Arthur N. Prior*

IF there is any contribution to moral philosophy which is more likely than any other to become permanently associated with the name of Professor G. E. Moore, it is the identification and refutation, in his Principia Ethica, 1 of what he calls the 'naturalistic fallacy'. I propose now to explain what it is to which Professor Moore gives this name, and what he considers to be involved in its fallaciousness; and I shall offer reasons for regarding his argument, not as disproving ethical naturalism itself, but as exposing an inconsistency into which some naturalists have fallen.

What Professor Moore means by the 'naturalistic fallacy' is the assumption that because some quality or combination of qualities invariably and necessarily accompanies the quality of goodness, or is invariably and necessarily accompanied by it, or both, this quality or combination of qualities is identical with goodness. If, for example, it is believed that whatever is pleasant is and must be good, or that whatever is good is and must be pleasant, or both, it is committing the naturalistic fallacy to infer from this that goodness and pleasantness are one and the same quality. The naturalistic fallacy is the assumption that because the words 'good' and, say, 'pleasant' necessarily describe the same objects, they must attribute the same quality to them. We might, with Mill, call the objects to which a term is applicable the denotation of the term, and the characteristics which an object must have for the term to be applicable to it, the connotation of the term. 2 What the man who commits the naturalistic fallacy fails to realize is that 'good' and some other adjective may denote or be applicable to the same things, and yet not con- note the same quality, i.e. describe the things in the same way. The difference between identity of denotation and identity of connotation may be brought out, as Professor Moore shows, by the following simple consideration: If the word 'good' and, say, the word 'pleasant' apply to the same things, but do not attribute the same quality to them, then to say that what is pleasant is good, or that what is good is pleasant, is to make a significant statement, however obvious its truth may appear to many people. But if the word 'good' and the word 'pleasant' not merely have the same application but the same connotation or 'meaning' - if, that is to say, the quality of pleasantness is identical with the quality of goodness - then to say that what is good is pleasant, or that what is pleasant is good, is to utter an empty tautology, or, as Mill would call it,3 a 'merely verbal' proposition; for both statements are on this supposition merely ways of saying that what is pleasant is pleasant.

From this consideration Professor Moore attempts to show that the term 'good' is incapable of definition. By 'definition' he means the exhibition of a quality referred to by some term as a combination of simpler qualities. And he argues that if we take any such combination of relatively simple qualities (such as the combination 'being what we desire to desire'), the statement that what possesses this combination of qualities is good (e.g. the statement that what we desire to desire is good) will always be found on careful inspection to be a significant statement and not a mere truism (like 'What we desire to desire, we desire to desire'). But this is not all that he claims to be able to show by this method. We may use it, he thinks, to show that goodness is not only simple, i.e. incapable of analysis into simpler parts, but unique. For even if we take a simple quality, such as pleasantness, we can always see that it is significant, and not a mere truism, to assert that what possesses this quality is good. (Despite his definition of 'definition', as analysis, he slips readily into calling 'Good means pleasant' a 'definition' too.)

This latter contention of Professor Moore's is exceedingly difficult to state with any precision. It plainly does not apply to the quality of goodness itself - it is a truism to assert that what is good is good. Nor does it apply to the quality of goodness itself when it is merely given another name, such as 'value' (which is often used as synonymous with 'goodness' by Professor Moore, as well as by many other writers). Yet if we merely say that goodness is not identical with any other quality, this is itself a truism - it merely tells us that goodness is not identical with any quality, simple or complex, with which it is not identical. It is a little ominous that Professor Moore quotes on his title-page the sentence from Bishop Butler, 'Everything is what it is, and not another thing'. For who would deny this? Even the man who identifies goodness with pleasantness, i.e. who regards 'good' as a mere synonym of 'pleasant', would not deny that it is in this sense 'unique'. For pleasantness also 'is what it is, and not another thing'; and to say that goodness is pleasantness is not, on such a view, to deny that it is what it is, or to affirm that it is another thing - it is merely to deny that pleasantness is 'another thing'.

Professor Moore's real aim, of course, is to show that goodness is not identical with any 'natural' quality. This is why he calls the kind of identification which he is opposing the 'naturalistic' fallacy. But what does he mean by a 'natural' quality? He attempts an answer to this in the Principia, but now says that the answer there given is 'utterly silly and preposterous', 4 as indeed it is (there is no need to reproduce it here). And at times it looks very much as if what he means by a 'natural' quality is simply any quality other than goodness or badness, or at all events other than goodness, badness, rightness, wrongness, and obligatoriness (if the last three are taken to be distinct from goodness and badness and from one another - which, in the Principia Ethica, they are not), and compounds containing these. But if this is what he means, are we not back where we were? - are we not still left with the truism that 'Everything is what it is, and not another thing'?

