22 The Meaning of Good (1942)

Discussing in Principia Ethica (ch. I, § 5) the question how “good” is to be defined, G. E. Moore considers it “impossible that, till the answer to this question be known, anyone should know what is evidence for any ethical judgment whatsoever… the main object of Ethics, as a systematic science, is to give correct reasons for thinking that this or that is good, and unless this question be answered, such reasons cannot be given” (italics in text; so in all subsequent quotations). It may be questioned, indeed, whether the giving of reasons depends on definitions; but at least we may argue that one of the main obstacles to the development of ethical science has been the difficulty of getting formal proof of ethical propositions — and perhaps we should add that the difficulty has been increased by acceptance of the doctrine that Moore goes on to uphold, that good is indefinable.

A question that confronts the ethical inquirer at the outset is whether there is a single sense of “good” or, allowing that there are several, whether one can be selected as fundamental to ethics — whether, in fact, there is such a subject as ethics. No one can deny the existence of “moralities” or codes, bodies of rules “recognised” (and backed, to varying extents, by “sanctions”) in societies or groups. But the use of such expressions as “legal ethics” and “medical ethics” gives us no ground for believing in ethics in general; the description of adherence to a given code as “good” and breach of it as “bad” is quite compatible with the denial of any absolute goodness or badness. In such cases, it may be argued, the group is concerned to maintain certain types of activity and takes as good whatever is favourable to them and as bad whatever is unfavourable (or, it might be said more broadly, the group takes as good and bad, respectively, what it supports and what it opposes); but such attitudes vary from group to group, and lend no colour to the treatment of things as good or bad in themselves.

Now this position of “ethical relativism” is quite widely accepted. According to it, we may say, “ethical” statements are incomplete; they signify relations one term of which has not been stated. Such abbreviations are common features of discourse and are obviously convenient. Equally obviously, they can lead to a great deal of confusion when the term “understood” by one party to a discussion is not the term “understood” by the other — and it is not only in the “ethical” field that such confusions occur. In this way people may be led to think that they are in agreement when they are not or to think that they are in disagreement when they

  ― 249 ―
are not. Thus (as Moore particularly stresses in his other book, Ethics) if the unstated term in any person's moral judgments is his own feelings, if his attaching a moral predicate to a certain subject means that he has certain feelings in relation to that subject, it would appear that two persons could never pass contradictory moral judgments. One might say “X is right” (or “Y is good”), and the other might say “X is not right” (or “Y is not good”), but they would be dealing with different issues. Each person would have his own moral terminology, and the apparent contradiction would imply no real contradiction in any one terminology of the common belief (shared by Moore) that the same action cannot be both right and wrong or the same thing both good and bad. It will still be possible, moreover, to translate any of the assertions into a common terminology; and if you say “X is the object (or occasion) of certain feelings in me”, there will be no logical objection, whatever objection there might be on the score of politeness, to my contradicting you, to my believing that X is not your sort of right or your sort of good.

But if there can be definite disagreement once the relation in which X is supposed to stand (“pleasing you”, let us say) has been clearly stated, this does not explain, it may be said, how confusion can ever arise, how I could ever imagine that “X is not right”, in my terminology, contradicts “X is right”, in yours. And, indeed, one may doubt whether confusion would ever arise if the omitted term were merely the feelings of the speaker. But the fact that people really think they are contradicting one another when they say “X is right” and “X is not right”, certainly gives us no ground for attributing rightness to X in itself, for denying that a relation to something unspecified is involved. The point can be illustrated from a set of cases, viz., gustatory differences, in which the primary reference is to the state or attitude of the speaker. A says “Oysters are nice”; B replies “No, they are nasty”. Here B is not denying that A likes oysters, but he takes A to have asserted more than that, to have implied, at least, that oysters are a proper object of liking — and this B denies. He is contending, in other words, that A is wrongly constituted, “abnormal”; he is appealing to a set of normal likings which A's particular taste contravenes. It is possible, of course, that A was merely expressing his own taste (though in that case he might have been expected to say “I like oysters”); if so, B was not really contradicting him. But it is also possible that A and B were expressing opposite views of the relation of a given taste to the “normal”.

We can now see how there can be confusion of issues, even though no question of what X is in itself (its qualities as contrasted with its relations) is involved — from the fact, namely, that many different terms may be “understood” in such judgments. And not only is there this variety of reference, but a particular reference, what it is that sets the standard we are upholding in a given case, may be far from definite. In general, as already suggested, the reference is to some institution or “movement”; the question is what forms of activity support or are supported by it. But social movements are not cut off from one another, and even the most

  ― 250 ―
well-marked movement has not a thoroughly worked out code but (like the most completely “codified” legal system) has its contradictions and unsolved problems; its policy, moreover, can change, and likewise the individual can change his adherences. Thus the judgments of A or B can have considerable uncertainty of reference, and even comparative certainty of reference does not imply a corresponding certainty of judgment. But this very uncertainty will make men more inclined to use an expression like “right” in an unqualified way; however vaguely the matter presents itself to them, they feel the pressure of “codes” and the urgency of adherence, and they are impelled to communicate this urgency to others without being able to indicate any precise ground for it.

Also, perhaps, without being willing — there is a dialectical advantage in ignoring the multiplicity of movements (or sources of norms); when we tell a person that some action is simply “right”, we represent it as favouring what he recognises as well as what we recognise, and, even if he remains uncertain about this, we at least confuse his mind so long as he has made no formal search for the missing terms. A common way out of the difficulty is, of course, to set up “society” as the single, ultimate determinant of what is correct or proper in behaviour. But it is not very hard to see that this is an evasion, that unqualified “social necessity” is on the same footing as unqualified “right”, that it is a means of concealing the opposition of forms of social organisation, that the delineation of a general “movement of society”, and the derivation from it of “norms” of behaviour, have not been carried out. It is one thing to recognise, as an outstanding feature of political life, the establishment of working arrangements among divergent movements; it is another thing to allow a postulated social unity to dictate to the social facts, to argue, as Idealists do, that there must be a unity “behind” the differences. To adopt this position is, theoretically, to abandon the problem of finding the “justification” of any specific form of action and, practically, to pass off some particular movement as “the whole” and so protect it from criticism.

