I agree with you.

1) Justice (i.e. people will be unhappy if a person who murders 100 people stays in jail undead.)
2) Deterring the criminal
3) Deterring others
4) Reforming the criminal
5) At the same time we shouldn’t dehumanize ourselves to accept torture, etc. as an appropriate punishment.
There might be others, but you get the point. There are some places where you just can’t avoid utilitarian decision-making.

Also taking an economic take on the “eye for an eye” notions of the law in general, since the loss of utility is essentially a subjective matter, it is difficult to assess the actual loss to the aggrieved party. This has been the cause of much debate over the extensive use of the law of torts in recent American history. It is difficult to say how accurately a group of disinterested people can assess how much the cost of a particular intangible loss is to a person, and also whether the loss to the criminal will be sufficiently large for it to qualify as a punishment or in fact be too large and unfair.

With specific regard to the death penalty again i think considering the fact that information may or may not be adequate it is better to have something like a life-imprisonment sentence which allows for a possibility of reversal in case some new information becomes available at a later stage, than to take the drastic and irreversible step of killing the person which may be a miscarriage of justice.

The Death Penalty also prompts the question “How most humanely to kill a person?”, which i find quite oxymoronic.

Considering the fallibility of human beings it is fairer that justice be a reversible process rather than a final one.

But since we are in the real world, the threat of punishment is an important tool for civil society though it’s objective is to overcome such threats to its constituents. Since we have to have a system based on the objectivisation of subjectivity we might as well have the one that has the least irreversible consequences.

Take for instance (not a very good instance) the Dreyfus affair, when the Jewish Officer was accused of spying for Germany in 1894, had he been killed immediatly after the first trial that would have been the end of the affair and the French government would have achieved a patriotic victory. But not many years later when it was found that he was infact the object of an anti-semitic conspiracy he was exonerated, and the instituitional bias against Jews was exposed. Though the effect was that he lost 5 years of his life. (He eventually went on to command one of the forts defending Paris in WWI and was later inducted in the Legion d’Honour).

The “humane” methods of killing are still quite oxymoronic in my view because the final consequence is murder, an irreversible condition caused by the wilfull enaction of force upon someone rendered defenceless. Whether there is social sanction (as in Sati or the Death Penalty) or only individual convictions (Honour Killings) and whether it is achieved by breaking the neck or slowly torturing the body it all leads to death.

Can you answer the question “Why not have death as punishment for petty theft?” without referring to fairness, i.e that the punishment should be somewhat proportionate to the crime?

If so, why not castration for rape and whipping as a punishment for assault? That’s roughly the line of reasoning I took. That’s where the multiple criteria thing came in.