Tuesday, February 28, 2012

Free will problem and the justification for punishment

To: David F.
From: Geoffrey Klempner
Subject: Free will problem and the justification for punishment
Date: 26 March 2007 08:55

Dear David,

Thank you for your email of 12 March, with your first essay for the Possible World Machine, in response to the question, 'In the light of the critique of 'free will', can blame and punishment ever be rationally justified? Consider hard cases, such as brainwashing, crimes of passion, the influence of drugs, medical or psychological conditions etc.'

This is a good attempt to deal with a difficult question. You have distinguished different purposes that punishment might serve under 2. (a)-(d), with the addition of reparation which you later consider as additional option - although I don't see that this would necessarily be an alternative to retribution. Punishment which served the purpose of compensating the victim of a crime could be carried out in the context of a 'no blame' (as you put it) view of punishment.

I am less clear about the consequences that you draw from the distinction between B. which is basically the view of 'compatibilism', that free will is compatible with determinism, and C. the view that given determinism our actions are not free, full stop.

Any system of punishment must recognize the difference between actions which are immediately coerced - the case you give is the gun against the head - and those which are not immediately coerced. If we think of punishment purely in instrumental terms, society does not have to be protected against individuals who do bad actions only when a gun is put against their head, nor would individuals in such a situation be deterred from doing the same thing in similar circumstances.

There is a problem, however, with a purely instrumental view which might not have occurred to you. This was put vividly by F.H. Bradley in his book 'Ethical Studies', where he gives the example of the Master of Hounds who gives his dogs 'a good thrashing' before they go out on a hunt 'just to show who's boss'. Pre-emptive punishment might be very effective indeed at not only preventing offences that might otherwise have occurred but also as a deterrent to others. Yet, only a very small minority would embrace this view. It is not fair to punish someone for something that they might or might not do in the future. But why insist on fairness if our only concept of punishment is instrumental?

But what about blame? In his British Academy lecture, 'Freedom and Resentment' (reprinted in 'Freedom and Resentment and Other Essays') P.F. Strawson argues that the practice of blame is part of our 'reactive attitudes' which are necessary for viewing one another as persons rather than things. This is consistent, Strawson argues, with a compatibilist view. Margaret, in dialogue 2 of the second unit starts from Strawson's position and attempts to explain why we 'argue against' something that someone has already done, and what that could possibly mean.

A purely instrumental view of punishment gives up all talk of blame. But as we have seen, it is not so easy to adopt this view consistently.

The 'hard cases' arise whether one adopts view B. or view C. Either way, we have to find a principled way of deciding whether or not to punish, or whether in some circumstances punishment is to be reduced and by how much. This is assuming, of course, that we accept that punishment must be meted out 'fairly'.

I find this issue very difficult. Patty Hearst, the erstwhile heiress turned bank robber is a famous case where a jury convicted someone who had been brainwashed because they were convinced by the argument that they were dealing with the person standing before them in the dock, the unrepentant member of the 'Symbionnese Liberation Army' - who had originally kidnapped her. The same principle applies to your Muslim extremists.

If someone takes drugs which lead them to murder, then it was their fault that they took the drugs but this is a far lesser offence (unless of course it is widely known that these particular drugs lead one to commit murder). In the famous 'Twinkie defence' in the USA (where else) a man was cleared of murder on the grounds that the packet of Twinkies he had eaten shortly before led to a rapid increase in blood sugar which caused his violent behaviour. No-one can be blamed for eating a packet of biscuits. Subsequent attempts to use this defence have apparently failed, however.

The most difficult issue to deal with, however, is the clear correlation between childhood deprivation or abuse and criminal behaviour. We don't consider these mitigating circumstances, on the principle cited above - that we are dealing with the person as they are now. Yet it still offends our sense of 'fairness'. Thomas Nagel has coined a term for this, 'moral luck'. It is also moral luck which leads one drunk motorist to lose six points after being breathalysed while another equally drunk, equally irresponsible motorist kills a child and ends up in prison. No-one simply gets the punishment they 'deserve' because accidental circumstances contribute to the outcome of an agent's intention.

All the best,

Geoffrey