Don’t weaken the conclusion!
But, it might sceptically be asked: do we need, or indeed can we ever have, a straight answer to this question? Certainly, a glance at the literature suggests that it has become acceptable for writers, attributing to Hume the ‘moderate’ version, to vacillate over its conclusion. Thus, while evidently wishing to endorse on his behalf the stronger conclusion (which, as we have seen, only follows on the ‘extreme’ version), they convert it to the weaker one at crucial junctures in their expositions. An example3 is Mackie (1980: 51). He first gives Hume’s conclusion as ‘morality is not based on reason’, and again as ‘moral distinctions are not derived from reason’, Drawing attention to the ambiguities surrounding Hume’s use of reason, he nevertheless goes so far as to entertain the possibility that it maybe meant to cover having any beliefs whatsoever. In this case, Hume’s conclusion could be ‘an explicitly nondescriptive (emotive or prescriptive) analysis of moral judgements’. A little later, however, (ibid.: 53), with a ‘moderate’ reading of premiss 2 clearly in view – ‘knowledge, beliefs and reasoning (of any kinds) alone do not influence action’ – he restates the conclusion. But now it has the following form: ‘the state of mind [expressed by a moral judgement] . . . does not consist wholly of knowledge, beliefs, and reasoning of any kinds’ (my italics). This is strictly compatible with moral judgements being just a combination of reason and desire, or even with them being essentially matters of reason.