These considerations apply to Moore's discussion (Ethics, chs. III and IV) of “the objectivity of moral judgments”. With reference to the view that each man's judgment of rightness is a judgment about his own feelings towards the given action (so that the judgments of two different men would never contradict one another), Moore maintains (p. 102) that, if we look at the question fairly, we must admit that such contradiction sometimes takes place, “that both men may use the word ‘right’ to denote exactly the same predicate, and that the one may really be thinking that the action in question really has this predicate, while the other is thinking that it has not got it”. How we go about “looking at a question fairly” is not indicated; but, if we allow that the two men could be referring to the same movement, it is clear enough that they could differ as to its “requirements” — though it is another question whether the point of their difference should be referred to as a “predicate” of the action. It is also possible, however, that they are referring to different movements; and it is on

  ― 251 ―
account of his ignoring of the multiplicity of movements that Moore's treatment of the question of a social reference fails.

Considering first the view that judgments of rightness are really concerned with the feelings prevailing in the society to which the person judging belongs, and noting that this would make contradiction on such matters between members of different societies impossible, Moore admits that it would be possible, on this view, for members of the same society to differ about the rightness of an act. But, he argues (p. 111), there is a fatal “psychological” objection to the view. “For, whatever feeling or feelings we take as the ones about which he is supposed to be judging, it is quite certain that a man may think an action to be right, even when he does not think that the members of his society have in general the required feeling (or absence of feeling) towards it.” To admit this, however, is not to get rid of the social reference. A man may consider, e.g., that he is acting rightly in going on strike, although most members of the society in which he lives are opposed to his action, but he may still take its rightness to consist in the fact that it serves a particular “cause” or movement. And similarly, even if we accept Moore's objection (pp. 142, 143) to the theory, which he thinks is often assumed without being expressly stated, “that to call an action right or wrong is the same thing as to say that an absolute majority of all mankind have some particular feeling (or absence of feeling) towards actions of that kind” — the objection, namely, “that it is quite certain, as a matter of fact, that a man may have no doubt that an action is right, even when he does doubt whether an absolute majority of all mankind have a particular feeling (or absence of feeling) towards it, no matter what feeling we take” — we can still take the man's judgment of the act as involving reference to a movement, perhaps a movement which is not limited to any particular society or age but can be regarded as running through human history, but at least a movement which is not identical with that history as a whole and which may at all times be a minority movement.

To say that Moore has failed to prove that judgments of the rightness of actions are not assertions of the requirements of some way of living (or of the policy of those who live in that way) is not, of course, to say that they do assert such requirements. And even if the expression “right” were commonly employed in that sense, it might have another sense which did not involve varying references. But the question would still be whether it involved some reference, whether it signified a relation to an unstated term (which might be the same in every instance) or signified a quality, something which really belonged to some subject itself. The latter position will hardly be maintained; it is apparent that, in any usage, to say that an action is right is to say that it “is to be done”, and that this raises the question on what consideration it is to be done. Now, as we have seen, this consideration may be a particular movement, i.e., it is the movement that “calls for” the action — but this is really to say no more than that the action does (or would, if performed) further the movement. Or it may be said more generally that the rightness of the action is its furthering

  ― 252 ―
something — something that can be agreed upon by the people who use the word “right” — but that, if no such thing were agreed upon or assumed to be agreed upon, the word would never be used.

Here it may be objected that justice is not being done to the conception of what “is to be done” or, more exactly, what “ought” to be done, that the question can always be asked, in regard to my action in furthering anything, “Ought I to further that?” But the assertion that I ought to perform some action, or that the action is binding on me, requires supplementation; it has still to be shown on what consideration it is binding. The notion of that which is unconditionally binding on me falls with every other notion of “that whose nature it is to have a certain relation”, viz., in that it treats a relation as if it were a quality. And if it is argued that the consideration in question is goodness, then it is my duty to promote what is good, the answer is that the assertion that certain actions promote good still gives no meaning to the contention that it is “my duty” to perform them — no meaning other than that they do promote good.

It could still be held that good is the unstated but constant term in judgments of rightness — such at least as are relevant to the science of ethics. And Moore follows this line in maintaining that the right action (the one which, in his view, it is my duty to perform) under given circumstances is that which, of all the actions “possible” to me in the circumstances, would have the best total consequences. But, apart from the peculiar difficulties of this formulation (the conceptions of “possible” actions and of a scale of goodness, the fact that the impossibility of calculating “total” consequences leaves us, as in the previously mentioned cases of unqualified right and unqualified social necessity, with the choice between uncertainty and dogmatism), it may be remarked in general (a) that it has not been shown that the promotion of goodness is not the furthering of a social movement, (b) that good might be treated as simply one example of an unstated term (something which can be furthered), the stating of which would render definite an issue of “right and wrong”, (c) that good also is commonly treated as a term of varying reference.

In this connection it was said earlier that people might take as good what is favourable to certain types of activity, without its being implied that anything is good in itself. But if so, it may be asked, what do they take it as? Is it merely meant that they take what is favourable as favourable? The point would be that they use “good” (like “right”) as an abbreviation, the term of reference (what is favoured) being “understood”. Now there is no doubt that the expression has relational uses; and all that Moore, in particular, maintains is (p. 162) that “to call a thing ‘good’ does not always mean merely that some mental attitude is taken up towards it” — or, he would argue similarly, that it contributes to some social movement. But his account of an alternative usage is unfortunate for his case. He appeals to judgments of “intrinsic value”, to the judgment concerning a particular state of things “that it would be worth while — would be ‘a good thing’ — that that state of things should exist, even if nothing else were to exist besides, either at the same time or

  ― 253 ―
afterwards. We do not, of course, so constantly make judgments of this kind, as we do some other judgments about the goodness of things. But we certainly can make them, and it seems quite clear that we mean something by them. We can consider with regard to any particular state of things whether it would be worth while that it should exist, even if there were absolutely nothing else in the Universe besides; whether, for instance, it would have been worth while that the Universe, as it has existed up till now, should have existed, even if absolutely nothing were to follow, but its existence were to be cut short at the present moment: we can consider whether the existence of such a Universe would have been better than nothing, or whether it would have been just as good that nothing at all should ever have existed” (pp. 162, 163).

