1 Of course some philosophers try to synthesize sceptical and naturalistic readings of Hume. But even in these situations one aspect is still considered more important than the other. For example, some claim that while Hume was primarily a naturalist, he employed scepticism to support his naturalism; clearly, though, this is still on the whole a naturalist interpretation. So although there are nuanced layers of interpreting Hume as a naturalist or a sceptic, there is still a substantial gap between the two camps.

2 We should note that some philosophers have begun to explore the internalism/ externalism issue with respect to Hume's ethics. See, for instance, Charlotte Brown, “Is Hume an Internalist?” Journal of the History of Philosophy, 26 (January 1988): 69–87.CrossRefGoogle Scholar

3 Fumerton, Richard, “The Internalism/Externalism Controversy,” in Philosophical Perspectives, Vol. 2:Google ScholarEpistemology, edited by Tomberlin, James (Atascadero, CA: Ridgeview, 1988), p. 443.Google Scholar

4 For a full defence of this characterization of internalism, see Kim, Kihyeon, “Internalism and Externalism in Epistemology,” American Philosophical Quarterly, 30 (1993): 303–16.Google Scholar Another recent interpretation of the internalist/ externalist distinction that is in accord with my initial characterization appears in Senor, Thomas D., “The Prima/Ultima Facie Justification Distinction in Epistemology,” Philosophy and Phenomenological Research, 56 (September 1996): 560–62.CrossRefGoogle Scholar

5 BonJour, Laurence, “Externalism/Internalism,” in A Companion to Epistemology, edited by Dancy, Jonathan and Sosa, Ernest (Cambridge, MA: Basil Blackwell, 1994), p. 132.Google Scholar

6 For a recent discussion of these issues, see Garrett, Don, Cognition and Commitment in Hume's Philosophy (New York: Oxford University Press, 1997), chaps. 4–5.Google Scholar

7 Hume, David, An Enquiry Concerning Human Understanding, edited by Selby-Bigge, L. A., 3rd ed., rev., edited by Nidditch, P. H. (Oxford: Clarendon Press, 1975), p. 27Google Scholar (hereafter cited as, e.g., Enquiry, p. 27).

8 Ibid., p. 63.

9 Ibid., pp. 73-74.

10 Ibid., p. 153.

11 Hume spells out this method in ibid., pp. 61-63. This method surfaces also in his discussion in A Treatise of Human Nature, edited by Selby-Bigge, L. A., 2nd ed., rev., edited by Nidditch, P. H. (Hong Kong: Oxford University Press, 1990), p. 155Google Scholar (hereafter cited as, e.g., Treatise, p. 155):

“Having thus explain'd the manner, in which we reason beyond our immediate impressions, and conclude that such particular causes must have particular effects; we must now return upon our footsteps to examine the question, which first occur'd to us, and which we dropt in our way, viz. What is our idea of necessity, when we say that two objects are necessarily connected together. Upon this head I repeat what I have occasion to observe, that as we have no idea, that is not deriv'd from an impression, we must find some impression, that gives rise to this idea of necessity, if we assert we really have such an idea” (Hume's emphasis).

12 For an interesting attempt to read Hume's thoughts on testimony as indicative of an externalist epistemology, see Saul Traiger, “Humean Testimony,” Pacific Philosophical Quarterly, 74 (June 1993): 135–49CrossRefGoogle Scholar, and Traiger, Saul, “Beyond our Senses,” Hume Studies, 20 (1994): 241–60.Google Scholar

13 Pollock, John, Contemporary Theories of Knowledge (Savage, MD: Rowman & Littlefield Publishers, 1986), pp. 23–24.Google Scholar

14 Such a characterization is not unprecedented in the epistemological literature. This is how Alvin Goldman depicts process types that are assessed for reliability in his “What Is Justified Belief?” in Justification and Knowledge, edited by Pappas, George (Dordrecht: D. Reidl, 1979), p. 11.CrossRefGoogle Scholar In addition, this is how John Pollock portrays internalism and externalism in Contemporary Theories of Knowledge, pp. 21-23.

15 “RI” stands for Radical Internalism.

16 By “internal,” I simply mean that which is introspectively accessible to the cognizer.

17 “RE” stands for Radical Externalism.

18 By “external,” I simply mean that which is not introspectively accessible to the cognizer.

19 “PI” stands for Partial Internalism.

20 “PE” stands for Partial Externalism.

21 Interestingly enough, this is roughly how John Pollock (Contemporary Theories of Knowledge) characterizes the essential feature of externalism: “Externalism is the denial of internalism. According to externalism, more than just the internal states of the believer enter into the justification of beliefs” (p. 23).

