Friday 13 April 2012

Can deontologists defend punishment?

27 years ago, Michael Philips set deontologists a challenge – to show how their principles could be reconciled with the institution of punishment. It’s an interesting paper, and I wonder if anyone has satisfactorily met that challenge yet.

A common objection to consequentialism is that it sometimes appears to justify punishing innocent people. This objection is vividly presented by H.J. McCloskey’s thought experiment, which invites us to imagine a scenario in which framing an innocent person for an atrocious crime will help to calm a furious mob and avert a riot. Utilitarians, fearing the possibility of death and destruction, are supposed to accept that punishing the innocent person might be the right thing to do. This apparent perversion of justice is often seen as grounds for rejecting consequentialism.

Yet as Philips points out, deontologists who accept the criminal justice system are in many ways in the same boat. It is a consequence of imperfect human institutions that the police and courts will make mistakes, and wrongly convict innocent people. This is more or less inevitable and predictable. A famous legal principle is that  it is “better that ten guilty persons escape than that one innocent suffer.” But how can it be justified for even a single innocent to suffer? The deontologists who condemn consequentialism for countenancing the possibility of punishing the innocent are (presumably) willing to defend a justice system that does exactly that – punish the innocent.

The obvious line of defence for the deontologist is to invoke the doctrine of double effect – the punishment of the innocent is permissible because it is a foreseen but unintended consequence of the institution of punishment. I have a couple of problems with this response. First, the problem that what is taken to be foreseen or intended seems to depend greatly on how an act is described. Insofar as the deontologist supports a policy that they know will result in punishing innocent people, they can be said to intend the punishment of the innocent. Alternatively, the consequentialist’s punishment of an innocent person can be taken as a side effect of averting a riot. Second, if I were an innocent person wrongly convicted of a crime I don’t think that the intentions of the person or system that punished me would make me feel any less wronged.

Another possibility is that the wrongness of letting guilty people escape unpunished is bad enough to justify occasionally punishing innocent people – perhaps not 1:1, but maybe in the 10:1 ratio suggested above. I don’t see how anybody who held this position could still criticise consequentialists with a straight face. Effectively what they’d be saying is that death and injury of those caught up in McCloskey’s riots is not of sufficient moral weight to count against punishing the innocent, but that punishing people is. Causing harm is more morally significant than preventing harm.

Maybe there are other ways out of the dilemma, but if it is a real one, which side should deontologists choose? Philips suggests that they should give up their objection to punishing the innocent, but Michael Zimmerman uses a similar argument (alongside a number of others) to urge us to relinquish the notion that punishment can ever be morally justified. Are there other options I’ve missed?

2 comments:

  1. I've added a problem: http://liberalreflections.wordpress.com/2012/04/15/moral-absolutes-and-punishment/

    ReplyDelete
  2. J.S Mill saw that punishability was the essence of Freedom for Liberalism. De Maistre had already pointed out that punishment was stochastic and that if this wasn't the case deontics or doxastic systems would collapse into Utilitarian or Consequentialist systems.
    Kipling, whose Aunt was a Radical, has his District Commissioner Petit quote Caiaphas- 'it is expedient that one man die for the sake of the people'- i.e. this isn't some new aporia but stuff everybody knows about and which gets subsumed under the rubric of evidentiary decision theory.
    Contra your title- Social Problems are maths, are stats, are probabilistic game theory. What they aint is stupid sophomore pi-jaw.
    Fuck is wrong with you?

    ReplyDelete