To: Mark S.
From: Geoffrey Klempner
Subject: Hume's claim that reason is the slave of the passions
Date: 12th February 2009 12:15

Dear Mark,

Thank you for your email of 2 February, with your essay for the University of London Ethics Historical Perspectives module, in response to the question, 'What does Hume mean by the claim that reason is the slave of the passions? What role does this notion play in his moral philosophy?'

I had a bit of a sense reading this closely argued piece of not being able to see the wood for the trees. This is partly because the essay reads like notes for a much longer essay. You could probably have written twice as much as you have written here (that's not an invitation!). Also (once again) I missed a bibliography.

One of the central issues is the role of belief for Hume. I don't think it can be correct to simply say, 'Hume regards beliefs as enlivened ideas that some situation or object holds pleasure or pain in store for the agent.' A belief is an enlivened idea which, as a result of enlivenment, acquires the same capacity to constrain our actions as our perceptions have. A belief that a lion is lurking in the bush and the perception of a lion produce the same behaviour -- conditional, of course, on what we want. (E.g. you are a lion tamer searching for the lion who escaped from the circus.)

How can belief, construed in these terms, ever cause an action? You refer to Hume's acknowledgement that our actions can be modified by our having prudential beliefs. Consider the following example: I may want an second whisky now, but I know that the breathalyser Police units have been very active in this area and the risk of being stopped on my way home from the bar is too great. However, the belief about the Police doesn't in itself motivate or create my desire to avoid having to take a breathalyser test when I am over the limit. If I didn't have this desire in the first place (say, I am a foreign diplomat confident of being able to avoid any charges for motoring offences), then the belief about the Police units would not deter me from having my second whisky.

Thomas Nagel in his book 'The Possibility of Altruism' argues that a hidden premise in prudential reasoning is the agent's 'belief in ones identity over time'. Assuming I care about failing a breathalyser test now, why should I care about failing a breathalyser test in two hours time? Or (a better example, since that seems a bit far fetched) if I care about avoiding a heart attack now, why should I care about avoiding a heart attack in ten years time? -- Because it will still be ME. It could be argued that this is one aspect of prudential reasoning which Hume simply doesn't consider. The apparatus of 'beliefs and passions' is not sufficient, in itself, to account for the possibility of prudential reasoning. Or, rather, it makes something look contingent (caring for what happens to one's future self) which is not contingent but part of what it is (in a constitutive sense) to be a 'rational agent'.

A starting point for any discussion of action (Humean or otherwise) is recognition that all action requires a combination of beliefs and desires. No desire on its own can motivate action without a relevant belief concerning how things stand in the world. I will not drink the glass of enticing amber liquid if I believe that it is brake fluid rather than whisky.

The crucial weakness in Hume's position is his assumption that passions are simply 'given'. This is the issue highlighted by Anscombe's criticism (the saucer of mud example). On the picture we get from Hume, any action requires a motivating passion, simply because we can imagine the very same state of affairs -- the same circumstances, the same agent -- minus the passion. This formula (illustrated, e.g. in the example of the merchant) is effective against any putative example of an action being motivated 'by reason'. As you state, for reason to play a motivating role it must be capable of 'formal causation', that is to say, determining the ends of action. Whereas for Hume, ends of action are always contingent on the passions.

There is a highly instructive critique of Hume's view of practical reasoning in a monograph by Richard Norman, 'Reasons for Action' (Blackwell) where Norman links Hume's assumption about the 'givenness' of passions to Wittgenstein's argument against the possibility of a private language. It is only through acquiring a shared language that we come to 'know' what our desires are. Such knowledge involves the capacity to give intelligible reasons (in Anscombe's sense). In other words, reason is intimately involved in the very notion of a 'passion', not merely the external instrument for calculating relations between ends and means.

I haven't said anything about your answer to the second part of the question. I think it is an error to look for any attempted link between the 'reason is slave to the passions' thesis and Hume's 'sentiment' theory, as opposed to any other subjective view (e.g. egoism, social conventions etc.) The essential consequence is Hume's denial of the possibility of deriving an 'ought' from an 'is'. To get an 'ought' you need a motivating desire. ''Tis not contrary to reason to prefer the destruction of the whole world to the scratching of my finger.'

All the best,

Geoffrey