G.E. Moore, answered from the texts and cited to the page.
The question "what is good?" is not a verbal question — not a request for the ordinary usage of a word.1 What I want is the nature of the object or idea that the word "good" stands for; and my answer, disappointing as it may seem, is that good is good, and that is the end of the matter.1 Good cannot be defined. This is not a trivial result: it amounts to saying that all propositions about the good are synthetic and never analytic, which means that nobody can foist upon us an axiom like "pleasure is the only good" on the pretence that this is simply the meaning of the word.2
Now here is where the argument bites. Suppose a philosopher tells you that good just means pleasure, or just means the object of desire. Consider what follows. If good simply means pleasure, then the question "is pleasure good?" reduces to "is pleasure pleasure?" — a trivial identity, incapable of being disputed or even meaningfully asked.
But that question is not trivial; it is precisely the central question hedonism proposes to answer.3 When people say "pleasure is good," we cannot believe they merely mean "pleasure is pleasure" and nothing more than that.3 The same difficulty afflicts every naturalistic proposal without exception. It may well be true that all things which are good also have some other property — just as all yellow things produce a certain kind of vibration in light.
But those vibrations are not what we mean by yellow; we should never have discovered them at all without first noticing the patent difference of quality between colours.4 The mistake is to suppose that naming those other properties constitutes a definition of good — that the properties are not "other" at all, but absolutely and entirely the same with goodness.4
That mistake is what I call the naturalistic fallacy.4 The argument also exposes an internal incoherence in naturalistic ethics taken on its own terms. If one philosopher says good means the object of desire, and then attempts to prove that good is not pleasure, he has contradicted himself: for if good just means the object of desire, his proof that it is not pleasure can only show that pleasure is not the object of desire — it cannot show that pleasure is not good, since he has already defined good away.5
The naturalists cannot even argue with one another coherently, because any definition they adopt makes their dispute with rival naturalists either a dispute about words or an impossibility.5 And if they retreat to saying they are merely reporting how most people use the word "good," they have abandoned ethics entirely — the advice "do this, because most people use a certain word to denote conduct of this nature" is not ethical teaching; it is lexicography dressed up as morality.6
The question remains permanently open: whatever natural property N you propose, it is always a further, substantive question whether N is really good. That openness is not a gap to be closed by better science or more careful sociology. It is a structural feature of the concept itself, and it is the reason good must be treated as simple, indefinable, and non-natural.
What I want to discover is the nature of that object or idea, and about this I am extremely anxious to arrive at an agreement.Principia Ethica, pp. 6–7
propositions about the good are all of them synthetic and never analytic; and that is plainly no trivial matter. And the same thing may be expressed more popularly, by saying that, if I am right, then nobody can foist upon us such an axiom as that 'Pleasure is the only good' or that 'The good is the desired' on the pretence that this is 'the very meaning of the word.'Principia Ethica, pp. 6–7
When they say 'Pleasure is good,' we cannot believe that they merely mean 'Pleasure is pleasure' and nothing more than that.Principia Ethica, pp. 11–12
It may be true that all things which are good are also something else, just as it is true that all things which are yellow produce a certain kind of vibration in the light... far too many philosophers have thought that when they named those other properties they were actually defining good; that these properties, in fact, were simply not 'other,' but absolutely and entirely the same with goodness. This view I propose to call the 'naturalistic fallacy.'Principia Ethica, pp. 10–11
One of them says that good is nothing but the object of desire, and at the same time tries to prove that it is not pleasure. But from his first assertion, that good just means the object of desire, one of two things must follow as regards his proof.Principia Ethica, pp. 10–11
'Do, pray, act so, because the word "good" is generally used to denote actions of this nature': such, on this view, would be the substance of their teaching... 'You are to do this, because most people use a certain word to denote conduct such as this.' 'You are to say the thing which is not, because most people call it lying.' That is an argument just as good!Principia Ethica, pp. 11–12