1 Earlier versions of this paper were presented at the University of Pennsylvania, Swarthmore College, William Paterson College of New Jersey, and Wake Forest University; I owe thanks to the audiences at those places for useful discussions. I am grateful to John Carriero, Bob Fogelin, Ed Minar, Sam Scheffler, Barry Stroud, Mike Williams, and especially Janet Broughton, for helpful comments on and criticisms of various ancestral drafts.

2 Here I follow Fogelin, who argues that the Treatise supplies a ‘natural history of philosophy’ — a narrative of the phenomenological and causal sequences of philosophical states of mind. See Hume's Skepticism in the Treatise of Human Nature (London: Routledge & Kegan Paul 1985), Ch. 7. See also Fogelin's, discussions in ‘The Tendency of Hume's Skepticism,’ in Burnyeat, ed., The Skeptical Tradition (Berkeley: University of California Press 1983)Google Scholar; and in ‘Hume's Scepticism,’ in The Cambridge Companion to Hume, Norton, David Fate ed. (Cambridge: Cambridge University Press 1993).CrossRefGoogle Scholar

3 These two aspects of the text are matched by two roles for Hume. First, he is the subject of philosophical states of mind; second, he is the author writing about Hume the subject. I try to make clear, either by context or more directly, which of these roles the word ‘Hume’ picks out in different parts of this paper.

4 Reid, Inquiry and Essays (Indianapolis: library of Liberal Arts 1975), 8–9Google Scholar

5 For a paradigm case of this sort of interpretation, see Chapter 1 of Strawson, Skepticism and Naturalism: Some Varieties (New York: Columbia University Press 1985).CrossRefGoogle Scholar

6 All parenthetical page references in the text are to the Selby-Bigge, /Nidditch, edition of the Treatise (Oxford: Oxford University Press 1978).Google Scholar

7 See Treatise: ‘the memory, senses, and understanding are, therefore, all of them founded on the imagination, or the vivacity of our ideas’ (265).

8 Two readings of ‘seemingly’ are possible. On one reading, Hume foreshadows a naturalistic refutation of skepticism even as he gives his skeptical arguments. The imagination's tendency to enliven some ideas beyond others only seems trivial; rightly understood, the tendency is an original propensity of human nature, not subject to the skeptic's condemnation. Cf. Immerwahr, John ‘A Skeptic's Progress: Hume's Preference for the First Enquiry,’ in Norton, Capaldi, and Robison, eds., McGill Hume Studies (San Diego: Austin Hill Press 1979), 233Google Scholar. But if that's what Hume means by ‘seemingly,’ why doesn't this naturalistic argument against skepticism ever become explicit in I iv 7? Instead, the rest of I iv 7 confirms the negative epistemic assessment of the imagination (though of course it also emphasizes that this negative assessment can have no durable practical effect). Therefore, a different reading of ‘seemingly’ is in order. On this second reading, that word points to: (1) the fact that the judgment of triviality is provisional, awaits confirmation (which it receives); and (2) the discrepancy between the small role we feel such a propensity should play in determining our beliefs, and the large role it in fact plays.

9 Note that Hume's derogation of the imagination in I iv 7 has a precedent. In I iv 2, on 217, he says that the qualities of imagination that lead us to believe in continued and distinct existence of objects (a natural, indispensable belief on Hume’ s account) are ‘trivial.’ Still, it seems to me that Hume makes the charge of triviality abruptly in both contexts. And in both contexts Hume might well have said: the imagination yields indispensable belief x, so is ipso facto not trivial.

10 To clarify two important points: (1) I am not arguing that all naturalists are stuck with pessimism and skepticism, but only that a form of pessimism that makes skepticism psychologically possible, and perhaps also argumentatively plausible, has a non-arbitrary connection to naturalism. (2) I do not try here to settle the issue how much of Hume's extreme skepticism flows only from his naturalistic commitments as deformed by the ‘theory of ideas,’ and how much of it flows from other features of his naturalistic commitments. I assume that something other than the theory of ideas does some work in driving Hume to skepticism - partly because, of the three skeptical arguments Hume cites in I iv 7, only that broached in I iv 4 seems to me to depend irrevocably on the theory of ideas; and partly because Hume seems inclined to think that any ‘anatomizing’ can ‘trivialize.’ On this last matter see (a) Treatise III iii 6, 620-1; and (b) Hume's letter of 17 September 1739, to Hutcheson, in Grieg, ed., The Letters of David Hume (Oxford: Oxford University Press 1932), 32-3.Google Scholar

11 I say a bit more on this score in my concluding section; but my aim there is not to spell out an understanding of the pessimistic shift, but to add some support to the claim that the shift is what we need to understand.

