In this last error, in the supposition that when I say You ought to do this I must mean You are commanded to do this, we have one of the reasons which has led to the supposition that the particular supersensible property by reference to which good must be defined is Will. And that ethical conclusion may be obtained by enquiring into the nature of a fundamentally real Will seems to be by far the commonest assumption of Metaphysical Ethics at the present day. But this assumption seems to owe its plausibility, not so much to the supposition that ought expresses a command, as to a far more fundamental error. This error consists in supposing that to ascribe certain predicates to a thing is the same thing as to say that that thing is the object of a certain kind of psychical state. It is supposed that to say that a thing is real or true is the same thing as to say that it is known in a certain way; and that the difference between the assertion that it is good and the assertion that it is real—between an ethical, therefore, and a metaphysical proposition—consists in the fact that whereas the latter asserts its relation to Cognition the former asserts its relation to Will. (§ 77 ¶ 1)

Now that this is an error has been already shewn in Chapter I. That the assertion This is good is not identical with the assertion This is willed, either by a supersensible will, or otherwise, nor with any other proposition, has been proved; nor can I add anything to that proof. But in face of this proof it may be anticipated that two lines of defence may be taken up. (1) It may be maintained that, nevertheless, they really are identical, and facts may be pointed out which seem to prove that identity. Or else (2) it may be said that an absolute identity is not maintained: that it is only meant to assert that there is some special connection between will and goodness, such as makes an enquiry into the real nature of the former an essential step in the proof of ethical conclusions. In order to meet these two possible objections, I propose first to shew what possible connections there are or may be between goodness and will; and that none of these can justify us in asserting that This is good is identical with This is willed. On the other hand it will appear that some of them may be easily confused with this assertion of identity; and that therefore the confusion is likely to have been made. This part of my argument will, therefore, already go some way towards meeting the second objection. But what must be conclusive against this is to shew that any possible connection between will and goodness except the absolute identity in question, would not be sufficient to give an enquiry into Will the smallest relevance to the proof of any ethical conclusion. (§ 77 ¶ 2)

It has been customary, since Kant's time, to assert that Cognition, Volition, and Feeling are three fundamentally distinct attitudes of the mind towards reality. They are three distinct ways of experiencing, and each of them informs us of a distinct aspect under which reality may be considered. The Epistemological method of approaching Metaphysics rests on the assumption that by considering what is implied in Cognition—what is its ideal—we may discover what properties the world must have, if it is to be true. And similarly it is held that by considering what is implied in the fact of Willing or Feeling—what is the ideal which they presuppose—we may discover what properties the world must have, if it is to be good or beautiful. The orthodox Idealistic Epistemologist differs from the Sensationalist or Empiricist in holding that what we directly cognise is neither all true nor yet the whole truth: in order to reject the false and to discover further truths we must, he says, not take cognition merely as it presents itself, but discover what is implied in it. And similarly the orthodox Metaphysical Ethicist differs from the mere Naturalist, in holding that not everything which we actually will is good, nor, if good, completely good: what is really good is that which is implied in the essential nature of will. Others again think that Feeling, and not Will, is the fundamental datum for Ethics. But, in either case, it is agreed that Ethics has some relation to Will or Feeling which it has not to Cognition, and which other objects of study have to Cognition. Will or Feeling, on the one hand, and Cognition, on the other, are regarded as in some sense co-ordinate sources of philosophical knowledge—the one of Practical, the other of Theoretical Philosophy. (§ 78 ¶ 1)

What, that is true, can possibly be meant by this view? (§ 78 ¶ 2)

First of all, it may be meant that, just as, by reflection on our perceptual and sensory experience, we become aware of the distinction between truth and falsehood, so it is by reflection on our experiences of feeling and willing that we become more aware of ethical distinctions. We should not know what was meant by thinking one thing better than another unless the attitude of our will or feeling towards one thing was different from its attitude towards another. All this may be admitted. But so far we have only the psychological fact that it is only because we will or feel things in a certain way, that we ever come to think them good; just as it is only because we have certain perceptual experiences, that we ever come to think things true. Here, then, is a special connection between willing and goodness; but it is only a causal connection—that willing is a necessary condition for the cognition of goodness. (§ 79 ¶ 1)

But it may be said further that willing and feeling are not only the origin of cognitions of goodness; but that to will a thing, or to have a certain feeling towards a thing, is the same thing as to think it good. And it may be admitted that even this is generally true in a sense. It does seem to be true that we hardly ever think a thing good, and never very decidedly, without at the same time having a special attitude of feeling or will towards it; though it is certainly not the case that this is true universally. And the converse may possibly be true universally: it may be the case that a perception of goodness is included in the complex facts which we mean by willing by having certain kinds of feeling. Let us admit then, that to think a thing good and to will it are the same thing in this sense, that, wherever the latter occurs, the former also occurs as part of it; and even that they are generally the same thing in the converse sense, that when the former occurs it is generally part of the latter. (§ 79 ¶ 2)