It is worth examining this sentence in its original context. Butler's argument in the paragraph from which it is taken 5 is directed against people who were putting it about that it can never be to any man's interest to be virtuous, since disinterestedness is of the essence of virtue. Mandeville, holding that nothing is virtuous but 'self-denial', went so far as to say that virtue was not only not in a man's own interests but generally not in anyone else's either, so that 'private vices' were 'public benefits'. So Butler sets out to show that virtue and disinterestedness are not the same thing (though virtue and self- interest are not the same thing either).

'Virtue and interest, are not to be opposed, but only to be distinguished from each other; in the same way as virtue and any other particular affection, love of arts, suppose, are to be distinguished. Everything is what it is, and not another thing. The goodness or badness of actions does not arise from hence, that the epithet, interested or disinterested, may be applied to them, any more than that any other indifferent epithet, suppose inquisitive or jealous, may or may not be applied to them; not from their being attended with present or future pleasure or pain; but from their being what they are; namely, what becomes such creatures as we are, what the state of the case requires, or the contrary.

Butler is not, I think, denying that the moral quality of an act is determined by its other qualities - he is not denying, for example, that in a given situation a certain intensity of jealousy is always wrong, i.e. 'unbecoming' to 'such creatures as we are'. But he is denying that anything of this sort - expressing jealousy of such-and-such an intensity in such-and-such a situation - is what we mean by calling an act good or bad. Its goodness or badness is its 'moral appropriateness' to our nature and our situation. That is, its goodness or badness is its goodness or badness; it is its 'being what it is', good or bad as the case may be. Goodness or badness cannot be identified with any 'indifferent' epithets.

But what kind of epithet is that? If we take 'indifferent' to mean merely 'non-moral' - i.e. if an 'indifferent' epithet is any one that does not mean the same as 'good' or 'bad' - is not Butler's argument open to the same objection as Professor Moore's? Certainly goodness and badness are not to be identified with any qualities that are other than goodness and badness; but how does this forbid us to identify goodness with disinterestedness? Does not the identification of goodness with disinterestedness merely remove the latter from the class of 'indifferent' epithets, i.e. from the class of the 'other things' which goodness is not (just as, on Butler's own view, what we have called 'moral appropriateness' is something that goodness is, 'and not another thing')?

I think we must take it that what Butler means by 'indifference', and Professor Moore by 'naturalness', is something more than mere non-identity with goodness or badness. Their view seems to be that all qualities other than goodness and badness have something positive in common - something which is so near to universal that we do not notice it until we compare the qualities marked by it with goodness and badness; and then it is intuitively evident. When we compare such qualities as goodness and badness with such qualities as pleasantness, pinkness, everlastingness - to take a quite random selection - we see that the former and the latter are not only individually non-identical, as pleasantness and pinkness are, but fall into two quite different categories or 'realms', namely, those which we sometimes call the realm of value (or of duty) and the realm of fact. These terms are not perhaps quite fortunately chosen, since it may be held - it is held by Professor Moore, for example, and was by Butler - that to say that something is our duty, or possesses value, is to state a fact, albeit of a very peculiar kind. (We shall see in the sixth and seventh studies that there are some writers who deny this; but such a denial seems to amount to saying that there is really only one realm - the 'natural' one - and this is not the position which we are at present trying to formulate.) But however we describe these two 'realms', their existence and distinctness is what seems to be referred to in Professor Moore's distinction between ethical predicates and all 'natural' ones, as it is in the old distinction betweeen the 'moral' perfections of the Deity and His 'natural' ones (omnipotence, omniscience, eternity, &c.), and in Aristotle's distinction between the 'ethical', the 'natural', and the 'logical' fields of inquiry. And Aristotle notes that 'the nature of each of the aforesaid kinds of proposition is not easily rendered in a definition, but we have to try to recognize each of them by means of the familiarity attained through induction, examining them in the light' of certain 'illustrations' given previously - 'ethical' questions being illustrated by 'Ought one rather to obey one's parents or the laws, if they disagree?' and 'natural' ones by 'Is the universe eternal or not? 6 (Aristotle is here using 'induction' to mean, not a process of reasoning, but the examining of instances until their common quality 'dawns' upon one - his appeal is to intuition.) But such an intuitively perceived difference between 'moral' qualities and all others plainly goes far beyond anything that can be proved from the principle that 'Everything is what it is and not another thing', since this principle would ' still apply within a single 'natural' realm even if there were no other. 7