12 For an illuminating account of extreme skepticism as the natural causal product of reason, see Broughton, Janet ‘Hume's Skepticism about Causal Inferences,’ Pacific Philosophical Quarterly 64 (1983) 3–18CrossRefGoogle Scholar.

13 Hume presents his skeptical arguments as additive in this way, so that each destroys cognitive hopes that the previous one left intact. But we might well think that 'knowledge’ of ‘necessary connections’ wouldn't furnish any solace if the faculty from which all beliefs (including beliefs about necessary connections) flow were irremediably unreliable. That is, we might think that the argument against the imagination already leaves us as badly off as we can be. So, Hume's presentation of the skeptical arguments aside, perhaps they are really best conceived of as assaults on cognition from different directions.

14 Why is it ‘our aim in all our studies and reflections’ to discern real necessary connections? Why must our more modest causal discoveries be unsatisfactory, insubstantial, beside the point somehow? Is this opinion about necessary connections another instance of (or product of) a shift of attitude?

15 Hume's use of ‘reason’ is notoriously slippery, and there have been fierce interpretive debates about whether and how Hume can think of some beliefs as reasonable. In broad terms and as a matter of overall interpretation, one can claim that Hume is in the end concerned about whether beliefs are reasonable to hold, that is whether the beliefs are useful for us, pleasant for us, or even inevitable for us. Or one can claim that he is primarily concerned to show, contra some species of rationalism, that reason does not generate our crucial beliefs. Or one can claim that Hume argues that reason impugns the content of our central beliefs -even when, as is the case with beliefs concerning matters of demonstration, the beliefs are both reasonable to hold and produced by the faculty of reason. Or one can mix and match these interpretive claims. Here I only want to claim that, in his descent into extreme skepticism, Hume comes to think that reason condemns the content of his beliefs.

16 A reminder: I am conceiving of the extreme skeptic as believing only that she cannot (reasonably) believe anything else-or, at any rate, as not believing much more than that.

17 See Strawson's discussion of Hume and Wittgenstein in Skepticism and Naturalism: Some Varieties.

18 For instance, in Treatise I iv 1, 183, Hume writes of skepticism:

19 This, again, is intended as a sketch of Strawson's view.

20 Hume must go on, it seems, either to substantiate the privilege, or to weaken his claims against skepticism (to something like ‘nature breaks the force of my skeptical reflections, even though the reflections are entirely correct’). But he does not explicitly take either course, which is why I iv 7 is in the end so unsatisfactory.

21 Immerwahr takes this passage to be a denunciation of dogmatism, a recommendation of probabilism: abandon certainty and make do with probability ('A Skeptic's Progress,’ 229-30). But it is more natural to take ‘certainly’ to modify ‘are,’ than to take it to modify ‘reason or believe’ (as Immerwahr does). For one thing, a happy-go- lucky probabilism hardly seems ‘splenetic.’ Also, the extreme skepticism whose 'remains’ still affect Hume tells equally against dogmatism and probabilism. Finally, it sounds odd to say ‘Smith reasons certainly’ rather than to say ‘Smith is certain about the conclusions of his reasoning.’

22 Hume uses the term ‘mitigated scepticism’ in the Enquiry Concerning Human Understanding, section Xll, part iii; see the Selby-Bigge, /Nidditch, edition of the Enquiries (Oxford: Oxford University Press 1975), 161-2Google Scholar. He does not use the term in the Treatise. But though the term is absent from the Treatise, the philosophical position or state of mind is present, and described in some detail. So, for the sake of convenience, I employ the well-known Enquiry term to describe the Treatise text.

