Yesterday I gave a talk within our department to fellow graduate students on the topic of impartiality in moral and political philosophy, and more specifically, on Bernard Williams’ famous ‘one thought too many objection’. I’ve been working on the subject on and off for a while, but this is the first time that I’ve collected my thoughts on the matter into a coherent(ish) whole. Below is a precis of my talk.
The classic problem raised against impartiality is that it clashes with our commonsense understanding of how we can, and should, act towards specific other people, most obvious in the case of friends and family. A crude version of impartial morality requires me to take the interests of my family and strangers to have equal importance, and forbids me from giving special preference to the needs of those close to me. This is intuitively wrong, and so any moral theory that implies such a conclusion must be wrong also. The most common response to this type of objection against impartialist morality is to make a distinction between first-order and second-order impartiality. Brian Barry's argument in Justice as Impartiality hinges on this distinction, in fact, he claims that once we have made it properly, the debate between partialists and impartialists will be neutralised. First-order impartiality refers to the level of action, whereas second-order impartiality refers to the level of principle. Barry argues that most critics of impartiality are directing their attacks at first-order impartiality, when in fact, most impartialists do not endorse this demanding form of impartiality in the first place.
Second-order impartiality demands that principles of morality are justified impartially. For a contractarian like Barry this is done by ensuring that principles cannot be reasonably rejected by any individual. Moral principles that are justified impartially will not necessarily require impartial action - for example, a principle requiring that parents favour their own children over the children of strangers could be justified impartially. What is different about partiality that is impartially justified at the level of principle is that is must hold equally for everyone - so if I think that I am right to favour my own children over those of others then I must recognise that other parents will also be right to favour their children over mine.
Once we accept the distinction between first and second-order impartiality the problems impartialists face do fade somewhat - second-order impartiality is much less demanding and conflicts less with our intuitively valuable personal relationships. However anti-impartialists are still critical. Probably the most famous criticism of impartialist approaches to ethics comes from Bernard Williams. Williams accepts that a second-order impartialist can explain why it is permissible for a husband to save his wife from drowning when faced with a choice between saving her and a stranger. But he argues that the attempt to justify his partiality by reference to a higher order principle is inappropriate, because all the justification that is needed is 'because she is my wife'. Any further justification, such as 'and in these situations it is permissible to save one's wife' is 'one thought too many'.
In the example of saving one’s wife then it seems that Williams is right that further justification is not needed, and that we think badly of someone who offered the further justification without prompting. But this doesn't mean that no justification is provided - by saying 'because she is my wife' the husband appeals to the obligations implicit within that relationship. If a passing alien who had no grasp of the concept of marriage enquired why the husband chose to save his wife over the other casualty then it wouldn't be out of place for him to offer the further justification that Williams wants to avoid. And, importantly, there are some cases in which it is at least an open question whether the husband would be justified in saving his wife over a stranger (if the stranger is someone extremely important to the survival of the human race, say). So we can examine the relationship of marriage and question how much partiality between spouses is legitimate.
However I think anti-impartialists like Williams would respond that this examination, and search of justification of partiality, is itself inappropriate. Natural sentiments and personal relationships are by nature not subject to demands for justification – and if we do so then we threaten the nature of these relationships.
It seems to me that this argument trades on confusion over what is being justified. Williams is right that love itself can’t be justified (and that we shouldn’t try), but this doesn’t mean that we shouldn’t justify it when love affects the distribution of goods and resources between people. The distribution of love doesn’t need justification, but the effect that love can have on the distribution of other goods does. When I fall in love with someone I don’t have reasons that are fully spelled out and it would be inappropriate to try and provide them. I don’t have to justify being in love with X rather than Y. But once I am in love with X rather than Y I do have to justify how I allow this to affect distribution of goods between me, X, and Y. If I want to spend all my money on luxuries for X whilst leaving Y to starve then I should justify this to Y, and if I can’t, then I am not free to do so.
It seems likely that ‘I am in love with X’ carries a certain justificatory weight, but also likely that this will be capable of being outweighed in some circumstances. It is certainly the case that I am not justified in favouring X in all circumstances just because I am in love with them – if I am a judge of a competition that X is taking part in then the fact that I am in love with X is not relevant at all. This is an obvious case, and other cases are obvious in the other direction. The one’s in the middle are the ones that we have to think about. So to bring the example of the husband saving the wife back in, he has to justify how he distributes his ability to save people between his wife and the stranger. Given that he can only save one person he is fully justified in saving his wife most of the time. But the fact that this is justified with reference to higher order impartial principles does not denigrate in any way the relationship between the husband and wife.
I have futher thoughts on how relevant the husband's motivation is, and whether it is problematic if his motivation and reasons are in conflict, but I will save those for another day.