The contention that it is not the same thing to say that somebody is pleased at the idea of X or desires X for its own sake and to say that it would be “worth while” that X alone should exist, is one that few would contest. But the usage “it is good that” cuts across the conception of “intrinsic” goodness, of a thing's being good in itself. We say “This is red” or “This is spherical”; but it would be nonsense to say “It is red that this, and this alone, should exist” or “It is spherical that this, and this alone, should exist”. The assertions “This is good” and “It is good that this should exist”, are intelligible (and compatible) only if we take a relational view of goodness — only if they have some such meaning as “This is demanded (or commanded)”, and “It is demanded (or commanded) that this should exist”. If, on the other hand, we regard good as a quality, as characteristic of a thing without further reference, then “It is good that this should exist” has no meaning. And, in this connection, the question what we should think of the thing if it made up the “Universe” is quite pointless; the sole issue is whether it has this particular quality, no matter what other things exist — though certainly it would not have any quality, it would not exist at all, except in an environment.

In fact, Moore's “large” way of treating the matter enables him to dodge the question “quality or relation?”; it enables him vaguely to suggest relations, without committing himself to the view that they are inherent in goodness. Thus, in spite of his rejection of the identification of being good with being desired, he speaks (p. 237) of a thing's goodness as a “reason for preferring” it, and, as we shall see, a similar confusion runs through Principia Ethica. And, on the face of it, to call a thing “worth while” is to say that it has some relation. But, if so, this may be, for all Moore shows, the relation of being the object of a certain mental attitude. It may be true, he is prepared to concede (Ethics, p. 166), “that there really is some very special feeling of such a nature that any man who knows that he himself or anybody else really feels it towards any state of things cannot doubt that the state of things in question is intrinsically good”. And if any one should maintain that when we call a thing intrinsically good we mean merely that this special feeling is felt towards it, “the only obvious argument” Moore can find against that view “is that it is surely plain that, even if the special feeling in question

  ― 254 ―
had not been felt by any one towards the given state of things, yet the state of things would have been intrinsically good”. Now, if “intrinsically” means “qualitatively”, the obvious and conclusive argument is that a quality is not a relation; but if it can have some relational meaning instead (or as well), then Moore's point is not “plain” at all. And his subsequent, less obvious argument from considerations of “duty” falls with that conception. There are, then, still the two possibilities (a) that good is a quality, (b) that “X is good” asserts a relation, one term of which is “understood” — leaving it an open question whether the missing term is “feelings” or something else.

It is important to observe that a person who denies that there is any quality, goodness, and asserts that “good” signifies a relation to something, e.g. to feelings, even to the feelings of the speaker, is not thereby denying the “objectivity” of moral judgments. (Presumably the class of judgments containing abbreviations like good and right could still be distinguished as moral judgments.) The question whether X causes pleasure in me is (assuming that an unambiguous meaning can be given to “pleasure”) just as objective, just as much a question of fact, as the question whether X is red. Briefly, relations are as objective as qualities. Thus the economist can make objective statements about “goods” (commodities) while recognising that, to be a commodity, a thing has to be demanded by a person, that nothing is a commodity in itself. It has, no doubt, been argued that certain things are of necessity commodities — the “necessaries of life”, the things that everybody demands or, at least, requires. But this further point does not prevent our using the conception of objects demanded, and recognising that the things a person demands, though other people do not, are “goods” in exactly the same sense as are objects of general demand. Again, many economists speak of “welfare”; they treat certain things as the proper objects of demand. But this can only mean either that these things, whether they are demanded or not, have some quality, or that the economists (and others) demand that people should demand these things. And these questions — the question of the demand for “goods” in some qualitative sense, and the question of demands for demands — can be fruitfully considered only if the initial, relational usage is quite clearly maintained, only if we are able to speak, with no presumption as to its own character or further conditions, of that which someone or other demands.

It is here, however, that not only economists but ethical theorists fall into the most serious confusion. The outstanding characteristic of traditional ethics is the passage from the demanded to the ultimately demanded (that for the sake of which anything else is demanded) and the identification of the supposed “final end” with the good. The position can be illustrated from Moore's criticism, with special reference to Mill's Utilitarianism, of the Hedonistic principle “that pleasure is the only thing at which we ought to aim, the only thing that is good as an end and for its own sake” (Principia Ethica, § 39). This is the view commonly described as “ethical hedonism” and distinguished from the view,

  ― 255 ―
“psychological hedonism”, that pleasure is the only thing at which we do aim, the only thing that is an end; and one main point in Moore's argument is that Mill confuses these two views, that he derives “ethical” conclusions from “psychological” premises. But the position cannot be cleared up until the notions of “ought” and “good as end” are abandoned; their retention renders ineffective a considerable part of Moore's criticism of hedonistic doctrine.

Mill, says Moore, “has already told us (p. 6)note that ‘Questions of ultimate ends are not amenable to direct proof. Whatever can be proved to be good, must be so by being shown to be a means to something admitted to be good without proof’ [Moore's italics]. With this I perfectly agree: indeed the chief object of my first chapter was to show that this is so. Anything which is good as an end must be admitted to be good without proof.” Now we should all agree that proof is from premises, i.e., from something which, as far as the given argument is concerned, is not proved — and, though a proof of it might be annexed, the chain of argument would take its departure from what we had not proved but simply believed. (This is not the place to contest the view that proof starts from the “self-evident”; but even those who believe this are not denying that the argument starts from something taken to be true and not derived from other propositions.) Thus, if we are proving something to be good, we must have a premise that asserts that something is good; we could argue that X is good because X is Y and Y is good, or because X is Y and Y is Z and Z is good, etc. — in any case, the assertion that anything of a certain sort is good is either a premise or derivable from a premise; i.e., “admitting something to be good without proof” is essential to the argument.