23 Richard Popkin argues that Hume advocates (as the most ‘natural’ position) an oscillation between dogmatism and skepticism, in ‘David Hume: His Pyrrhonism and His Critique of Pyrrhonism,’ reprinted in Chappell, V.C. ed., Hume: A Collection of Critical Essays (Garden City, NY: Doubleday 1966) 53–98CrossRefGoogle Scholar. Stroud, Barry criticizes Popkin in ‘Hume's Scepticism: Natural Instincts and Philosophical Reflection,’ Philosophical Topics 19 (1991) 271-91CrossRefGoogle Scholar. According to Stroud, Hume is recommending that we reflect on both our ‘dogmatic’ and ‘skeptical’ moments, and so bring ourselves to a new and distinctive state of mind, one that will be healthier and more pleasant than a mere oscillation (however ‘natural’) could be.

24 How, then, can this be a principle? How can the determination to let nature or expedience determine my beliefs make any difference, if these were bound in any case to determine my beliefs? This is a difficult problem for Hume. He might want to employ his distinction between two senses of ‘natural': one all-embracing, the other denoting health (see 225-6). Then the mitigated skeptic would be the person who, aware that beliefs are always determined by human nature, took care that her beliefs were determined by healthful natural principles. However, this line merely postpones the problem, for now we need to know what makes some principles healthful and others not.

25 What is it like to become less fervent about one's everyday beliefs, to hold them less strongly than before? One possibility is that the mitigated skeptic is less certain, more tentative, than the non-skeptic. The non-skeptic thinks that fire certainly warms, the extreme skeptic has no opinion about whether fire actually warms, and the mitigated skeptic thinks that fire probably warms. That is to say, the weakening of belief amounts to a shift from dogmatism to probabilism.

26 Perhaps the mitigated skeptic replies: ‘My view too has a claim to priority, because it is founded on wide experience. My reaction is caused by all that has preceded it, including both the plain person's beliefs and the skeptic's doubts.’ But this apparently amounts at best to practical, not to theoretical or rational, priority; the extreme skeptic seems to have the only good claim to this latter sort of priority.

27 This position does not blend in well with natural dogmatism. For the natural dogmatist is uneasy that her beliefs are based on nothing more or other than expedience, and so guilty about dismissing philosophy only because of its inexpedience. But the mitigated skeptic I envision is content that her beliefs are based only on expedience, and so also content to dismiss (provisionally and non-dogmatically) any beliefs or activities she finds inexpedient.

28 Baier, Annette discusses this passage in ‘Doing Without Moral Theory?’ in Postures of the Mind (Minneapolis, MN: University of Minnesota Press 1985), 237Google Scholar. She there reads the passage only as a plea for more solid common sense in philosophy. Surely it is that, but also more; Hume is poking fun at these ‘honest gentlemen,’ in effect congratulating them for having minds so tranquil and impoverished that they can 'rest, like those of beasts, in that narrow circle of objects, which are the subject of daily conversation and action.’ (Perhaps it is significant that Hume says these gentlemen are in England, not in Scotland?)

29 In other writings, Hume tries to use mitigated skepticism as a critical standpoint; he seems to want to say that the mitigated skeptic is on better cognitive ground than the non-skeptic. (See, for instance, the section on miracles in the first Enquiry, and the whole of the Dialogues.) But he runs into a problem: how can anybody be on better cognitive ground than anybody else, since no one has any reply to extreme skeptical arguments? It is deeply unsatisfying to say that rational criticism of superstition is impossible, but that we can nevertheless safeguard ourselves from superstition. Saying this leaves unanswered (and unanswerable) the question why we should prefer not to be superstitious.

30 ‘Most critical examination’ must be either ironic, or elliptical for ‘most critical examination possible when we hold fixed the fundamental beliefs of common life.’

31 Note the distinction between Hume's project in writing the Treatise, and the Treatise itself. I would not claim that the optimistic Introduction to the Treatise contains no irony; but I do claim that the Introduction reports the experience of straightforward hope and ambition.

32 This distinction suggests two interpretations of ‘preserving skepticism in all the affairs of life': (1) being skeptical at every moment, never believing wholeheartedly or straightforwardly. (2) Having one's life shaped by skepticism, but for that very reason having moments of wholehearted, straightforward belief.