Professor Moore's appeal to this truism, and the little dialectical device which he bases upon it, are not, however, entirely pointless. For there are occasions when men implicitly deny logical truisms, and need to be reminded of them; namely, when they are inconsistent. It is not against the naturalist as such, but the inconsistent naturalist, the man who tries to 'have it both ways', that Professor Moore's type of argument is really effective and important. And such people are not uncommon. Professor Moore himself mentions them - the people who begin by laying it down as a truth of primary importance, perhaps even as something rather revolutionary, that nothing is good but pleasure, or that nothing is good but what promotes biological survival, and who, when asked why they are so certain of this, reply that 'that is the very meaning of the word'. To such people it is certainly legitimate and necessary to reply that if pleasantness, or the promotion of survival, is what 'goodness' means, then the fact that only pleasure is good, or that only what promotes survival is good, is hardly worth shouting from the house-tops, since nobody in his senses ever denied that what is pleasant, and only what is pleasant, is pleasant, or that what promotes survival, and only what promotes survival, promotes survival. What these people would plainly like to hold is that goodness is both identical with pleasantness and not identical with it; and, of course, it cannot be done. They want to regard 'What is pleasant is good' as a significant assertion; and it can only be so if the pleasantness of what is pleasant is one thing, and its goodness another. On the other hand they want to make it logically impossible to contradict this assertion - they want to treat the opposing assertion that what is pleasant may not be good as not merely false but logically absurd - and this can only be done if pleasantness and goodness are taken to be identical. To represent an opponent's position in such a way as to make it not only false but self-contradictory is a dialectical triumph which can never be obtained without being duly paid for; and the price is the representation of one's position as not only true but a truism. 'If a denial is to have any value as a statement of matter of fact', as Dr. J. N. Keynes says, 8 then what it denies 'must be consistent with the meaning of the terms employed.... The denial of a contradiction in terms . . . yields merely what is tautologous and practically useless.'

It is sometimes pointed out by naturalists that there is never more than one ethical statement which is rendered trivial by a naturalistic definition of 'good'. If, for example, we use 'good' as synonymous with 'conducive to biological survival', then, while it is a truism to say that what is conducive to biological survival is 'good' in this sense, it is not a truism to say that pleasure is, since it is not a truism to say that pleasure is conducive to survival. We shall find shortly that there is a point at which this consideration is important; but if Professor Moore's argument is regarded as a criticism of the attempt to deduce significant assertions from definitions, this answer to it is irrelevant, since the statement which the definition makes trivial is always precisely the one which it is put forward to 'prove', in a sense in which it is not trivial but significant. A man who has defined 'good' as 'conducive to biological survival', with the express purpose of establishing it as an ethical principle of primary importance that only what conduces to survival is good, will not be greatly cheered by the consideration that it is 'only' this principle which the definition renders insignificant.

Confronted with Professor Moore's argument, an inconsistent ethical naturalist has two courses open to him. He may clear himself of inconsistency, on the one hand, by abandoning his naturalism - he may continue to insist that only pleasure, or conduciveness to survival, or whatever it may be, is good, but may preserve the significance of this assertion by sacrificing its certainty, admitting that its denial, though still in his opinion false, is not self-contradictory. Professor Moore writes as if this is what any naturalist who really grasps his argument will do - he seems to consider his argument a refutation of naturalism. But a naturalist can preserve his naturalism if he wants to, even in the face of Professor Moore's argument - he can do so by admitting that the assertion that, say, pleasure and nothing but pleasure is good, is for him a mere truism; and that if Ethics be the attempt to determine what is in fact good, then the statement that what is pleasant is good is not, strictly speaking, an ethical statement, but only a way of indicating just what study is to go under the name of 'Ethics' - the study of what is actually pleasant, without any pretence of maintaining that pleasure has any 'goodness' beyond its pleasantness. He might add at the same time that he is not only not going to discuss goodness as a 'non-natural' quality, but that in his belief there is no such quality, and that this is worth shouting from the housetops, as it liberates us from a transcendental notion which has haunted us too long. (He might say that this is what he really means by the assertion that 'Nothing is good but pleasure' - he means, not that what is pleasant alone possesses some other quality called 'goodness', but that there are no qualities beyond 'natural' ones such as pleasantness to which the word 'goodness' could be applied.) Indeed, he is bound to say something of this sort if he is to justify his appropriation of the word 'good' for the purpose to which he puts it. And such a man, it seems to me, should be prepared to state his position in an alternative way, namely, as a denial that there is such a study as Ethics - he should be prepared, for the sake of clarity, and to further the mental 'liberation' in which he is primarily interested, to call his inquiry into the sources of pleasure, not Ethics, but some such name as 'Hedonics'; or if he defines goodness as 'conduciveness to survival', to call his substitute for Ethics 'Biological Strategy'.