1 Chisholm, Roderick, Theory of Knowledge, 3rd ed. (Englewood Cliffs, NJ: Prentice Hall, 1989), p. 84.Google Scholar

23 This “x” is omitted in the text, doubtless due to a typographical error.

24 Chisholm, Roderick, “A Version of Foundationalism,” in Midwest Studies in Philosophy, Vol. 5, edited by French, Peter A., Uehling, Theodore E. Jr., and Wettstein, Howard K. (Minneapolis: University of Minnesota Press, 1980), p. 556.Google Scholar

25 Ibid. For a similar comment, see p. 558.

26 Another epistemologist who has recently affirmed the compatibility of externalism and internalism is Keith Lehrer. In Lehrer's own words:

“The dispute between externalism and internalism is not so clear as the de”fenders of each position think…. Enlightened internalists, those maintaining the justification of a belief can be completely determined by reflection, Chisholm, for example, are left with the need to deal with the Gettier problem….

In so doing, they naturally require that justification not be connected with falsity, and this requirement imposes an external condition. The purest internalists who bar external conditions from entering into an account of justification find externalism creeping in through Gettier's back door. Enlightened externalists, Goldman, for example, find it necessary to require that a person not believe anything which would undermine beliefs arising from external processes, thus introducing reference to background beliefs, that is, to internal conditions. As a rough and ready aid to the classification of theories, I have no objection to the distinction between externalism and internalism, but the dispute over the matter is more useful for igniting dialectic passion than for reaching truth in the matter. Knowledge arises when there is the appropriate sort of match between all of what a person believes and external reality. The object of epistemology is to find the right mix of internal factors and external relationships to explicate what is required” (Lehrer, Keith, “Knowledge Reconsidered,” in Knowledge and Skepticism, edited by Clay, Marjorie and Lehrer, Keith [Boulder, CO: Westview Press, 1989], pp. 131–32).Google Scholar

For more on this general theme, see Sennett, James, “Towards a Compatibilist Theory of Internalist and Externalist Epistemologies,” Philosophy and Phenomenological Research, 52 (1992): 641–55.CrossRefGoogle Scholar

27 Dauer, Frances W., “Hume's Skeptical Solution and the Causal Theory of Knowledge,” The Philosophical Review, 89 (July 1980): 357–78.CrossRefGoogle Scholar

28 Goldman, Alvin, “A Causal Theory of Knowing,” The Journal of Philosophy, 64 (1967): 357–72.CrossRefGoogle Scholar

29 Dauer, “Hume's Skeptical Solution and the Causal Theory of Knowledge,” p. 369. Dauer also does not deny that Hume offers such an account.

30 We should note that Dauer has recently reiterated that his reading is a possible externalist solution to Hume's scepticism. See his “Hume's Scepticism with Regard to Reason: A Reconsideration,” Hume Studies, 22 (November 1996): 229.Google Scholar

31 Enquiry, p. 110.

32 David Hume, Dialogues Concerning Natural Religion, Part XI. I will assume that even if Philo does not “speak for” Hume, this is an epistemic principle that Hume would accept.

33 We should note that Pollock's portrayal of the internalist/externalist divide is slightly different from most philosophers. For he does not define this distinction epistemically; rather he proposes “to define internal states as those states that are directly accessible to the mechanisms in our central nervous system that direct our reasoning. … It seems clear that directly accessible properties must be in some sense “psychological,” but I doubt that we can say much more than that from the comfort of our armchairs. What properties are directly accessible is an empirical question to be answered by psychologists” (Contemporary Theories of Knowledge, pp. 134-35).

This characterization notwithstanding, Pollock spends considerable time criticizing those theories that would be considered externalist even on more epistemic readings of the internalist/externalist distinction. In addition, he brands himself and Chisholm as “paradigmatic internalists” (Contemporary Theories of Knowledge, p. 141). So the label “naturalistic internalist” still seems sufficiently apt when applied to Pollock.

