In disputes, therefore, as to whether particular kinds of value are or are not "subjective," I think that the issue which is really felt to be important, almost always by one side, and often by both, is not really the issue between "subjective" and "non-subjective," but between "intrinsic" and "non-intrinsic." And not only is this felt to be the more important issue; I think it really is so. For the difference that must be made to our view of the Universe, according as we hold that some kinds of value are "intrinsic" or that none are, is much greater than any which follows from a mere difference of opinion as to whether some are "non-subjective," or all without exception "subjective." To hold that any kinds of value are "intrinsic" entails the recognition of a kind of predicate extremely different from any we should otherwise have to recognise and perhaps unique; whereas it is in any case certain that there are "objective" predicates as well as "subjective."
But now what is this "internality" of which I have been speaking? What is meant by saying with regard to a kind of value that it is "intrinsic?" To express roughly what is meant is, I think, simple enough; and everybody will recognise it at once, as a notion which is constantly in people's heads; but I want to dwell upon it at some length, because I know of no place where it is expressly explained and defined, and because, though it seems very simple and fundamental, the task of defining it precisely is by no moans easy and involves some difficulties which I must confess that I do not know how to solve.
I have already given incidentally the main idea in speaking of that evolutionary interpretation of "goodness," according to which, as I said, goodness would be "objective" but would not be "intrinsic." I there used as equivalent to the assertion that 'better,' on that definition, would not be 'intrinsic,' the assertion that the question whether one type of being A was better than another B would not depend solely on the intrinsic natures of A and B, but on circumstances and the laws of nature. And I think that this phrase will in fact suggest to everybody just what I do mean by "intrinsic" value. We can, in fact, set up the following definition. To say that a kind of value is "intrinsic" means merely that the question whether a thing possesses it, and in what degree it possesses it, depends solely on the intrinsic nature of the thing in question.
But though this definition does, I think, convey exactly what I mean, I want to dwell upon its meaning, partly because the conception of 'differing in intrinsic nature which I believe to be of fundamental importance, is liable to be confused with other conceptions, and partly because the definition involves notions, which I do not know how to define exactly.
When I say, with regard to any particular kind of value, that the question whether and in what degree anything possesses it depends solely on the intrinsic nature of the thing in question, I mean to say two different things at the same time. I mean to say (1) that it is impossible for what is strictly one and the same thing to possess that kind of value at one time, or in one set of circumstances, and not to possess it at another; and equally impossible for it to possess it in one degree at one time, or in one set of circumstances, and to possess it in a different degree at another, or in a different set. This, I think, is obviously part of what is naturally conveyed by saying that the question whether and in what degree a thing possesses the kind of value in question always depends solely on the intrinsic nature of the thing. For if x and y have different intrinsic natures, it follows that x cannot be quite strictly one and the same thing as y; and hence if x and y can have a different intrinsic value, only where their intrinsic natures are different, it follows that one and the same thing must always have the same intrinsic value. This, then, is part of what is meant; and about this part I think I need say no more, except to call attention to the fact that it involves a conception, which as we shall see is also involved in the other part, and which involves the same difficulty in both cases—I mean, the conception which is expressed by the word 'impossible.' (2) The second part of what is meant is that if a given thing possesses any kind of intrinsic value in a certain degree, then not only must that same thing possess it, under all circumstances, in the same degree, but also anything exactly like it, must, under all circumstances, possess it in exactly the same degree. Or to put it in the corresponding negative form: It is impossible that of two exactly similar things one should possess it and the other not, or that one should possess it in one degree, and the other in a different one.
