^{page 24 note 1} Cf. D. Daiches Raphael, The Moral Sense, Chapters V and VI.

^{page 24 note 2} Cf. Moore, G. E., Principia Ethica, p. 7et seqGoogle Scholar.

^{page 28 note 1} Cf. Charles Stevenson, Ethics and Language, Chapter 1.

^{page 29 note 1} Ross, , Foundations of Ethics, pp. 83–86Google Scholar; Broad, , “Some of the Main Problems of Ethics,” Philosophy, 1946, p. 117Google Scholar.

^{page 30 note 1} One desperate expedient might occur to North. He might say that it is not the bare presence of the promise-keeping feature that entails the rightness of the act, but the presence of this feature, coupled with the absence of any features which would entail its wrongness. His general rules would then be, not of the form “ ‘χ has φ’ entails ‘χ is right’,” but of the form “ ‘χ has φ and χ has no ψ such that “ ‘χ has φ” entails “χ is wrong”’ entails ‘ χ is right’.” But the suggestion is inadmissable, since (i) the establishment of the general proposition “χ has no ψ, etc.” would require the enumeration of all those features which would make it wrong to keep a promise, and (ii) any rule of the form “ ‘ χ has ψ’ entails ‘ χ is wrong’ ” would require expansion in exactly the same way as the “right-making” rule; which would involve an infinite regress of such expansions. Besides having this theoretical defect, the suggested model is, of course, practically absurd.

^{page 31 note 1} E.g. There was a certain plausibility in saying “My feeling morally obliged to pursue such a course (or end) presupposes my believing that it is right (or good),” and thence concluding that this belief cannot be “reduced to” the feeling which it arouses. (For examples of this sort of argument, see Ross, , op.cit., pp. 261–262Google Scholar, and Broad, , op. cit., p. 115Google Scholar.) But the weakness of the reasoning is more clearly exposed when the sentence is re-written as “My feeling morally obliged to pursue such a course presupposes my believing that I am morally obliged to pursue it.” The point is that “presupposes” and “believing” are both ambiguous. If “presupposes” means “causally requires” and “believing” is used in its ordinary sense, then it is obviously false that the beliefs which occasion such a feeling invariably include some belief which would be correctly described in these terms. (Compare: “My feeling frightened presupposes my believing that I am frightened.”) But the argument begins to have weight against the “analysability” of beliefs correctly so described only if they are invariably present as occasioning factors. If, on the other hand, “presupposes” means “logically requires,” then “believing” might be used in a queer sense such that the sentence is tautologically true. But this result is secured only by defining “believing” (used in this sense) in terms of feeling (compare the sense in which “thinking χ funny” means “being amused by χ”): and this was precisely the result which North sought to avoid.

^{page 32 note 1} Cf. Wisdom, , “Metaphysics and Verification,” Mind, 1938Google Scholar.

^{page 33 note 1} Hardie, , “The Paradox of Phenomenalism,” Proceedings of the Aristotelian Society, 1945–1946, p. 150Google Scholar.

^{page 33 note 2} Wisdom.