These facts may seem to give countenance to the general assertion that to think a thing good is to prefer it or approve it, in the sense in which preference and approval denote certain kinds of will or feeling. It seems to be always true that when we thus prefer or approve, there is included in that fact the fact that we think good; and it is certainly true, in an immense majority of instances, that when we think good, we also prefer or approve. It is natural enough, then, to say that to think good is to prefer. And what more natural than to add: When I say a thing is good, I mean that I prefer it? And yet this natural addition involves a gross confusion. Even if it be true that to think good is the same thing as to prefer (which, as we have seen, is never true in the sense that they are absolutely identical; and not always true in the sense that they occur together), yet it is not true that what you think, when you think a thing good, is that you prefer it. Even if your thinking the thing good is the same thing as your preference of it, yet the goodness of the thing—that of which you think—is, for that very reason, obviously not the same thing as your preference of it. WHether you have a certain thought or not is one question; and whether what you think is true is quite a different one, upon which the answer to the first has not the least bearing. The fact that you prefer a thing does not tend to shew that the thing is good; even if it does shew that you think it so. (§ 80 ¶ 1)

It seems to be owing to this confusion, that the question What is good? is thought to be identical with the question What is preferred? It is said, with sufficient truth, that you would never know a thing was good unless you preferred it, just as you would never know a thing existed unless you perceived it. But it is added, and this is false, that you would never know a thing was good unless you knew that you preferred it, or that it existed unless knew that you perceived it. And it is finally added, and this is utterly false, that you cannot distinguish the fact that a thing is good from the fact that you prefer it, or the fact that it exists from the fact that you perceive it. It is often pointed out that I cannot at any given moment distinguish what is true from what I think so: and this is true. But though I cannot distinguish what is true from what I think so, I always can distinguish what I mean by saying that it is true from what I mean by saying that I think so. For I understand the meaning of the supposition that what I think true may nevertheless be false. When, therefore, I assert that it is true I mean to assert something different from the fact that I think so. What I think, namely that something is true, is always quite distinct from the fact that I think it. The assertion that it is true does not even include the assertion that I think it so; although, of course, whenever I do think a thing true, it is, as a matter of fact, also true that I do think it. This tautologous proposition that for a thing to be thought true it is necessary that it should be thought is, however, commonly identified with the proposition that for a thing to be true it is necessary that it should be thought. A very little reflection should suffice to convince anyone that this identification is erroneous; and a very little more will shew that, if so, we must mean by true something which includes no reference to thinking or to any other psychical fact. It may be difficult to discover precisely what we mean—to hold the object in question before us, so as to compare it with other objects: but that we do mean something distinct and unique can no longer be matter of doubt. That to be true means to be thought in a certain way is, therefore, certainly false. Yet this assertion plays the most essential part in Kant's Copernican revolution of philosophy, and renders worthless the whole mass of modern literature, to which that revolution has given rise, and which is called Epistemology. Kant held that what was unified in a certain manner by the synthetic activity of thought was ipso facto true: that this was the very meaning of the word. Whereas it is plain that the only connection which can possibly hold between being true and being thought in a certain way, is that the latter should be a criterion or test of the former. In order, however, to establish that it is so, it would be necessary to establish by the methods of induction that what was true was always thought in a certain way. Modern Epistemology dispenses with this long and difficult investigation at the cost of the self-contradictory assumption that truth and the criterion of truth are one and the same thing. (§ 80 ¶ 2)

It is, then, a very natural, though an uterrly false supposition that for a thing to be true is the same thing as for it to be perceived or thought of in a certain way. And since, for the reasons given above, the fact of preference seems roughly to stand in the same relation to thinking things good, in which the fact of perception stands to thinking that they are true or exist, it is very natural that for a thing to be good should be supposed identical with its being preferred in a certain way. But once this coordination of Volition and Cognition has been accepted, it is again very natural that every fact which seems to support the conclusion that being true is identical with being cognised should confirm the corresponding conclusion that being good is identical with being willed. It will, therefore, be in place to point out another confusion, which seems to have had great influence in causing acceptance of the view that to be true is the same thing as to be cognised. (§ 81 ¶ 1)