But we cannot agree with Mill and Moore in their further characterisation of the argument. What is required, in addition to the admission in question, is the assertion (or a number of assertions which imply) that X is of a given sort; to say that X is a means to something of that sort is entirely irrelevant to the proof of the required conclusion. We might say, indeed, that “Y is good” and “X is a means to Y” prove something, viz., that X is a means to something good; but how can this be taken to prove that X is good, let alone to be typical of the method by which anything can be proved to be good? If X were a means to Y and Y were red, we should not take this as giving any sort of proof that X was red. It is only on a relational view of goodness that the question of means to ends comes into the matter, only if, e.g., being good signifies being demanded that this “character” could be regarded as transferable from a given thing to what brings that thing about. Even so, of course, the proof would not be adequate if our information were merely that X brings about Y; given that Y is demanded, X would have to be taken as not only a means but the only means to Y if the passing on of the character of being demanded from Y to X was to be justified. Still, under these

  ― 256 ―
conditions, consideration of means would be relevant, though proof without reference to means (syllogistic proof from the premises that any Y is demanded and X is Y) would also be possible.

It may be contended that the above argument passes over the distinction, on which Moore insists, between being an end simply and being good as an end, or the distinction, which (§ 40) he accuses Mill of concealing in his use of the expression “desirable”, between that which can be desired and that which ought to be desired or which it is good to desire. But the question is what Moore's “ethical” expressions can possibly mean — what can be meant by saying that X is good as an end or is that which it is good to desire. Is it meant, in the former case, that X is good and is chosen? If so, consideration of its being chosen is quite distinct from consideration of its being good — and this whether its goodness is a quality or another relation (for the presumption is that it is not the same relation, that being good as an end is different from merely being an end). Again, if it is meant that X merely may be chosen and that this would be a good choice, then to say that choosing X would be choosing a good thing adds nothing to the characterisation of X as good and gives no justification for the carrying over of “goodness” from the chosen to the choosing. And this would be so even if it were always by being chosen that good things come about.

Moore, then, can give no meaning to “good as an end”, and the same applies to his notion of the “desirable”. This can scarcely mean that which is good when it is desired, for, as before, whether its goodness is a quality or a relation other than that of being desired, it would presumably be good even when it was not desired — or if we held (curious as this might appear) that the sort of thing in question was always desired throughout its existence or as long as it remained good, still its being good would be different from its being desired and speaking of its goodness separately would obviate confusion. If, on the other hand, the “desirable” means that the desiring of which is good, then we are attributing goodness not to certain things chosen, but to relations of choosing — or we might attribute it to that which chooses — but in neither case would any connection have been indicated between goodness and the character of the things chosen, so that they could be called “desirable”. In other words, unless goodness simply means choosing certain things (which then would not themselves be called “good”), the fact that what is good chooses, or is the choice of, something will be beside the point.

There would be no difficulty in taking a consistent view, of course, if the matter were considered relationally — as a question, say, of support or advocacy. Thus the “desirable” might be taken as that which is advocated by certain people (or is in harmony with a certain movement) or as that the desiring of which is so advocated; and there would be a ready passage from either of these meanings to the other. Something like this, I would suggest, is the ordinary usage; the assertion that X is desirable always invites the question — desirable for the sake of what? But Moore, in upholding the absolutely desirable or the notion of “good as an end”,

  ― 257 ―
is running together relational and non-relational notions. He is falling into the sort of confusion he himself refers to as “the naturalistic fallacy”, exemplified in what he takes as Mill's dual use of “desirable” and again in “evolutionary ethics” which at once identifies and distinguishes the good and the product of evolution, maintaining that “good” means neither more nor less than what evolution produces and yet that the product of evolution is good. Certainly, Moore holds that some ends are not good; but he cannot distinguish good and bad ends except by their qualities — in which case their being ends has nothing to do with the matter — and thus, when he takes “finality” as essential to goodness, he can be accused of identifying, as well as distinguishing, being good and being an end. Incidentally, whatever may be said of his later argument to show that we desire other things than pleasure, he cannot be said, in the passage above referred to, to have convicted Mill of any blunder based on a misunderstanding of “desirable”; for, if the only thing that can be an end is pleasure, it is equally the only thing that can be a good end.

In fine, Moore does treat good as the ultimate object of demand, as that for the sake of which anything else is demanded; for if he should change this to that for the sake of which anything else is properly demanded, in other words, that the bringing about or production of which is “proper production”, the only meaning he could give to this expression would be production of good — and thus he would be saying that good is that the production of which is production of good, which is quite uninformative and, in particular, indicates no connection between production (or the employment of means to ends) and good. But he also wants to treat good as qualitative. And it is this predicament that is at the basis of his doctrine (developed in the first chapter of Principia Ethica) of the “indefinability” of good.

In the first place, we may say, it has to be taken as indefinable, there is nothing we can set out as its character, because there is no ultimate object of demand. But, secondly, there can be nothing common and peculiar to it, because that could be taken as a reason for demanding it (something common but not peculiar to good could not be such a reason, for it would equally be a reason for demanding what is not good), and, being that for the sake of which some person demanded “good”, this “reason”, or something to which it in turn was subordinate, would really be the person's standard or ultimate end; i.e., it, and not the thing first called so, would really be the person's “good”. To put the matter otherwise, Moore has implicitly defined good as that which is ultimately demanded (or desired), but he cannot make this explicit without exposing himself to criticism on the score of the irreducible multiplicity of demands — and, as already said, he has a sense of good as a quality. In fact, in his notion of good he amalgamates quality and relation (this being the procedure which, in my view, is properly described as “relativism”) and so cannot give either a relational or a qualitative account of it — hence its “indefinability”.

