^{page 329 note 1} See Urmson, J. O., ‘On Grading’, in Logic and Language, II. ed. by Flew, A. (Oxford, 1955), pp. 159–86;Google Scholar Hare, R. M., The Language of Morals (Oxford, 1952), chaps. 6-9;Google Scholar Nowell-Smith, P. H., Ethics (London, 1954), pp. 164–80 passim;Google Scholar Toulmin, S. E., The Uses of Argument (Cambridge, 1958), pp. 30 ff.:Google Scholar Sparshott, F. E., An Enquiry into Goodness (Chicago, 1958), pp. 83, 173–77. The argument is also adumbrated in Stevenson, C. L., Ethics and Language (New Haven, 1944), pp. 85 ff.Google Scholar

^{page 330 note 1} This distinction is related to but not identical with the traditional distinction between intension and extension. The major differences stem from the ‘subjective’ reference of the commendatory meaning and from the fact that the descriptive criteria of application refer to properties at least as much as to objects.

^{page 331 note 1} Analysis of Knowledge and Valuation (LaSalle, 111., 1946), p. 133.Google Scholar For other philosophers who likewise locate ‘criterion’ on the ‘subjective’ rather than the ‘objective’ side, in that they view it as something had or done by the knower rather than as properties of the object of knowledge, see Peirce, C. S., Collected Papers (Cambridge, Mass., 1932), II, para. 30 (especially the last sentence where Peirce defines ‘a criterion’ as ‘a method of experiment’);Google Scholar Wittgenstein, L., Philosophical Investigations (Oxford, 1953), I, sec. 141.Google Scholar (Contrast, however, The Blue and Brown Books \Oxford, 1958], pp. 24–25.)Google Scholar This ‘subjective’ interpretation goes back at least to Sextus Empiricus; see Outlines of Pyrrhonism, I, 21 ff.; II, 14 ff.; Against the Logicians, I, 27 ff.Google Scholar

^{page 334 note 1} There is not space here to go into the question of the sense in which, or the extent to which, Latrone's statement is a ‘moral’ judgment. Some recent philosophers have been trying to combine a morally neutral definition of the concept of morality with the insistence that morality (in this morally neutral sense of the term) must encompass requirements ranging from universalizability to concern for the common good. I do not think such a combination can consistently be made, for these requirements are not morally neutral. Moreover, the requirements in question seem to overlook the fact that the concept of morality has traditionally been regarded as having among its species such views as ethical egoism, including that version of it whereby a man is unwilling to grant to others what he claims for himself. To dispute whether Callicles and some of the other sophists depicted by Plato were taking a ‘moral’ position is at least in part verbal; in any case, Plato seems to have thought that the disagreement between them and Socrates was a moral one. For a further discussion of this question, see my ‘Meta-Ethics and Normative Ethics’, Mind, LXIX (1960), 187–205, especially pp. 191-97.Google Scholar

^{page 338 note 1} See my ‘Positive “Ethics” and Normative ‘Science’,’ Philosophical Review’ LXIX (1960), 311-30.Google Scholar

^{page 338 note 2} A related example of disparateness is found in Hare's argument that if ‘x is good’ means the same as ‘x has descriptive characteristics C’, then it becomes impossible to use 'x is good’ in order to commend x, as we do in fact intend to do (op. cit., pp. 84-91). It has not been noticed, however, that this argument from unfulfilled function cuts both ways. For if to say ‘x is good’ means the same as to commend x, then it becomes impossible to commend x for being good, since in saying ‘x is good’ we shall already have commended x. Yet often we do want to commend something for being good, as when we say, ‘You ought to commend Andrew for being such a good student’; and in such cases the ‘good’ for which we commend consists in descriptive characteristics. Thus Hare's argument against a purely naturalistic or descriptive definition of ‘good’ applies just as much against a purely commendatory definition. Since writing this note 1 have come upon a similar point in Cox, J. W. Roxbee, ‘Commending and Describing’, Philosophical Quarterly, 11 (1961), p.44.CrossRefGoogle Scholar

^{page 339 note 1} It might seem possible to hold that the descriptive criteria for applying ‘good’ have in common something far more specific than merely being descriptive features, namely, the property of satisfying wants or answering to interests. (For recent versions of these, see, for the first, Sparshott, op. cit., pp. 122 ff.; for the second, Ziff, P., Semantic Analysis \Ithaca, N. Y., 1960], pp. 212 ff.)Google Scholar It is to be noted, however, that such a view reverts the ‘objective’ features of what is good back to ‘subjective’ features (wants or interests) of the persons who say that something is good. In the terms being used in this paper, it is as if we were to hold that all good things have in common the property of being objects of commendation. This still tells us nothing about the things as such but only about the subject's reaction to them. The case would be different if we were to insist that only what satisfies ‘real’ needs or answers to ‘real’ interests is good; but such a normative view of ‘good’ would lose the generality aimed at by those who wish to grasp the meaning common to all uses of ‘good’, without differentiating between morally better and worse uses or users.

^{page 344 note 1} In view of my emphasis on the moral significance of the ‘how’ of commending, I wish to point ou some respects in which the above analysis differs from other recent discussions of ethical methods or procedures which refer to a ‘competent judge’ or an ‘ideal observer’ or a ‘qualified attitude’. There are three main differences: (1) The procedures of evaluation, as I conceive them, characterise the moral agent himself and not only or primarily a judge or observer of the agent. (2) In so far as the procedures do characterise a judge or observer of the moral situation, the latter's moral quality is independent of the observer; that is, persons, acts, or qualities have their moral goodness or badness in themselves and not in virtue of how they are evaluated by a judge or observer. (3) The procedures of evaluation, as I view them, are themselves normatively moral, and not morally neutral. That is, I have not attempted to present the traits of an observer, judge, or attitude in morally neutral terms, such that these traits could then be used without circularity to differentiate the morally good from the morally bad. These attempts seem to me to fail in one of two ways: either they do not succeed in making the desired differentiation, or if they do succeed, they are circular.