¹ Smith, Norman Kemp, “The Naturalism of Hume”, Mind (April 1905), pp. 149–73, 335–47CrossRefGoogle Scholar. I have put references to this article into the text using “S” followed by the page number, so that S149 is to be read Smith, p. 149.

² Reid, Thomas, Essays on the Powers of the Human Mind, printed for Bell & Bradfute (Edinburgh: 1819), Vol. I, p. 267Google Scholar.

³ ibid., p. 270.

⁴ I want to isolate and then ignore a third theory of the general nature of Hume's philosophy. This is that Hume develops the principles of his predecessors to a sceptical conclusion not to reject the question of justification as irrelevant, but to make clear the necessity of a new justificatory principle, which he then introduces and which shows our suppositions to be justified. By this new principle our suppositions are justified not in terms of being derived from experience, but in terms of their practical consequences: Hume is thus some sort of pragmatist. This theory was developed by Bayley, F. C. in Causes and Evidence of Beliefs, privately printed (Mount Hermon, Mass.: 1936)Google Scholar, and Bayley's main points have been repeated by Randall, John Herman in his Career of Philosophy, Columbia University Press (New York: 1964), pp. 640–49Google Scholar. This theory also has its sources in Kemp Smith's article, I think, for he speaks of the alternative to the failure of the empiricist principle being “not scepticism, but the practical test of human validity” (S152). If this third theory is there mingled with the one I attribute to Kemp Smith, I have not extracted it. I think that it lacks textual support, that a look at Bayley's systematic attempt to apply it to the details of Hume's analyses shows a systematic distortion of them, and that what I say in the paper, if reworked, is sufficient to refute it.

⁵ References to Hume's works have been put into the text using “T” for the Treatise (Selby-Bigge edition) and “E” for the Enquiry (Selby-Bigge edition). T79–80 thus refers to pp. 79–80 of the Treatise.

⁶ Hume's claim is contentious. As John W. Cook has pointed out, Hume's explanation of the effects of constancy presupposes that it is a necessary truth that our perceptions be perceived to exist (“Hume's Scepticism with Regard to the Senses,” American Philosophical Quarterly, Vol. 5 (1968), pp. 1–17).Google Scholar

⁷ Hume introduces his discussion of the effects of constancy by saying he is “here accounting for the opinions and belief of the vulgar” and so is not going to distinguish between “the objects and perceptions of the senses” (T202). It is one of my reasons for examining only the effects of coherence that no such limitation on the scope of the effects is claimed: what infects the vulgar also infects the philosophers. I think it true, however, that Hume later extends his account of the effects of constancy to the philosophers (see T215), and it is also true that some commentators have extended Hume's stated limitation to his prior remarks on coherence. John W. Cook thus quotes T202 to support a claim that Hume is defining the vulgar notion of body at T188 (Op. cit., p. 1). If one does that, it is difficult to explain what it is tfiat the philosopher supposes the distinct existence of.

⁸ This claim needs further support. For Hume does talk of “trivial qualities of the fancy” in regard to our supposition (T217; see also T226 in regard to the supposition of substance). The claim I make rests upon Hume's saying we “suppose an external universe” because of “a natural instinct or prepossession” (E151) and upon one of the examples he gives when he talks of permanent principles of the imagination being an example of the supposition of external objects (T225, at the bottom).

⁹ The inference from principles being permanent, universal, and irresistible to the results of those principles having those qualities is an inference sanctioned by Hume. For he introduces the distinction between those principles and “changeable, weak, and irregular” ones to account for his claim that the suppositions of the ancient philosophers are neither “unavoidable to mankind, nor necessary, or so much as useful in the conduct of life” (T225).

¹⁰ This obviously presupposes a distinction between “inferring a further coherence” and inferring causally, but though Hume is committed to there being such a distinction, I cannot now see what it could be. The examples Hume cites of “inferring a further coherence” all seem to be cases where we fill in gaps in our past experience: there was a door there because we heard a creak. The cases are thus instances of inferring a cause from an effect, but it would seem that inferring causally includes that sort of inference as well as inferring effects from causes (see T225). Even if not, surely the difference is not sufficient to make the one inference reasonable and the other unreasonable.

¹¹ This analysis of the activating conditions differs from Robert Paul Wolff's in his “Hume's Theory of Mental Activity”, reprinted in Hume, ed. Chappell, V. C. (Anchor Books: 1966). See pp. 127–28CrossRefGoogle Scholar.

¹² Hendel, Charles W., Studies in the Philosophy of David Hume (Bobbs-Merrill: 1963), p. 199Google Scholar. Hendel has injudiciously lifted the remark from Hume's views on his problems with the idea of personal identity and put it about our views on the existence of unperceived objects.

¹³ Hume makes this same point about causal inferences. See T222–23. This is one of the items I would cite in order to prove my wider contention that Hume is the same sort of sceptic in regard to causation and personal identity.

¹⁴ See e.g. Bitzer, Lloyd F., “Hume's Philosophy in George Campbell's Philosophy of Rhetoric”, Philosophy and Rhetoric, Vol. 2 (Summer 1969), p. 145Google Scholar.

¹⁵ This conception of the source of scepticism can be found in Pierre Daniel Huet's Traité de la faiblesse de l'esprit humain, printed in English for Gysbert Dommer in 1725. See pp. 18, 22, 24, 128, 133–4, but see p. 151. Hume was clearly familiar with Huet's work, for he mentions it in A Letter from a Gentleman to his friend in Edinburgh, ed. Mossner, E. C. and Price, J. V. (Edinburgh: 1967), p. 21Google Scholar.

¹⁶ See e.g. E164. See S164ff. for Kemp Smith's analysis of Hume's two senses of “being reasonable”, the one with regard to relations between ideas and the other the one I refer to here. I am convinced that Hume has a relatively well articulated theory of what is reasonable. It has its sources in his distinction in the Treatise between proofs and probabilities and is further refined in e.g. “Of Miracles” and Dialogues Concerning Natural Religion. (For some of what is contained in the latter regarding the matter, see Nelson Pike's commentary to his edition of the Dialogues (Bobbs-Merrill: 1970), pp. 129, ff.) That I need to prove my convictions with some rather extended remarks was made clear to me by a number of people, but especially by Pall S. Ardal, the commentator on my paper at the APA meetings.

¹⁷ For Hume's remarks on the difference between his view and Descartes's see E149–50 and A Letter from a Gentleman to his friend in Edinburgh, op. cit., p. 19.