This view is borne out by his treatment of good as peculiarly a

  ― 258 ―
predicate. In the passage first quoted (§ 5) he considers the assertion that this or that is good as the sort of thing that ethics would seek to establish. One would think that ethics, regarded as specially concerned with good, would try to discover predicates as well as subjects of it, how it operates as well as where it occurs. But it cannot stand as a distinct subject if it is really conceived relationally, if in its use “as a predicate” it is an abbreviation the expansion of which would introduce a hitherto unstated term, if it is not an operating thing at all but a way in which one thing stands to another. Not being prepared to undertake (or even admit the possibility of) this expansion, Moore has to fall back on “intuition”, the arbitrary attribution of goodness to certain subjects, with no possibility of discussion. But such attribution, if it cannot be expanded, can never convey an absolute obligation on persons to endeavour to bring such subjects into existence; on the other hand, any actual expansion would still not impose an absolute obligation, would not show that anything is obligatory in itself but would merely present something that some persons would be inclined, and others disinclined, to promote. It is more convenient, if there are objects one wishes above all to promote, dogmatically to call them good and let it be “understood” that a certain obligatoriness attaches to them on that account.

Moore supports his position (§ 9) by distinguishing between good and the good (that which is good), the latter of which he believes to be definable though the former is not. “I suppose it may be granted that ‘good’ is an adjective. Well ‘the good’, ‘that which is good’, must therefore be the substantive to which the adjective ‘good’ will apply: it must be the whole of that to which the adjective will apply, and the adjective must always truly apply to it. But if it is that to which the adjective will apply, it must be something different from that adjective itself; and the whole of that something different, whatever it is, will be our definition of the good.” Now, Moore says, “many people appear to think that, if we say ‘Pleasure and intelligence are good’, or [and?] if we say ‘Only pleasure and intelligence are good’, we are defining ‘good’”. But this is not his own view; whatever definition of the substantive may be found, the adjective, he considers, remains undefined. “It may be true that all things which are good are also something else…And it is a fact, that Ethics aims at discovering what are those other properties belonging to all things which are good. But 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” (§ 10). And it is this view that Moore proposes to call “the naturalistic fallacy”.

In dealing with these contentions, we need not dwell on the fact that they would make any term indefinable — so that Moore must be simply wrong about the meaning of definition. The essential point is that the adjective does not “apply” to the substantive (“that which is good is good” is not a proposition), is not predicable of it as one term is predicable of another, but adjective and substantive are the very same term, and when we

  ― 259 ―
define the substantive we are defining the adjective. The adjectival form is ordinarily an indication that the term is being used as a predicate, and, when we use it as a subject, we find it more natural to employ the substantive form; but such usages involve no logical distinction. Thus “Some animals are men” is the very same assertion as “Some animals are human”, and though, in converting the second form, we commonly say “Some human beings are animals”, we could say, without alteration of content, “Some human are animals”, or “Some human are animal”. Again, we could derive the syllogistic conclusion “Socrates is mortal” from the premises “Socrates is human” and “All men are mortal”, or “Socrates is a man” and “All human are mortal”, as easily as from those we ordinarily employ; in fact, it is a question of different formulations of the same premises, not of different premises. The whole point of syllogistic arguments is that the same term may function as a predicate (“adjectivally”) and as a subject (“substantively”), that it characterises other terms and is characterised by other terms.

Equally, a term is defined by other terms. If “rational animals” were a correct definition of “men”, that would mean that we could substitute “rational animals” for “men” in every true proposition in which “men” occurs as subject or as predicate and the resulting propositions would be true. But they would be different propositions from those we started with. In defining A as B, or, more strictly, as BC, we are not proposing never to use the term A again but always to use the term BC instead; we are setting forth a complex relationship among certain sorts of things, A, B and C. Thus there may be some force in Moore's opposition to those who say that good means “nothing but” something or other, e.g., nothing but the object of desire. But the main question in this particular case would be whether good is to be taken qualitatively or relationally. If in the former way, it cannot be described as nothing but the possessor of a relation to something else (it would have to have some other description, or there would be nothing to be so related), whether the proposed “total” description is called a definition or not. If in the latter way, the correct expansion of an abbreviation can certainly not be called a definition, but at least it makes for understanding of the matters talked about.

It is true, as Moore goes on to say (§ 11), that such an explanation of a usage cannot direct us to act in one way rather than another. But it is no less true that the recognition of a quality cannot so direct us. Thus, when he says of people who explain how the word “good” is used, that “in so far as they tell us how we ought to act, their teaching is truly ethical, as they mean it to be”, but that it is absurd of them to offer the usage as a reason for acting as they recommend, he is showing that he himself has no clear conception of ethics or of goodness. The point is connected with his assertion that good is not a “natural” object, and with his consequent description of theories which identify good with some natural object as committing the naturalistic fallacy (though the same “fallacy” is committed, he considers, if good is treated as “absolutely and entirely the same” as anything whatever). For we can take as natural

  ― 260 ―
having a certain quality, and we can take as natural being in a certain relation, but we cannot take as natural that in which being in a certain relation and being of a certain quality are merged. But this is not because it is “non-natural” (perhaps, supernatural); it is because it is nothing at all. The description of good as non-natural, like the description of it as indefinable, is a way of avoiding the clearing up of the ambiguities in Moore's conception of it.

A parallel case is found in the discussion between Socrates and Thrasymachus in Republic, Book I. Thrasymachus demands of Socrates after he has demolished the definition of justice which Polemarchus has tried to uphold, that he should give his own definition of it; “and don't dare to tell me that it is the obligatory, or the expedient, or the profitable, or the lucrative, or the advantageous, but make your answer precise and accurate, for I will not have any rubbish of that kind from you” (Lindsay's translation). The point would seem to be that, if we give an “ethical” definition of any ethical term, we have what is really no more than an identity (“the worth while is worth while”, say), whereas if we give a “non-ethical” definition of it (if we define it as a “natural object”), we seem to be destroying ethics; hence moralists like Socrates (and Moore) are unwilling to define such terms.