This confusion is due to a failure to observe that when we say we have a sensation or perception or that we know a thing, we mean to assert not only that our mind is cognitive, but also that that which it cognises is true. It is not observed that the usage of these words is such that, if a thing be untrue, that fact alone is sufficient to justify us in saying that the person who says he perceives or knows it, does not perceive or know it, without our either enquiring whether, or assuming that, his state of mind differs in any respect from what it would have been had he perceived or known. By this denial we do not accuse him of an error in introspection, even if there was such an error: we do not deny that he was aware of a certain object, nor even that his state of mind was exactly such as he took it to be: we merely deny that the object, of which he was aware, had a certain property. It is, however, commonly supposed that when we assert a thing to be perceived or known, we are asserting one fact only; and since of the two facts which we really assert, the existence of a psychical state is by far the easier to distinguish, it is supposed that this is the only one which we do assert. Thus perception and sensation have come to be regarded as if they denoted certain states of mind and nothing more; a mistake which was the easier to make since the commonest state of mind, to which we give a name which does not imply that its object is true, namely imagination, may, with some plausibility, be supposed to differ from sensation and perception not only in the property possessed by its object, but also in its character as a state of mind. It has thus come to be supposed that the only difference between perception and imagination, by which they can be defined, must be a merely psychical difference: and, if this were the case, it would follow at once that to be true was identical with being cognised in a certain way; since the assertion that a thing is perceived does certainly include the assertion that it is true, and if, nevertheless, that it is perceived means only that the mind has a certain attitude towards it, then its truth must be identical with the fact that it is regarded in this way. We may, then, attribute the view that to be true means to be cognised in a certain way partly to the failure to perceive that certain words, which are commonly supposed to stand for nothing more than a certain kind of cognitive state, do, in fact, also include a reference to the truth of the object of such states. (§ 81 ¶ 2)

I will now sum up account of the apparent connections between will and ethical propositions, which seem to support the vague conviction that This is good is somehow identical with This is willed in a certain way. (1) It may be maintained, with sufficient show of truth, that it is only because certain things were originally willed, that we ever came to have ethical convictions at all. And it has been too commonly assumed that to shew what was the cause of a thing is the same thing as to shew the thing itself is. IT is, however, hardly necessary to point out that this is not the case. (2) It may be further maintained, with some plausibility, that to think a thing good and to will it in a certain way are now as a matter of fact identical. We must, however, distinguish certain possible meanings of this assertion. It may be admitted that when we think a thing good, we generally have a special attitude of will or feeling towards it; and that, perhaps, when we will it in a certain way, we do always think it good. But the very fact that we can thus distinguish the question whether, though the one is always accompanied by the other, yet this other may not always be accompanied by the first, shews that the two things are not, in the strict sense, identical. The fact is that, whatever we mean by will, or by any form of the will, the fact we mean by it certainly always includes something else beside the thinking a thing good: and hence that, when willing and thinking good are asserted to be identical, the most that can be meant is that this other element in will always both accompanies and is accompanied by the thinking good; and this, as has been said, is of very doubtful truth. Even, however, if it were strictly true, the fact that the two things can be distinguished is fatal to the assumed coordination between will and cognition, in one of the senses in which that assumption is commonly made. For it is only in respect of the other element in will, that volition differs from cognition; whereas it is only in respect of the fact that volition, or some form of volition, includes a cognition of goodness, that will can have the same relation to ethical, which cognition has to metaphysical, propositions. Accordingly the fact of volition, as a whole, that is, if we include in it the element which makes it volition and distinguishes it from cognition, has not the same relation to ethical propositions which cognition has to those which are metaphysical. Volition and cognition are not coordinate ways of experiencing, since it is only in so far as volition denotes a complex fact, which includes in it the one identical simple fact, which is meant by cognition, that volition is a way of experiencing at all. (§ 82 ¶ 1)

But, (3) if we allow the terms volition or will to stand for thinking good, although they certainly do not commonly stand for this, there still remains the question: What connection would this fact establish between volition and Ethics? Could the enquiry into what was willed be identical with the ethical enquiry into what was good? It is plain enough that they could not be identical; though it is also plain why they should be thought so. The question What is good? is confused with the question What is thought good? and the question What is true? with the question What is thought true? for two main reasons. (1) One of these is the general difficulty that is found in distinguishing what is cognised from the cognition of it. It is observed that I certainly cannot cognise anything that is true without cognising it. Since, therefore, whenever I know a thing that is true, the thing is certainly cognised, it is assumed that for a thing to be true at all is the same thing as for it to be cognised. And (2) it is not observed that certain words, which are supposed to denote only peculiar species of cognition, do as a matter of fact also denote that the object cognised is true. Thus if perception be taken to denote only a certain kind of mental fact, then, since the object of it is always true, it becomes easy to suppose that to be true means only to be object to a mental state of that kind. And similarly it is easy to suppose that to be truly good differs from being falsely thought so, solely in respect of the fact that to be the former is to be the object of a volition differing from that of which an apparent good is the object, in the same way in which a perception (on this supposition) differs from an illusion. (§ 82 ¶ 2)

