The Question That Stays Open

Suppose you want to close Hume's gap by definition. You say that good simply means, as a matter of what the word means, whatever maximises pleasure. Then a fact about pleasure is a fact about goodness, the gap disappears, and morality becomes a branch of the study of what produces pleasure.

Moore, in Principia Ethica in 1903, offered a test.

Take something that plainly maximises pleasure. Now ask: it maximises pleasure, but is it good?

If the definition were correct, that question would be closed in the way asking whether a bachelor is unmarried is closed. Somebody who asks it has not understood the word. But the question about pleasure is not closed. You can ask it, seriously, without any confusion at all, and reasonable people do.

Moore's conclusion is that the two do not mean the same thing, and that the proposed definition has failed.

The power of the test is that it does not care which definition you offer. Substitute anything you like. Good is what society approves of: this society approves of it, but is it good? Good is what evolution selected for: we evolved to do it, but is it good? Good is what a fully informed impartial person would choose: they would choose it, but is it good?

Every one of those questions stays open. That is a remarkable result for an argument this simple, and it is why the twentieth century of moral philosophy is largely a series of attempts to get around it.

Now the weakness, which is real and instructive.

Consider: water is H2O. That is a correct identity. And somebody who has not studied chemistry can ask, without obvious confusion, whether this H2O is really water. The question feels open to them.

So openness may show only that a definition is not obvious to the person asking, rather than that it is wrong. A correct but informative identity will feel open until you have grasped why it holds, and moral definitions might be like that.

This is generally accepted as a serious blow. What survives is something more modest and still useful: the test converts a proposed definition into a demand. If good just means this, explain why the question still feels open to competent speakers. That demand can be met, and meeting it is real work, and nobody has met it in a way that has persuaded the field.

There is one more thing to notice, because the next chapter collects on it.

If Moore is right that goodness is not identical with any natural property, then goodness is some other sort of property, and he owes an account of what sort and of how anybody comes to be aware of it. He proposed that we know it by a kind of direct intuition.

That answer is what made the argument of the next chapter available.