But what we are destroying, if we point definitely to some existing thing as the subject of ethics, if we take good as a merely descriptive term, is its mandatory character. “Goods are just those things” does not direct us to act in any way but it permits us to investigate their ways of working, and so ethical science can go on. But “the good is the mandatory” tells us nothing at all; it invites the rejoinder of Thrasymachus, implied in his definition of justice as “what is advantageous to the stronger” (or “the interest of the stronger”: Jowett), that such a relation among persons as the command of a ruler to his subjects is all that we can positively mean by an imperative. This is something we can observe, it is a type of situation in interaction with which we can develop a theory of it, whatever qualities that which is imposed by command may have. Similarly, we could study things qualitatively describable as good, whatever relations they might have. But an imperative quality is something we could never observe or study; it has to be left in obscurity because it is basically ambiguous — though we may still consider that the postulation of such a “non-natural" entity covers the pushing of certain quite natural objects.

The remainder of Moore's first chapter raises a few further points, which may be briefly dealt with. His distinction (introduced in § 15) between the two types of ethical judgments, those which take something to be “good in itself” and those which take something to be “good as a means”, suggests that carrying over of a character from end to means which has already been criticised. Here it may be said that judgments of the former type, which take a “unique property” as attaching to something, would also have to take it as attaching in a unique way if it is a non-natural property of a natural object. In fact, it is impossible to see how

  ― 261 ―
Moore's “intuitions” would work, how an unobservable predicate could be found to attach to one observable subject rather than another, or how we could ever know that it was the same thing that had some natural property and had the non-natural property. But if we are able to take a thing and its ethical and non-ethical properties in the same view, there is no question of “intuition” (except in the sense of observation), and good is found in things in the same way as red or yellow may be found — the alternative being that it does not “attach” to them.

The point is reinforced by consideration of judgments of the second type, “that the thing in question is a cause or necessary condition for the existence of other things to which this unique property does attach”. For a thing is not a means to good unless it brings about not merely things which have the property but the property itself, unless good comes into existence through the operation of the thing in question. (It may, of course, already exist elsewhere, but the thing has caused it to exist in a given place.) Otherwise, the existence (or subsistence) of good is quite indifferent to that of the things to which it is said to attach, and in that case there is no sense in saying that it attaches to them. But if a natural thing, in its natural operation, can bring about the result, good, there is no sense in saying that this is not a natural result, no way of distinguishing its natural results from its non-natural results; in each case it produces something somewhere. And it should be noticed that what is in question here is not the theory of universals; it may, indeed, have some relevance to the case, but Moore admits that there are natural properties (e.g., yellow), and he is not entitled to treat some properties in terms of one theory of universals and others in terms of another, to make out that there are peculiarly “subsistent” properties. The conclusion to which he is forced, then, is that good is natural.

His doctrine of organic wholes (introduced in § 18) is of importance only as a confirmation of his relativism. “The value of a whole must not be assumed to be the same as the sum of the values of its parts”; it can be greater or less. Moore is speaking here of “intrinsic value”, of the “unique property” also referred to in the phrase “good in itself”. Thus he is recognising degrees of goodness — as he also does in taking the right act to be that which has the best total consequences. Now, such recognition presents no difficulties if good is taken relationally, if, in particular, the good is identified with the wanted. For clearly there can be degrees of “wantedness”; we can want one thing more than we want another. Thus we could speak of a thing as “better” than something else or as “the best” of a number of things, using the same sort of abbreviation as we do in calling it “good”, i.e., without specifying who it is that does the wanting — though, while we can speak of a thing as a “good” (commodity) if it is wanted by any one at all, specification is more urgent in the case of degrees since different people have different preferences or, as we may put it, since there are different rates of exchange on different markets. The very use of the term “value” suggests that the question is an economic one —

  ― 262 ―
remembering always that there is no place, in strict economic science, for the conception of “intrinsic” value.

But if good is taken qualitatively, there can be no question of degrees; if two things have a certain quality, one of them cannot have it “more”, in a higher degree, than the other has it. This is a point on which there is a great deal of confusion, and in common speech we constantly make such assertions as “Gum leaves are not so green as grass”. Now one possible meaning of this assertion is that, while grass is green, gum-leaves are not green but have certain resemblances in colour to what is green — more generally, that various colours can be said more or less nearly to “approach” a given colour. But, whatever may be meant by “approaching”, this will not entitle us to speak of degrees of possession of a given colour. Alternatively, it may be meant that each is of some shade of green but one shade is more “typically” green than the other. But in whatever way the type may be determined, if both shades are really shades of green, one is not more possessed of greenness than the other. In other words, if it is in greenness that they agree, it is not in greenness that they differ; and any contrary view would simply involve the use of “greenness” in different senses. Again, we often speak of a thing as “partly green”, but this would seem to mean that parts of it are (unambiguously) green and other parts are (unambiguously) not green — and in any case the thing itself is not green; it has a green part but it has not a part of greenness. More could be said on this question, but it will always be found, I would argue, that the notion of degrees of any quality is a confused one — that the question of a thing's possession of the quality is confused with the question of its possession of other qualities (or of certain relations) as well.

The persistence of the belief in degrees of goodness, then, is due to the fact that it is not taken qualitatively, or not only so, that it is taken as jointly qualitative and relational. But that means that it is taken in an inconsistent fashion. This inconsistency is present in the very notion of “intrinsic value”, i.e., something which is such in itself that it requires to be valued (by some other being); for, even if it could be significantly said that the thing “calls for” valuation (this, incidentally, being regularly taken to mean that it calls for support, not merely for recognition of it as of a certain character), that would not be what the thing is in itself. Two consistent attitudes can be adopted. One is to deny that good is a quality, to take a purely relational view of it, e.g., that it is the demanded. In that case we can work out a positive theory of it, while recognising different degrees of “demandedness” (or relative strengths of demands). Equally we can have a positive theory of authority while recognising that some things exercise greater authority over us than others but denying that anything is authoritative in itself. The other is to take good simply as a quality, to recognise goods as things existing in certain places and going on in certain ways. On this view, though not on the other, there will be a distinct science of ethics, but it too will be a positive or natural science. This, however, will be no reason for confusing it with

  ― 263 ―
other natural sciences, e.g., the science of the operation of our demands.