I think this second proposition also is naturally conveyed by saying that the kind of value in question depends solely on the intrinsic nature of what possesses it. For we should naturally say of two things which were exactly alike intrinsically, in spite of their being two, that they possessed the same intrinsic nature. But it is important to call attention expressly to the fact that what I mean by the expression 'having a different intrinsic nature' is equivalent to 'not exactly alike' because here there is real risk of confusion between this conception and a different one. This comes about as follows. It is natural to suppose that the phrase 'having a different intrinsic nature' is equivalent to the phrase 'intrinsically different' or 'having different intrinsic properties.' But, if we do make this identification, there is a risk of confusion. For it is obvious that there is a sense in which, when things are exactly like, they must be 'intrinsically different' and have different intrinsic properties, merely because they are two. For instance, two patches of colour may be exactly alike, in spite of the fact that each possesses a constituent which the other does not possess, provided only that their two constituents are exactly alike. And yet, in a certain sense, it is obvious that the fact that each has a constituent, which the other has not got, does constitute an intrinsic difference between them, and implies that each has an intrinsic property which the other has not got. And even where the two things are simple the mere fact that they are numerically different does in a sense constitute an intrinsic difference between them, and each will have at least one intrinsic property which the other has not got—namely that of being identical with itself. It is obvious therefore that the phrases 'intrinsically different' and 'having different intrinsic properties' are ambiguous. They may be used in such a sense that to say of two things that they are intrinsically different or have different intrinsic properties does not imply that they are not exactly alike, but only that they are numerically different. Or they may be used in a sense in which two things can be said to be intrinsically different, and to have different intrinsic properties only when they are not exactly alike. It is, therefore, extremely important to insist that when I say: Two things can differ in intrinsic value, only when they have different intrinsic natures, I am using the expression 'having different intrinsic natures' in the latter sense and not the former:—in a sense in which the mere fact that two things are two, or differ numerically, does not imply that they have different intrinsic natures, but in which they can be said to have different intrinsic natures, only where, besides differing numerically, they are also not exactly alike.
But as soon as this is explained, another risk of confusion arises owing to the fact that when people contrast mere numerical difference with a kind of intrinsic difference, which is not merely numerical, they are apt to identify the latter with qualitative difference. It might, therefore, easily be thought that by 'difference in intrinsic nature' I mean 'difference in quality.' But this identification of difference in quality with difference in intrinsic nature would also be a mistake. It is true that what is commonly meant by difference of quality, in the strict sense, always is a difference of intrinsic nature: two things cannot differ in quality without differing in intrinsic nature; and that fact is one of the most important facts about qualitative difference. But the converse is by no means also true: although two things cannot differ in quality without differing in intrinsic nature, they can differ in intrinsic nature without differing in quality; or, in other words, difference in quality is only one species of difference in intrinsic nature. That this is so follows from the fact that, as I explained, I am using the phrase 'different in intrinsic nature' as equivalent to 'not exactly like for it is quite plain that two things may not be exactly alike, in spite of the fact that they don't differ in quality, e.g. if the only difference between them were in respect of the degree in which they possess some quality they do possess. Nobody would say that a very loud sound was exactly like a very soft one, even if they were exactly like in quality; and yet it is plain there is a sense in which their intrinsic nature is different For this reason alone qualitative difference cannot be identified with difference in intrinsic nature. And there are still other reasons. Difference in size, for instance may be a difference in intrinsic nature, in the sense I mean, but it can hardly be called a difference in quality. Or take such a difference as the difference between two patterns consisting in the fact that the one is a yellow circle with a red spot in the middle, and the other a yellow circle with a blue spot in the middle. This difference would perhaps be loosely called a difference of quality; but obviously it would be more accurate to call it a difference which consists in the fact that the one pattern has a constituent which is qualitatively different from any which the other has; and the difference between being qualitatively different and having qualitatively different constituents is important both because the latter can only be defined in terms of the former, and because it is possible for simple things to differ from one another in the former way, whereas it is only possible for complex things to differ in the latter.
I hope this is sufficient to make clear exactly what the conception is which I am expressing by the phrase "different in intrinsic nature." The important points are (1) that it is a kind of difference which does not hold between two things, when they are merely numerically different, but only when, besides being numerically different, they are also not exactly alike and (2) that it is not identical with qualitative difference; although qualitative difference is one particular species of it. The conception seems to me to be an extremely important and fundamental one, although, so far as I can see, it has no quite simple and unambiguous name: and this is the reason why I have dwelt on it at such length. "Not exactly like" is the least ambiguous way of expressing it; but this has the disadvantage that it looks as if the idea of exact likeness were the fundamental one from which this was derived, whereas I believe the contrary to be the case. For this reason it is perhaps better to stick to the cumbrous phrase "different in intrinsic nature."
So much for the question what is meant by saying of two things that they "differ in intrinsic nature." We have now to turn to the more difficult question as to what is meant by the words "impossible" and "necessary" in the statement: A kind of value is intrinsic if and only if, it is impossible that x and y should have different values of the kind, unless they differ in intrinsic nature; and in the equivalent statement: A kind of value is intrinsic if and only if, when anything possesses it, that same thing or anything exactly like it would necessarily or must always, under all circumstances, possess it in exactly the same degree.
As regards the meaning of this necessity and impossibility, we may begin by making two points clear.