Home Page
Conferences / Working Parties


Response to Neil Sinclair's review of my book Hume, Reason and Morality: A Legacy of Contradiction? in Mind Vol.116, July 2007

The first question takes off from the ‘practicality’ argument formulated by Hume in Treatise 3.1.1, and the troublesome ambiguities surrounding its second premise. I proposed that, in order to resolve them, we should take Hume, when he stresses the utter motivational impotence of reason, to have in mind the moral rationalists’ notion . Sinclair agrees (734) that Hume both “has rationalist opponents in mind in 3.1.1” and “objects to their view that ‘eternal fitnesses’ can influence the will (3.1.1:23)”. But he rejects that proposal on grounds that “at no point in this discussion does Hume even hint at using ‘reason’ to refer to [their] fitnesses”. The proposal is not to the effect that Hume explicitly announced such a use of ‘reason’ - nothing so deliberate or systematic on Hume‘s part. But, quite independently of the proposal’s merits, the implication of Sinclair’s claim, that Hume never did on any occasion use ‘reason’ to refer to the rationalists’ notion, and that we are never entitled to substitute ‘reason’ for ‘fitnesses’, is contentious.

Given that the rationalists are the main target of 3.1.1, it would be astonishing if Hume never used ‘reason’ in this section to refer to their concept. Indeed we need only look as far as paragraph 4 to find an example. Here Hume expounds their position in terms of (Clarke‘s) “eternal fitnesses and unfitnesses of things” and (Cudworth’s) “immutable measures of right and wrong”. But he prefaces his exposition by making the general point that “according to [all these systems] virtue is nothing but a conformity with reason”. What else could ‘reason’ refer to here but his rationalist opponents’ concept?1

Sinclair will perhaps point out that he does not claim, or imply, that Hume never uses ‘reason’ to refer to his rationalists’ opponents’ reason, but rather that he never uses it to refer to their ‘fitnesses and unfitnesses’. But this would suggest that Hume drew a sharp distinction between the moral rationalists’ concept of reason, or rationality, and their concept of ‘fitnesses and unfitnesses’. But this is not true. Hume is quite explicit that (Clarke’s) “eternal immutable fitnesses and unfitnesses of things” and (Cudworth’s) “eternal measures of good and evil”, though they are relations existing outside us, are rational . Thus he does not content himself with saying (para.4) that these “eternal measures” are the same to “every rational being that considers them”, which would be consistent with reason as merely a mental activity of ours. He speaks of them (para.23) as themselves “eternal rational measures of right and wrong” (my italics). In this way, as any great philosopher should, he does justice to Clarke’s and Cudworth’s repeated identification of ‘reason’ with ‘the reason of things’2 or ‘the reason of the world’3 , at the same time as utterly repudiating their view.

Sinclair might now suggest that his own phrase “in this discussion” refers only to paragraph 23 which he cites as the source of his quotation. Actually Hume neither refers to ‘eternal fitnesses’ nor denies their influence on the will in that paragraph. But even if we concentrate on paragraph 22, where he does say these things, it is unclear how this could help. We have seen, first, that it is unwise, if not impossible, to try to prize apart the rationalists’ notion of reason from their notion of ‘the fitnesses and unfitnesses of things’, since for them ‘reason’ is essentially ‘the reason of the world’ or ‘the reason of things’. Secondly, there is no evidence that Hume was ever so foolish as to make this mistake. We are therefore entitled to take his rejection of their view that these ‘eternal fitnesses’ could influence the will as also a rejection of their view that reason could do so. But it is then hard to maintain that he would have regarded his famous descriptive phrase “perfectly inert” as inapplicable to the rationalists’ notion of reason. On the contrary, it must have seemed, from his point of view, highly applicable. The rationalists cannot, he might have said, posit such an inert concept of reason as the basis of anything so active as morality. We should not of course immediately jump to the conclusion that when Hume described reason as “perfectly inert” (3.1.1:8), the rationalist concept was, or even may have been, implicated. That demands much more argument which I will not repeat here.

The second question of interpretation relates to the ‘direction of fit’ account of desire and belief. Sinclair does not agree that this account presupposes that ‘it is a good thing for there to be a fit between desire and the world…..in every instance of desire ’. He comments (734): “It would certainly be a surprise to defenders of the direction-of-fit account that it implied such evaluative claims”. But these “evaluative claims” are merely paraphrases, or ways of drawing out the implications (since otherwise they are intolerably obscure) of Humberstone’s remark, in his classic discussion4 that, when there is this fit “We have a sense of things going right”, and Zangwill’s remark5 that when there isn’t this fit, “That’s a fault in the world”. It is a real question whether the ‘direction of fit’ account can be given without substantive use of evaluative language.


  1. Nor is this the first time Hume has used “reason” in a sense which he proposes to repudiate. To cite just one other example, at 2.3.3:1, he speaks of those systems of morality of which the chief distinguishing feature is “the supposed pre-eminence of reason over passion” , and goes on that in these systems “The eternity, invariableness and divine origin of the former have been displayed to the best advantage”. Clearly “the former” refers to “reason”.
  2. Clarke, Samuel (1728) A Discourse of Natural Religion, selections reprinted in D.D. Raphael (ed.) (1969) British Moralists 1650-1800, I: Hobbes-Gay: 231.
  3. Cudworth, Ralph (1731) A Treatise Concerning Eternal and Immutable Morality, reprinted in Raphael, ibid: 134.
  4. Humberstone, L.(1992) “Direction of fit”, Mind 101:72
  5. Zangwill, N.(1996) “Direction of fit and normative functionalism” Philosophical Studies 91:177.

SOPHIE BOTROS, Institute of Philosophy, School of Advanced Studies, London University.

Dr Botros can be contacted at sophie.botros@sophiebotros.com.

This page last updated: 6th November 2017.