It is true, no doubt, that those who recognise such a quality and such a science will have in the end to give some explanation of the confusions that constantly arise, of the fact that the expression “goods” which they use qualitatively is used relationally in economics, of the fact that “justice” has jointly or alternately a political and an ethical meaning — on the one hand, social procedure according to recognised rules and, on the other hand, a good way of proceeding. But what they will have to do in the first place is to explain their own usage — and to point out what, in accordance with it, they take as the most serious errors of other thinkers.

Thus they will accuse Socrates, in the Republic, of confusing between political and ethical “justice”, between the system of rules laid down in a given society — this being always a compromise, a working arrangement among divergent movements, and varying from time to time as well as from society to society — and the modes of operation of goods, limited always by a hostile environment but incapable of existing except in the struggle with difficulties. They will accuse him of obscuring the relation between the two (obscuring, e.g., the concrete question how far the operation of goods is necessary for any sort of social cohesion) by erecting an imaginary system in which political and ethical justice are one, in which goodness dominates and the “rules” of its operation are in no way contravened. And they will observe that all this is facilitated by the taking up of a preceptual attitude, the attribution of a mandatory character to goodness, that the questions of how societies work and how goods work are amalgamated in the notion of how societies ought to work. They will say, then, that progress in ethics and allied studies depends on the rejection of “mandatoriness”, which can only arbitrarily be attached to any line of action. Not that they will accuse Socrates, or again Moore, of mere advocacy — but until, in the work of such thinkers, the recognition of goodness as a quality of certain human activities is disentangled from the advocacy, discussion will not be materially advanced and ethics will remain a free field for any one who thinks he can tell people what to do.

It is in fact a standing obstacle to the acceptance of ethics as a positive science that people simply will not be persuaded that, when we say “X is good”, we are not urging them to promote X or to exhibit activities of the character X — that there is no more advocacy in our statement than in the statement “X is red”. When they demand proof of such a statement, what they really want us to give them is some reason for supporting X — a reason which, of course, could be given only by connecting X with something (no matter what its nature might be) which they actually support, and thus a reason having nothing to do with the qualitative characterisation of things. Such people, on the other hand, as recognise that it is a quality that is in question will see that formal proof of its belonging to X depends on its being known to belong to something else which may be predicated of X. And proof of this kind will be of no use to those who have still to be persuaded that there is “such a thing” as the quality, good. It has already been observed that there is no question

  ― 264 ―
of a mere quality, that good, on the view under consideration, would have to be taken as a “sort of thing” (or term) which could be either subject or predicate. The question is, then, how any one who does not yet admit it, can be brought to see that there is the sort of thing that another person calls “good”.

First of all, there is the possibility that he is acquainted with the sort of thing (or, putting it extensively, with the class of things) in question but calls it something else, so that all he is brought to do is to alter his usage. But he would not do that if he thought that calling it “good” was an idiosyncrasy of the person trying to persuade him, but only if he considered that it came fairly close to a common usage or perhaps more particularly to a learned usage, that many views of good were concerned with something closely connected with the given sort of thing and, indeed, might be represented as not fully successful attempts to clarify the conception of it. And this leads on to the second possibility, that it is something of which he himself has not a clear conception, something he “implicitly” recognises but has only partially disentangled from other sorts of things (or qualities). Persuading him in such a case would largely consist of showing him that, by drawing lines of distinction where it was proposed that he should do so, he could clear up many uncertainties, could give coherence to his own and other people's views and usages, could, as before, represent them as attempts to grapple with what the suggested position clearly sets forth. It may be said, in fact, that this is a regular feature of criticism and the development of knowledge, and that it must be involved in the establishment of a positive theory of good — that no one who has not some acquaintance with the field (whatever his confusions may be) is going to understand, let alone accept, a view of it. The essential point is that there are degrees of clarity and confusion.

It must not be thought, however, that it is solely (perhaps even mainly) by formal argument that confusion is dissipated. Here we may consider how in general we come to distinguish a quality, recognise a sort of thing, “use a term”. What is in question here is not the use of words, but can be illustrated by reference to the use of words; the learning of a language exemplifies the characteristics of learning in general. For, while the use of a word may be described as arbitrary in the sense that what we call “green”, for example, could conceivably have been (is, in fact, in a language other than English) referred to by some other word, we are not using it as a word unless we refer by means of it to a particular sort of thing. And this implies that we are directly acquainted with that sort of thing or with things of that sort, i.e., with situations. Further, we have to be acquainted with the word as a noise of a certain sort; and the “reference” of this to the other sort of thing is a further situation with which we become acquainted.

As has been pointed out by J. B. Watson, the principle of the learning of a language is that “the word brings the thing”, this being a particular example of the way in which one sort of thing signifies another (as black clouds mean rain or fire means heat). We learn, that is to say, by having

  ― 265 ―
desires and expectations and having them satisfied. Of course, we are often disappointed; but it is in trying to overcome disappointments that we learn to make finer discriminations. So we may make mistakes in the use of language, i.e., in the usages of the persons from whom we learn it; a situation being complex, we may “mean” one feature of it when they “mean” another. And there are other reasons why the word does not always bring the thing. But if it had not, in the first instance, led to the presentation of something we recognised and if it did not later elicit certain reactions (forms of communication) from other people, we should never learn the language. The essential point is that it is in our operations (involving, for the most part, co-operation, and marked by fulfilled and unfulfilled expectations) upon the situations in which we find ourselves, that we learn and, incidentally, correct our errors (as far as we do correct them), i.e., our failures to get the expected response from things or from persons.