33 ‘The Tendency of Hume's Skepticism,’ 404

34 Does mitigated skepticism resolve or just dodge the epistemological problems Hume has raised? Is it a coherent psychological state, fit to do the work he assigns it? As the text of this paper hints, I am inclined to pessimism about mitigated skepticism on both these scores. But these issues deserve separate and extended discussion. Here my aim has been to clarify Hume's final position in I iv 7, not to assess that position in detail.

35 A certain school of Hume interpretation will (correctly!) emphasize Hume’ s positive naturalistic program, and also the scorn he often heaps on extreme skepticism as an untenable, unlivable disposition, and argue that therefore Hume never takes extreme skepticism seriously. For him extreme skepticism must be at most a tool for overthrowing his philosophical opponents, or a dialectical moment to be transcended with alacrity. (That is, Hume is ‘skeptical’ in just the way that Descartes is 'skeptical’ in the Meditations.) This interpretation depends on Hume's being in substantial control of his philosophizing all along; but Hume's work strikes me as including elements that are in tension with each other, and a struggle to fit these elements together into a unified package. That is, even though he has a positive naturalistic program, Hume's reflections lead him to take extreme skepticism seriously, as a certain sort of real threat (despite the fact that it is an untenable, unlivable disposition). Then Hume must turn his energies to taming extreme skepticism, to somehow assimilating it into his more general positive program. But the dialectical contortions of Treatise I iv 7 indicate to me that Hume does not have a leash round the neck of extreme skepticism from the start.

36 In Hume (London: Routledge and Kegan Paul 1977); see especially 116-117, and 248-250.

37 Fogelin's tendency to speak about the constructive point or uses or upshot of extreme skepticism is most pronounced inHume's Skepticism (see especially Chapter 11 of that book). Fogelin's recent essay ‘Hume's Scepticism’ seems less concerned than was his earlier work to discern such constructive aspects of extreme skepticism. This may be just a matter of omission rather than of change in interpretation. Still, on p. 94 of ‘Hume's Scepticism,’ Fogelin says of nature's victory over Pyrrhonism: 'The irony is that the ways of nature, when revealed, hardly fill us with confidence or with a sense of human dignity.’ And if the experience of extreme skepticism makes us acutely aware of nature's warped workings, it is hard to see how extreme skepticism supports or coheres with naturalism, or is a likely cause of a stable and cheerful mitigated skeptical disposition.

38 I do not deny that naturalism and extreme skepticism are closely connected for Hume; but the connection is not one of mutual support. For Hume, naturalistic investigation leads inexorably to extreme skepticism, and extreme skepticism is disastrous for naturalistic investigation. It is practically disastrous, because it renders the skeptic (temporarily) incapable of such investigation; and it is theoretically disastrous, because it constitutes an entirely negative epistemic assessment of human cognition. (Naturalism is the thesis that we can explain human beings and human phenomena in purely natural terms. Naturalism is then a belief that we can adequately and reasonably explain certain things; and a successful naturalistic project would include a wide variety of beliefs, most of which would have to be reasonable and held as reasonable.)

39 Wittgenstein, Philosophical Investigations (New York: Macmillan 1968)Google Scholar, section 124 40 Dialogues Concerning Natural Religion, Smith, Norman Kemp ed. (New York: Macmillan Library of Liberal Arts, 1989), 154Google Scholar. (The speaker is Cleanthes, in Part III.)

41 There is, of course, a great deal of controversy on this point. For instance, Baier, Annette in A Progress of Sentiments: Reflections on Hume's Treatise (Cambridge, MA: Harvard University Press 1991)Google Scholar, argues that I iv 7 does refute extreme skepticism by means of a narrative reductio. Baier's case is, roughly, this: at the start of I iv 7, Hume adopts a rationalist persona. He is solipsistic (or at least isolated), rationalistic, and neurotically questing after truth. But these habits of thought lead Hume's rationalist persona straight to extreme skepticism. He therefore adopts new habits of thought: he turns clubbable, naturalist, and good-natured fallibilist.