34 Ibid., p. 168.

35 Ibid.

36 Treatise, pp. 181-82. For a technical reconstruction of this argument, see Wilson, Fred, “Hume's Skeptical Argument against Reason,” Hume Studies, 9 (November 1983): 90–129.CrossRefGoogle Scholar A sophisticated attempt to show how the argument of this section fits in with the naturalist view of Hume appears in Morris, William Edward, “Hume's Scepticism about Reason,” Hume Studies, 15 (April 1989): 39–60.CrossRefGoogle Scholar For a sceptical reading of this argument, see my “Hume's Iterative Probability Argument: A Pernicious Reductio,” Journal of the History of Philosophy, 38, 2 (April 2000): 221–38.CrossRefGoogle Scholar

37 Treatise, p. 183.

38 Pollock, , Contemporary Theories of Knowledge, p. 168.Google Scholar

39 Ibid, p. 162.

40 Ibid, p. 161.

41 Pollock himself seems to acknowledge this. For elsewhere he says: “Reasoning is not, strictly speaking, an action, but it is something we do, and we do it by doing other simpler things. We reason by adopting new beliefs and rejecting old beliefs under a variety of circumstances.” (This quotation is from Pollock's “Epistemic Norms,” Synthese, 71 [1987]: 75.)Google Scholar Because these “quick and dirty” systems allow us to embrace new beliefs and rebuff old ones (albeit without time to deliberate), they presumably would count as a type of reasoning.

42 Goldman, Alvin, “Naturalistic Epistemology and Reliabilism,” in Midwest Studies in Philosophy, Vol. 19, edited by French, Peter A., Uehling, Theodore Jr., and Wettstein, Howard (Notre Dame: University of Notre Dame Press, 1994), pp. 304–305.Google Scholar

43 Or, more precisely, any true belief that is not susceptible to some kind of Gettier example counts as knowledge because all beliefs are justified

44 Enquiry, p. 110.1 believe that I could make the same point if we considered the norms that Hume discusses in “Of Unphilosophical Probability” (Treatise, I, iii, 13) and “Rules by which to Judge of Causes and Effects” (Treatise, I, iii, 15). But a consideration of these sections would take us too far afield.

45 Kitcher, Philip, “The Naturalists Return,” The Philosophical Review, 101 (January 1992): 59–60.CrossRefGoogle Scholar

46 Goldman, Alvin, Epistemology and Cognition (Cambridge, MA: Harvard University Press, 1986), p. 378.Google Scholar

47 Ibid, p. 6.

48 Goldman, Alvin, “Strong and Weak Justification,” in Philosophical Perspectives, Vol. 2:Google ScholarEpistemology, edited by Tomberlin, James (Atascadero, CA: Ridgeview, 1988), p. 54.Google Scholar For more discussion of this non-undermining clause, see Goldman, , Epistemology and Cognition, pp. 62–63 and 111-12.Google Scholar

49 Goldman, Epistemology and Cognition, p. 89.

50 Ibid., p. 92.

51 For more details of how Goldman's approach takes a direction different from that of Hume's, see ibid., pp. 97ff.

52 Baier, Annette [A Progress of Sentiments (Cambridge and London: Harvard University Press, 1991)Google Scholar] has recently argued that for Hume those norms that we accept after reflection are the “correct” norms. More specifically, Baier maintains that “[t]he whole of the Treatise searches for mental operations that can bear their own scrutiny, sorting those that can … from those that get into 'manifest contradictions' or self-destructive conflict when turned on themselves. The whole enterprise is a search for norms with the sort of grounding that a reflective naturalist can accept” (p. 97). Gill, Now Michael B. (“A Philosopher in His Closet: Reflexivity and Justification in Hume's Moral Theory,” Canadian Journal of Philosophy, 26 [June 1996]: 231–56)CrossRefGoogle Scholar and Singer, Ira (“Hume's Extreme Skepticism in Treatise I, IV 7,” Canadian Journal of Philosophy, 25 [1995]: 595–622)CrossRefGoogle Scholar have plausibly argued that Baier's views about Hume cannot withstand close textual scrutiny. But suppose that Baier is correct. It is still difficult to see how Baier's suggestion helps with a theory of epistemic justification—where epistemic justification helps convert true belief into knowledge. In other words, why should a belief formed while following a norm that a reflective naturalist can accept have any positive epistemic status? We simply need more information before we can evaluate Baier's theory given the context of this paper. Now perhaps Baier is on the right track; but so far as I can tell, until further information is forthcoming, it is not sufficiently detailed as a theory of epistemic justification to provide a “naturalistic internalism” reading of Hume.

53 Earlier versions of this paper were read at the 23^rd Hume Conference and the 1997 Pacific Division Meeting of the American Philosophical Association. I am grateful to my commentators (Rupert Read and Kenneth Merrill) and to the members of the audiences for their helpful feedback. Thanks are also due to Michael Bergmann, Philip L. Quinn, and Leopold Stubenburg for their critical comments, to Saul Traiger for a conversation that prompted me to explore these issues, and to the Vrije Universiteit in Amsterdam for a research fellowship that allowed me time to work on this paper. Finally, I would like to thank two anonymous referees for their probing suggestions and criticisms.