This means that we can acquire a knowledge of good, in particular, only as something upon which we can act and which can act on us — only as something “natural”, present in our environment. Unless good is one description of certain things, helping us to recognise them just as their being green might do, we can have and communicate no knowledge of it — assuming, that is, that it is not something relational; but, if it were, our knowledge of it would still depend on our encountering such relations in the situations that confronted us. The treatment of it as “non-natural”, not open to ordinary observation, is due, as we have seen, to the running together of incompatible meanings; and on that view we could never have become acquainted with it. It will be urged, in this connection, that the expressions “good” and “bad” come to us as hortatory terms, that the initial meaning of “X is good” is “I want you to do (or to promote) X”, while that of “Y is bad” is “I want you not to do (or to avoid) Y”. These, of course, would be perfectly natural meanings; we regularly encounter facts of these types in our mature experience. But it may be questioned whether the notion of exhortation can be clearly grasped in early experience, and consideration of what is then conveyed may lead to a modification of the relational view.

The position adopted by members of the Freudian school is of importance here. The child, they consider, being urged to do what is “good” and to avoid what is “bad”, finds that the doing of the acts described as “bad” leads to the withdrawal from him of his parents' affection, while the doing of the acts described as “good” involves its retention or restoration. Thus “bad” brings or means for him loss of love, and “good” means love. The general effect of the parents' exhortations, then, is to attach the notion of “goodness” to something of which the child has direct experience — which is, indeed, the main form of his communication with others. It may be contended that the admonitions the child receives are of such a mixed character that he can draw no clear line between what is good and what is not. But, the answer is, unless he can draw some line, the admonitions will have no meaning for him and he will never be “morally

  ― 266 ―
trained”; and he has his own experience, of love and the loss of it, to draw upon. Thus, while the mixture of admonitions and, later, of social pressures (and this would vary in different cases) may be responsible for some abiding confusion on the subject of goodness, he has his own way of meeting the situation; according to the Freudians, he divides the parental figure into a “good” and a “bad” parent — the latter being the original of the “wicked step-parent” of the legends.

The “ethical” distinction he adopts can thus correspond very closely to a real difference of quality, and so can be the beginning of a positive theory. It is another question how far the theory will develop, whether the distinction will become more or less clear-cut as time goes on, whether similar differences will be observed in other cases — i.e., whether love will be taken as the only good (the view that the Freudians themselves tend to adopt) or merely as the first recognised or first strongly operative good. And here I would suggest that the distinction between thoroughgoing communication (co-operation) and repression is of considerable importance, that, while this may seem on the face of it to be a distinction between relations, there are distinctive qualities that go with them, that a quality, goodness, is involved in communication itself, whether it be the communication of love, the communication of knowledge, or any other. On this view, good is not merely something that we discover but is that by which we discover things — or, if inquiry is taken as one particular good, it at least communicates with other goods and they assist its operations (a view which I have maintained, in slightly different terms, in an earlier article: “Determinism and Ethics”, A.J.P.P., December, 1928).note

On this view, too, the consideration of social movements is of considerable importance for ethics and may assist the recognition of good as a quality. It may be recognised first that there are qualitative differences among ways of life, and secondly that a way of life is not something that we adopt, by a voluntary decision, but something that adopts us, takes us as a vehicle, kindles a certain “spirit” in us. Thus the scientific spirit, the spirit of inquiry, may be said to be kindled in us by the scientific movement, by a social phenomenon which no individual or set of individuals could have planned and which, in operating through an individual, never completely absorbs him but strengthens the communicating, as against the divisive, tendencies in him. The other ingredients in culture, like art and industry, are also generative of a communicating spirit and of institutions in which it may be expressed, and this, it may be said, is what prevents the taking of “culture” in a non-qualitative sense, as merely what is established among a group of human beings at any given time. So with “progress”; it cannot be taken simply as the approach to any desired object, for it implies a continuing tradition and the working further, without arriving at any end, along the line of that tradition, being thus opposed alike to stagnation and to mere innovation. And only certain activities can proceed in this way, as science, e.g., goes from solutions to further problems.

  ― 267 ―
Progress, then, is not the achievement of “greater good” but is a characteristic of goods; goods are “progressive” in that they continually grapple with new problems on the basis of their previous history. They are the continuing features of social life, the “causes” to which men can devote themselves. The position may be summed up in Croce's description of history as the story of liberty or, as we may put it, liberty as the subject of history. Liberty resides in “causes”, and they alone have a history because they alone continue as long as there is society. We may thus approach a definition of good. Goods, we may say, are those mental activities, or those social activities, which are “free” or enterprising, which exhibit the spirit of enterprise. It might be better to come down on the mental side. The main point is that ethics penetrates both the psychological and the sociological field, but is nevertheless a distinct and positive inquiry. And the recognition of a class of things to be inquired into is more immediately important than a formal definition of good. But definition has its own scientific importance, and it should be understood that there is nothing “indefinable” about good, whether we actually succeed in defining it or not.

That many people would be unconvinced, by the above outline of a view, of the existence of a natural quality, good, is obvious; but others may see that it is something with which they have long been in certain ways acquainted. It may be seen, too, that certain popular “moral” conceptions can be accounted for as approaches to what I have taken to be ethical facts — that “happiness”, e.g., may be understood not as the receiving of what we want, so that we want no more, but as the continuance of an activity, securing its material as it goes along; that “freedom” may be taken not in the metaphysical sense of release from causation but as a power of devoting oneself to what transcends oneself (a social movement or “cause”); that even “duty” may be considered as expressing the fact that individuals may fall away from movements and be painfully brought back. The vital point is the rejection of “good as end”, of the notion that goods come about by being wanted. This, the individualistic or “consumer's” view, is the main obstacle to the development of a positive science (i.e., a science) of ethics.

That there should be such an obstacle is not surprising. For the goodness of inquiry implies the goodness of ethical inquiry. And, in recognising that goods themselves operate in us and out of us, we have also to recognise that they encounter obstacles both in us and out of us. We cannot, in other words, make the world safe for goodness; it exists and develops in struggle with evils. But it is important also to observe, in rejecting the doctrine of conscious intention (or “planning” of good), that good can exist without consciousness of good, that a person can engage in good activities without being an ethical theorist.