Monday, March 25, 2013

Are moral assertions merely expressions of emotion?

To: Chris M.
From: Geoffrey Klempner
Subject: Are moral assertions merely expressions of emotion?
Date: 23rd September 2009 11:01

Dear Christian,

Thank you for your email of 13 September with your one hour timed essay in response to the question taken from the 2008 the University of London 'Ethics Historical Perspectives' paper, 'Are moral assertions merely expressions of emotion?'

I like the way you have structured this essay, representing Ayer's and Stevenson's emotivism as a response to G.E. Moore's 'naturalistic fallacy'. As Moore presents the problem, in response to any factual statement, we can always raise the question, but is that good? So in the example you give, to say that action A leads to the 'greatest happiness for the greatest number' still leaves open the question whether we think that is a good thing or not.

We can recast Moore's argument in the following way: moral statements have motivational force. From the acceptance of a moral statement, action follows. The problem is explaining the link between the two. Any given belief entails action only on the assumption of a desire. Two people can have the same belief, e.g. 'War involves wanton killing' but different desires. One hates killing, the other enjoys it.

With emotivism, however, a third aspect emerges, besides belief and (first-order) desire, namely *attitude*. Not only do we do things that we desire, and avoid things that we don't desire; we also express our approval or disapproval of the desires and actions of others. Human beings are influenced by expressions of attitude. By expressing my attitude about war, e.g. that wanton killing is wrong, I (hope to) bring it about that others change their desires in relation to the subject of war.

According to emotivism, the motivational force of moral judgements is entirely accounted for by the expression of attitude/ emotion. This is subjectivist (because my moral judgements are contingent on my attitudes), but with an extra component. If I say, 'War is bad', I am not merely stating that I have a negative attitude towards war; I am trying to influence you. Hence, 'I don't like war' is not a moral judgement but 'war is bad' is a moral judgement.

So far, so good. You mention in your essay that there appears to be a problem with explaining how there can be 'rational discourse and analysis' in relation to moral judgements. You express agreement with Ayer and Stevenson's reply to this point, namely that the substance of the argument concerns the facts, in other words getting a clearer picture of the situation which calls for approval/ disapproval.

Here you could have gone further, and mentioned the objection raised by Peter Geach, based on Fregean semantics, namely that moral statements can occur as the antecedent of conditionals. If 'War is wrong' is merely an expression of emotion, then how do you account for the conditional statement, 'If war is wrong then it is wrong to threaten a country with war if they fail to agree to your demands'?

There has been much discussion of this point, which relates to another point which you do make regarding different conceptions of 'truth'. Arguably, the emotivist can help themself to a sufficiently 'minimalist' conception of truth as mere agreement without conceding that the statement, 'War is wrong' does, after all, have a 'factual' (= correspondence truth) component.

You also raise an objection regarding 'the excess of freedom that the basic expressivism implies: if morality is about expressing emotions -- how can we then reject e.g. Nazis? The Nazis probably do not feel bad about what they do.' However, this is not an objection to emotivism as such, but rather to any moral subjectivist theory. There is no final 'court of appeal' apart from subjective human attitudes.

To someone who is sceptical about the possibility of any 'objective' account, however, this objection falls on deaf ears. When I condemn the Nazis, I'm not interested in what they think about what they do. This isn't something on which one takes a vote. My attitude is my attitude. The fact is that human beings agree to a remarkable extent on the kinds of behaviour which are approved or disapproved of. That's all there is.

In citing Jackson and Petitt and also Horgan and Timmons, the point here seems to connect with another important line of argument which you could have mentioned, the idea of 'thick concepts'. Many supposedly 'descriptive' words contain an emotive or valuational component. To accept the use of the word, is to agree to the emotion which that use normally expresses. You are 'buying in' to the language game with that word. If you disagree with that particular evaluation, you need to find alternative language, which makes it clear that you don't accept the word in question.

You also offer an idea of your own, that a possible world type of semantics could be used to account for the 'minimal truth' of moral judgements, on the emotivist theory. 'It seems we could take a subset of possible worlds and call them 'moral model world for a subject S'. This would contain all actions that the subject is morally engaged with... The moral agent... wants the 'is-world' to overlap with its 'ought-world'.'

From your description, I'm not sure exactly how this is supposed to work. Recall what I said earlier about the difference between desires and attitudes. Emotivism is about the expression of attitude -- the attempt to influence the desires of others. It is not just that I have a picture of the world or worlds I would ideally 'desire' the actual world to 'overlap with'. I want others to share my view.

I can see how one might use imaginary worlds in this sense to account for second-order desires; I can approve or disapprove of my own first-order desires as well as the desires of others. For example, after reading Aristotle's 'Ethics' I decide I want to live the 'good life'. I can't do this so long as my desires are the way they are. So I form the second-order intention to bring it about that my desires change.

What I don't understand is how the possible world idea solves a 'semantic' problem in relation to the nature of ethical 'truth'. Maybe I'm just missing something obvious here -- by all means run the idea by me again.

All the best,