wrong and B is right unless there is some property P which A has, and which B has not, such that he is willing to say, for all cases of the occurrence of P, that the presence of P makes the situation morally worse than it would otherwise be. Failure to follow this rule in one's moral utterances is not playing the game to which one commits himself in the use of moral language. Differences of opinion on matters of ethics we tolerate; but it is hard to tolerate moral inconsistency when a person is aware of it, for such inconsistency shows he is deliberately misusing moral language for the sake of personal gain or prejudice. What anyone is definitely committing himself not to do, when he uses moral language, is to give expression to what he personally wants; there is, as Hume said, a generality about moral language.
Let us consider now the second truth criterion. Suppose a person P says, "A is wrong," and shortly after says, "A is right." We might be puzzled by this, but perhaps we should say: "P has seen a light; now we know where he stands." There is no rule in English against saying A is right if one once said it was wrong, although there is a rule which prescribes that now one must say that formerly he was mistaken, since it would not make sense to say that the same action could have been wrong at an earlier time and right at a later time. This rule, however, is not my point. What I do wish to emphasize is that, while English usage does not prohibit changing one's mind in ethics, it does prohibit uttering an ethical sentence on the basis of a momentary impulse. When we say something is wrong, we definitely give the impression of a more than temporary attitude and commitment.
(Richard B. Brandt, "The Emotive Theory of Ethics," The Philosophical Review 59 [July 1950]: 305-18, at 315-6 [italics in original])
Note from KBJ: Philosophers are experts, but their expertise is neither evaluative nor factual. It is conceptual. Philosophers are trained to analyze concepts. The tool with which they do this is logic. The only argumentative leverage a philosopher has is the law of noncontradiction, which says that no proposition can be both true and not true. Thus, if you believe proposition p, then you cannot consistently believe not-p or anything that entails not-p. Suppose I wish to induce you to believe proposition p. I will have to show you that p follows logically from some other proposition (or conjunction of propositions) that you (1) already believe but (2) are not willing to give up. If you are willing to give it up, then you may refrain from believing p (or even believe not-p). There is always more than one way to eliminate an inconsistency. If I show you that three of your beliefs are inconsistent, then you must get rid of one of them. Which one you get rid of is up to you, not me. In a sense, philosophers are bearers of bad tidings, which may explain why they annoy people. They are constantly telling people that they (the people) have inconsistent beliefs. If they go beyond this and tell the people which belief(s) to give up, they have stopped being philosophers and started being preachers.