But how - as Mr. E. F. Carritt pertinently asks 9 - can we be 'liberated' from a notion which we cannot ever have had? For how can we have had a 'transcendental' notion of goodness if the word which is alleged to have called it up is also alleged to have no meaning, or none beyond ones which are not 'transcendental' at all? Even this question it is not beyond the power of a consistent naturalist to answer.

'A name', as J. S. Mill points out, 'is not imposed at once and by previous purpose upon a class of objects, but is first applied to one thing, and then extended by a series of transitions to another. By this process ... a name not unfrequently passes by successive links of resemblance from one object to another, until it becomes applied to things having nothing in common with the first things to which the name was given; which, however, do not, for that reason, drop the name; so that it at last denotes a confused huddle of objects, having nothing whatever in common; and connotes nothing, not even a vague and general resemblance. When a name has fallen into this state, . . . it has become unfit for the purposes either of thought or of the communication of thought; and can only be made serviceable by stripping it of some part of its multifarious denotation, and confining it to objects possessed of some attributes in common, which it may be made to connote.' 10

And this, a naturalist may say, is precisely what has happened with the word 'good', and what needs to be done about it. At present, when we call a thing good we may mean that it is pleasant, or that it is commanded by someone, or that it is customary, or that it promotes survival, or any one of a number of things; and because we use the same term to connote all these characteristics, we think there must be some other single characteristic which they all entail; but in fact there is not. When it is said that being good means promoting survival, we are dissatisfied; we feel that it is still significant to say that promoting survival is good; and the same thing happens with every identification that is suggested; but this is just because, in each case, the other meanings are still hovering in our minds - to say that promoting survival is good is significant because it means that to promote survival is what we desire; to say that what we desire is good is significant because it means that what we desire promotes survival; and so on. Once we realize this, we may either recommend and adopt a more consistent usage; or we may leave the word with its present 'flexibility', but with the misleading suggestions of that flexibility removed. The naturalist who proposes some unambiguous definition is taking the first course. 11

This way of dealing with words like 'good' is characteristic of the 'therapeutic positivism' developed at Cambridge in the past few decades under the influence of Professor Wittgenstein. While this is unquestionably a useful philosophical technique, there are obvious limits to its applicability. For it is plain that in some cases in which diverse objects are called by a common name there is a common characteristic on account of which the name is given to them all. We need some principle enabling us to decide when such a common characteristic exists and when it does not; and what principle we use for this purpose will depend upon our general philosophical position. Analyses of the sort just given cannot therefore replace philosophical inquiry, as 'therapeutic positivists' seem at times to think they can, but both aid it and depend upon it. If we have other reasons for regarding the distinction between the 'natural' and the 'moral' realms as an illusory one, then tricks of language may explain how the illusion has come about; but it may still, as a matter of fact, be real.

It remains true, however, that a naturalist can extricate himself from Professor Moore's trap if he is bold enough and tough enough. And in imagining that in his refutation of what he calls the 'naturalistic fallacy' he has refuted naturalism, Professor Moore has himself fallen into a fallacy not unlike it. For if Professor Moore's own non-naturalism is a significant belief, then it must be possible to formulate the naturalism which it contradicts in a significant way; and if naturalism itself, and not merely the inadvertent combination of naturalism with something inconsistent with it, is senseless, then the denial of it is trivial. A significant non-naturalism, in other words, must comprise more than mere freedom from the 'naturalistic fallacy'.

* This paper was first published as chapter 1 of "Logic And The Basis Of Ethics" (ISBN 0 19 824157 7) by Oxford University Press, 1949.

1. pp. 6-17.

2. J. S. Mill, System of Logic, 1. ii. 5. The importance of Mill's distinction in the interpretation of Professor Moore's account of the naturalistic fallacy is rightly emphasized in Dr. D. Daiches Raphael's The Moral Sense, pp 111-14; though on p. 113 Dr. Raphael attributes to modern mathematicians a confusion in regard to this point, of which I do not think they are really guilty.

3. System of Logic, 1. vi.

4. The Philosophy of G. E. Moore, p. 582.

5. Sermons on Human Nature, Preface, par. 39.

6. Topics, 105b21-9.

7. This point is elaborated in an article on 'The Naturalistic Fallacy', byW. K. Frankena, in Mind, 1939, pp. 472 ff.

8. FormalLogic, pp. 119-20.

9. Ethical and Political Thinking, PP. 33-4.

10. Systemof Logic, 1. viii. 7; see also IV. iv. 5, V. 2.

11. For an answer to Professor Moore along these general lines see adialogue by E. and M. Clark entitled 'What is Goodness?' in theAustralasian Journal of Psychology and Philosophy, 1941.