Being good, then, is not identical with being willed or felt in any kind of way, any more than being true is identical with being thought in any kind of wy. But let us suppose this to be admitted: Is it still possible that an enquiry into the nature of will or feeling should be a necessary step to the proof of ethical conclusions? If being good and being willed are not identical, then the most that can be maintained with regard to the connection of goodness with will is that what is good is always also willed in a certain way, and that what is willed in a certain way is always also good. And it may be said that this is all that is meant by those metaphysical writers who profess to base Ethics upon the Metaphysics of Will. What would follow from this supposition? (§ 83 ¶ 1)

It is plain that if what is willed in a certain way were always also good, then the fact that a thing was so willed would be a criterion of its goodness. But in order to establish that will is a criterion of goodness, we must be able to shew first and separately that in a great number of the instances in which we find a certain kind of will we also find that the objects of that will are good. We might, then, perhaps, be entitled to infer that in a few instances, where it was not obvious whether a thing was good or not but was obvious that it was willed in the way required, the thing was really good, since it had the property which in all other instances we had found to be accompanied by goodness. A reference to will might thus, just conceivably, become of use towards the end of our ethical investigations, when we had already been able to shew, independently, of a vast number of different objects that they were really good and in what degree they were so. And against even this conceivable utility it may be urged (1) That it is impossible to see why it should not be as easy (and it would certainly be the more secure way) to prove that the thing in question was good, by the same methods which we had used in proving that other things were good, as by reference to our criterion; and (2) That, if we set ourselves seriously to find out what things are good, we shall see reason to think (as will appear in Chapter VI) that they have no other property, both common and peculiar to them, beside their goodness—that, in fact, there is no criterion of goodness. (§ 83 ¶ 2)

But to consider whether any form of will is or is not a criterion of goodness is quite unnecessary for our purpose here; since none of those writers who profess to base their Ethics on an investigation of will have ever recognised the need of proving directly and independently that all the things which are willed in a certain way are good. They make no attempt to shew that will is a criterion of goodness; and no stronger evidence could be given that they do not recognise that this, at most, is all it can be. As has been just pointed out, if we are to maintain that whatever is willed in a certain way is also good, we must in the first place be able to shew that certain things have one property goodness, and that the same things also have the other property that they are willed in a certain way. And secondly we must be able to shew this in a very large number of instances, if we are to be entitled to claim any assent for the proposition that these two properties always accompany one another: even when this was shewn it would still be doubtful whether the inference from generally to always would be valid, and almost certain that this doubtful principle would be useless. But the very question which it is the business of Ethics to answer is this question what things are good; and, so long as Hedonism retains its present popularity, it must be admitted that it is a question upon which there is scarcely any agreement and which therefore requires the most careful examination. The greatest and most difficult part of the business of Ethics would therefore require to have been already accomplished before we could be entitled to claim that anything was a criterion of goodness. If, on the other hand, to be willed in a certain way was identical with being good, then indeed we should be entitled to start our ethical investigations by enquiring what was willed in the way required. That this is the way in which metaphysical writers start their investigations seems to shew conclusively that they are influenced by the idea that goodness is identical with being willed. They do not recognise that the question What is good? is a different one from the question What is willed in a certain way? Thus we find Green explicitly stating that the common characteristic of the good is that it satisfies some desire. If we are to take this statement strictly, it obviously asserts that good things have no characteristic in common, except that they satisfy some desire—not even, therefore, that they are good. And this can only be the case, if being good is identical with satisfying desire: if good is merely another name for desire-satisfying. There could be no plainer instance of the naturalistic fallacy. And we cannot take the statement as a mere verbal slip, which does not affect the validity of Green’s argument. For he nowhere either gives or pretends to give any reason for believing anything to be good in any sense, except that it is what would satisfy a particular kind of desire—the kind of desire which he tries to shew to be that of a moral agent. An unhappy alternative is before us. Such reasoning would give valid reasons for his conclusions, if, and only if, being good and being desired in a particular way were identical: and in this case, as we have seen in Chapter I, his conclusions would not be ethical. On the other hand, if the two are not identical, his conclusions may be ethical and may even be right, but he has not given us a single reason for believing them. The thing which a scientific Ethics is required to shew, namely that certain things are really good, he has assumed to begin with, in assuming that things which are willed in a certain way are always good. We may, therefore, have as much respect for Green’s conclusions as for those of any other man who details to us his ethical convictions: but that any of his arguments are such as to give us any reason for holding that Green’s convictions are more likely to be true than those of any other man, must be clearly denied. The Prolegomena to Ethics is quite as far as Mr Spencer’s Data of Ethics, from making the smallest contribution to the solution of ethical problems. (§ 84 ¶ 1)