To: Mark S.
From: Geoffrey Klempner
Subject: Hume's claim that justice is an artificial virtue
Date: 25th February 2009 11:49

Dear Mark,

Thank you for your email of 16 February, with your essay for the University of London Ethics Historical Perspectives module, in response to the question, 'Why does Hume describe justice as an artificial virtue? Is what he says defensible?'

I found this essay very clear and well-argued. (Thanks for the bibliography.) I am also genuinely gripped by the first of the five challenges which you raise against Hume's account of justice as an artificial virtue.

In terms of strategy, in an exam when faced with a question like, 'Is so-and-so's account defensible?' I would tend to concentrate effort on the objections which I consider problematic, the ones which one can say something interesting (and possibly original) about rather than those against which the account or theory may be relatively easily defended. This isn't in any way a criticism of the essay: the essay is enhanced by the fact that you do canvas various objections, and this is genuinely instructive. However, given what you say about the first objection, this is where the real 'meat' of the essay lies, and it deserves more space.

Why am I gripped? This isn't just a question of the plausibility of claiming that a sense of justice is primitive rather than socially derived. It is plausible. As you say, 'Whether most humans would act like sensible knaves in dealing with strangers in a pre-social state is a moot point that cannot be answered by research now that all the world's inhabitants have been socialised.' But, then, how *is* this kind of question to be decided?

Notwithstanding Hume's claim to be giving a 'theory of human nature', his theory of ethics is not based merely on empirical observation. Having ruled out any possibility that ethics can be derived from reason, he is faced with the challenge of giving an alternative account, based on the *minimum* that one needs to assume about human nature. This assumption is, in a sense 'a priori' in that (in putatively Kantian terms) it is a necessary condition for the possibility of ethical discourse. Given the primitive motivations of self-love and benevolence, the complex apparatus of our system of ethical judgements may be logically derived, through philosophical analysis.

One can indeed challenge Hume on the need for benevolence: there are systems in which every virtue is 'artificial' in Hume's sense. Everything is ultimately explained by self-love.

The problem is clear: there doesn't seem to be any clear methodology, available to the moral subjectivist, for deciding between the various rival accounts: pure self-love, or self-love plus benevolence, or self-love plus benevolence plus justice, etc. etc.

The reason why there is ultimately no room for a Kantian-style 'transcendental argument' here is that this goes against Kant's strict rule: 'I have prescribed to myself the maxim, that in this kind of investigation it is in no wise permissible to hold opinions. Everything, therefore, which bears any manner of resemblance to an hypothesis is to be treated as contraband; it is not to be put up for sale even at the lowest price, but forthwith confiscated, immediately upon detection.'

(Sorry, I don't have the page reference to hand. See http://www.follydiddledah.com/image_and_quote_15.html)

It seems to me that this is the fatal weakness in Hume's account, and indeed in any similar attempt to reconstruct ethics on a subjective basis. What we get is not a philosophical theory concerning the foundations of ethics, but merely a more or less plausible (although ultimately untestable) psychological theory of human motivation.

There are two possible responses to this: The first is to simply admit that one's preferred theory has an ineradicable empirical element. So what? It is the 'best explanation', given the unacceptability of an alternative, and given the difficulty, or impossibility of gathering sufficient empirical evidence for or against the theory.

The second response is to attempt to give something like an a priori account of the genesis of the 'self' or 'person' -- the genesis of self-consciousness -- as a phenomenon which necessarily comes into being only through social interaction. I am thinking of the kind of account you would find in Hegel's 'Phenomenology of Mind' (e.g. the famous discussion of 'master and slave'). This is a route one might take if one were attempting to defend the view that a sense of justice is primitive. The problem with this is that, if successful, the resulting theory would amount to the very thing that Hume rejects: a rational basis for ethics.

I remember seeing a brilliant early 80's play by Tom Stoppard, 'Professional Foul' in which one of the characters, a persecuted dissident Czech philosophy student, has written a thesis arguing that (in Hume's terms) a sense of justice is primitive, and therefore not to be determined -- as the Communist party line claimed -- by considerations of what best serves the interests of the state.

All the best,

Geoffrey