General Rules and the Justification of Probable Belief in Hume's Treatise

Hume Studies Volume 27, Number 2, November 2001, pp. 247-277 General Rules and the Justification of Probable Belief in Hume's Treatise JACK C. LYONS By the conclusion of Book I of the Treatise, Hume faces something of a dilemma . Because of the skeptical arguments of part 4, he is "ready to reject all belief and reasoning, and can look upon no opinion even as more probable or likely than another" (T 268-9).l Yet on the other hand, he clearly does think that some methods of belief-formation are better than others. Five paragraphs after the passage just cited, he proclaims, "I make bold to recommend philosophy , and shall not scruple to give it the preference to superstition of every kind and denomination" (T 271). Although it is difficult to take much of what Hume says in part 4 (and especially section 7) at face value, it is clear that Hume is sincere in his endorsement of philosophy here. Book I, after all, is only the first of the three books of the Treatise, and the other two books begin just four (Selby-Bigge) pages after this endorsement. The skeptical arguments examined throughout the Treatise seem to indicate that we cannot show that many, if any, of our beliefs have a high probability of being true, and yet Hume wants to maintain a distinction between better and worse methods of belief-formation. On what could such a distinction be founded? Notoriously, there are several different themes in the Treatise that look like they might play some role; at various points in the Treatise, Hume mentions the involuntariness of belief, the pleasure derived from philosophy, the love of truth, the distinction between the more and less universal workings of the mind, and so forth. Not surprisingly, there is a good deal of debate among Hume's commentators as to which of these themes is actually at work Jack C. Lyons is Assistant Professor of Philosophy, University of Arkansas, Fayetteville, AR 72701, USA. e-mail: jclyons@uark.edu 248 Jack C. Lyons in Hume's epistemology.21 think that there is a way to resolve many of these disputes by distinguishing among various components of epistemic theories, or stages of epistemological theorizing. I will try to use these distinctions to find a place for several of these different themes in Hume. Space prohibits a treatment of all of Hume's epistemology. Intuition and demonstration, for example, receive relatively little attention in the Treatise, and I will completely ignore them here. Nor will I address Hume's later epistemological views. My concerns here will be restricted to the epistemic principles involved in Hume's theory of probable belief as defended in the Treatise. I. The Epistemological Framework It will be helpful to begin by laying out the general framework with which I hope to explicate Hume's view. An "epistemic norm," as I will use the term, is simply a principle licensing a certain method of belief formation or revision .3 Examples of some epistemic norms might be: "If you (justifiably) believe that|> and that Ï• implies q, then, ceteris paribus, it is permissible to believe that q" and "If your psychic says thatp, then it is permissible to believe that p." Presumably, in this case, the former is a correct norm and the latter is not. Yet this raises an important question: what are the criteria for epistemic norm correctness? That is, what is it that distinguishes the correct from the incorrect norms? Some possible answers to this question would be, "The correct norms are those that when followed produce the greatest ratio of true beliefs ," or "The correct norms (for me) are those that are endorsed by my culture." The criterion, if it is to be at all illuminating, should be framed in non-evaluative terms. We can think of the project of formulating a criterion of norm correctness as an effort to distinguish those epistemic norms of which we approve from those of which we disapprove. Seen in this light, however, it is quite another project to defend, in the sense of advocating, the use of norms that have whatever property the criterion specifies as belonging to the...