And, first of all, there is an ambiguity in the very question: What is good? to which it seems some influence must be attributed. The question may mean either: Which among existing things are good? or else: What sort of things are good, what are the things which, whether they are real or not, ought to be real? And of these two questions it is plain that to answer the first, we must know both the answer to the second and also the answer to this question: What is real? It asks us for a catalogue of all the good things there are in the Universe and also which of them are good. Upon this question then our Metaphysics would have a bearing, if it can tell us what is real. It would help us to complete the list of things which are both real and good. But to make such a list is not the business of Ethics. So far as it enquires What is good? its business is finished when it has completed the list of things which ought to exist, whether they do exist or not. And if our Metaphysics is to have any bearing upon this part of the ethical problem, it must be because the fact that something is real gives a reason for thinking that it or something else is good, whether it be real or not. That any such fact can give any such reason is impossible; but it may be suspected that the contrary supposition has been encouraged by the failure to distinguish between the assertion This is good, when it means This sort of thing is good, or This would be good, if it existed, and the assertion This existing thing is good. The latter proposition obviously cannot be true, unless the thing exists; and hence the proof of the thing's existence is a necessary step to its proof. Both propositions, however, in spite of this immense difference between them, are commonly expressed in the same terms. We use the same words, when we assert an ethical proposition about a subject that is actually real, and when we assert it about a subject considered as merely possible. (§ 70 ¶ 1)

In this ambiguity of language we have, then, a possible source of error with regard to the bearing of truths that assert reality upon truths that assert goodness. And that this ambiguity is actually neglected by those metaphysical writers who profess that the Supreme Good consists in an eternal reality may be shewn in the following way. We have seen, in considering the possible bearing of Metaphysics upon Practical Ethics, that, since what exists eternally cannot possibly be affected by our actions, no practical maxim can possibly be true, if the sole reality is eternal. This fact, as I said, is commonly neglected by metaphysical writers: they assert both of the two contradictory propositions that the sole reality is eternal and that its realisation in the future is a good too. Prof. Mackenzie, we saw, asserts that we ought to aim at the realisation of the true self or the rational universe: and yet Prof. Mackenzie holds, as the word true plainly implies, that both the true self and the rational universe are eternally real. Here we have already a contradiction in the supposition that what is eternally real can be realised in the future; and it is comparatively unimportant whether or not we add to this the further contradiction involved in the supposition that the eternal is the sole reality. That such a contradiction should be supposed valid can only be explained by a neglect of the distinction between a real subject and the character which that real subject possesses. What is eternally real may, indeed, be realised in the future, if by this be only meant the sort of thing which is eternally real. But when we assert that a thing is good, what we mean is that its existence or reality is good; and the eternal existence of a thing cannot possibly be the same good as the existence in time of what, in a necessary sense, is nevertheless the same thing. When, therefore, we are told that the future realisation of the true self is good, this can at most only mean that the future realisation of a self exactly like the self, which is true and exists eternally, is good. If this fact were clearly stated, instead of consistently ignored, by those who advocate the view that the Supreme Good can be defined in these metaphysical terms, it seems probable that the view that a knowledge of reality is necessary to a knowledge of the Supreme Good would lose part of its plausibility. That that at which we ought to aim cannot possibly be that which is eternally real, even if it be exactly like it; and that the eternal reality cannot possibly be the sole good—these two propositions seem sensibly to diminish the probability that Ethics must be based on Metaphysics. It is not very plausible to maintain that because one thing is real, therefore something like it, which is not real, would be good. It seems, therefore, that some of the plausibility of Metaphysical Ethics may be reasonably attributed to the failure to observe that verbal ambiguity, whereby This is good may mean either This real thing is good or The existence of this thing (whether it exists or not) would be good. (§ 70 ¶ 2)