I have said that those systems of Ethics, which I propose to call Metaphysical, are characterised by the fact that they describe the Supreme Good in metaphysical terms; and this has now to been explained as meaning that they describe it in terms of something which (they hold) does exist, but does not exist in Nature—in terms of a supersensible reality. A Metaphysical Ethics is marked by the fact that it makes the assertion: That which would be perfectly good is something which exists, but is not natural; that which has some characteristic possessed by a supersensible reality. Such an assertion was made by the Stoics when they asserted that a life in accordance with Nature was perfect. For they did not mean by Nature, what I have so defined, but something supersensible which they inferred to exist, and which they held to be perfectly good. Such an assertion, again, is made by Spinoza when he tells us that we are more or less perfect, in proportion as we are more or less closely united with Absolute Substance by the intellectual love of God. Such an assertion is made by Kant when he tells us that his Kingdom of Ends is the ideal. And such, finally, is made by modern writers who tell us that the final and perfect end is to realise our true selves—a self different both from the whole and from any part of that which exists here and now in Nature. (§ 67 ¶ 1)

Now it is plain that such ethical principles have a merit, not possessed by Naturalism, in recognising that for perfect goodness much more is required than any quantity of what exists here and now or can be inferred as likely to exist in the future. And moreover it is quite possible that their assertions should be true, if we only understand them to assert that something which is real possesses all the characteristics necessary for perfect goodness. But this is not all that they assert. They also imply, as I said, that this ethical proposition follows from some proposition which is metaphysical: that the question What is real? has some logical bearing upon the question What is good? It was for this reason that I described Metaphysical Ethics in Chapter II as based upon the naturalistic fallacy. And that a knowledge of what is real supplies reasons for holding certain things to be good in themselves is either implied or expressly asserted by all those who define the Supreme Good in metaphysical terms. This contention is part of what is meant by saying that Ethics should be based on Metaphysics. It is meant that some knowledge of supersensible reality is necessary as a premise for correct conclusions as to what ought to exist. This view is, for instance, plainly expressed in the following statements: The truth is that the theory of Ethics which seems most satisfactory has a metaphysical basis......If we rest our view of Ethics on the idea of the development of the ideal self or of the rational universe, the significance of this cannot be made fully apparent without a metaphysical examination of the nature of self; nor can its validity be established except by a discussion of the reality of the rational universe. The validity of an ethical conclusion about the nature of the ideal, it is here asserted, cannot be established except by considering the question whether that ideal is real. Such an assertion involves the naturalistic fallacy. It rests upon the failure to perceive that any truth which asserts This is good in itself is quite unique in kind—that it cannot be reduced to any assertion about reality, and therefore must remain unaffected by any conclusions we may reach about the nature of reality. This confusion as to the unique nature of ethical truths is, I have said, involved in all those ethical theories which I have called metaphysical. It is plain that, but for some confusion of the sort, no-one would think it worth while even to describe the Supreme Good in metaphysical terms. If, for instance, we are told that the ideal consists in the realisation of the true self, the very words suggest that the fact that the self in question is true is supposed to have some bearing on the fact that it is good. All the ethical truth which can possibly be conveyed by such an assertion would be just as well conveyed by saying that the ideal consisted in the realisation of a particular kind of self, which might be either real or purely imaginary. Metaphysical Ethics, then, involves the supposition that Ethics can be based on Metaphysics; and our first concern with them is to make clear that this supposition must be false. (§ 67 ¶ 2)