Hume's Pivotal Argument, and His Supposed Obligation of Reason

Abstract

Treatise 1.4.1, "Of scepticism with regard to reason", argues that whenever we attempt to assess our probability of error, "we are oblig'd by our reason" to consider also the probability of error in that assessment, leading to a fatal regress whichbut for irresistible naturewould extinguish all belief. The argument plays a huge role in the Treatise, and has recently attracted many defenders, rejecting the previously standard objection that such iterated reflection need not imply reduction of probability. This paper, however, presses a more fundamental objection that there is no obligation of reason to iterate in the first placesomething obscured by the failure of previous analyses to focus on specific examples. Unlike the Treatise, Hume's Enquiry of 1748 is richly illustrated with examples, making it likely that he himself would have encountered these problems. And there are traces in the Enquiry of a fundamental change of view, corroborating the significance of this argument for his philosophical development. Hume's argument in Treatise 1.4.1, "Of scepticism with regard to reason", is one of the most important in the entire work, setting an extreme sceptical tone which persists into the Conclusion of Book 1, and delivering a result that appears to wreak havoc there. The significance Hume accords it in 1739 is also witnessed by the number of times that he references or alludes to it in other sections of the Treatise. 1 Yet it does not feature in his 1748 Enquiry concerning Human Understanding, and became in time (arguably) "the only major argument of Book I of the Treatise not repeated in his later writings". 2Accordingly, it was for many years neglected by scholars, and omitted from most books on Hume, which typically aspired to give a smooth picture of his epistemology that encompassed both the Treatise and the Enquiry. Other likely factors here were the argument's tension with the "naturalist" view of Hume popularised by Norman Kemp Smith and later Barry Stroud, and its mismatch with the conventional structure of topics (ideas, induction, causation, external world, personal identity, etc.) around which 1 Hume explicitly refers to this argument no fewer than three times in other sections of the Treatise (T 1.3.13.5 n. 24, 1.3.13.17 n. 26, and 1.4.7.7 n. 53), and twice implicitly (in the first and last paragraphs of the following section, at T 1.4.2.1 and 1.4.2.57). T 1.3.13.5 talks of it as exhibiting a "very memorable" point "which is of vast consequence" for the understanding, presumably alluding to its huge significance in the Conclusion of Book 1. 2 Immerwahr (1979, p. 234), from whom this quote is taken, draws the reasonable conclusion that this gives good "reason for thinking that Hume is dissatisfied with this argument". 2 many books on Hume have been organised. 3 This neglect, however, came to an end in the 1980s, with three authors -Fred Wilson, Richard DeWitt, and Ted Morrisall seeking to rehabilitate Hume's argument against what they considered to be an overwhelmingly dismissive orthodoxy, 4 and a f...