Feinberg lays out three justifications for punishment: retributivism, utilitarianism, and vengeance. He explained how these views work to justify punishment and some of the major problems that each one faces.
Feinberg identifies retributivism with the following points:
The necessary condition is the least controversial of the bunch. But it’s not clear that any human institution of punishment is compatible with it since punishment always harms the innocent: think of family members of the person being punished.
There are significant questions about the sufficient condition. Most importantly, why should a society be committed to punish when doing so either does no good or costs the society more than it thinks it gains? Those being punished may deserve it, but why is that more important than all of the other things a society may care about?
The third point, about the amount of punishment to be given faces a variety of questions about how apportioning the punishment to the crime works.
The chief problem with the utilitarian view is that guilt is not a necessary condition of punishment.
There are also persistent questions about fairness. Two people may commit the same crimes but receive different treatment, depending on the circumstances their society finds itself in.
Wing made a very interesting point about deterrence. This is usually thought of as a strong point for utilitarian theories of punishment. But, as he observed, the retributivists are the ones who are more committed to seeing that every crime is punished. The utilitarians could neglect to punish for a variety of reasons. On the other hand, as Callum pointed out, utilitarians can make up for this by imposing punishments that are far out of proportion to the original crime: the deterrent effect would come less from the certainty of punishment than from its possible cost.
Retributivism looks at what those who receive punishment deserve. Vengeance looks at what those who inflict punishment get out of doing so. (And utilitarianism looks at the effects of punishment on everyone, including third parties.)
Feinberg takes the punisher’s perspective into account with the expressive theory that we will read next time. Murphy’s chapter in the textbook is also interesting.
As Joseph noticed, Feinberg presents “pure” versions of the retributive and utilitarian theories of punishment. One way of dealing with the chief problems of both would be to drop the sufficient conditions and combine the necessary conditions: punishment can only be used against the guilty and only when it is to the advantage of society.
This impure alternative isn’t going to solve everything. It doesn’t give us a sufficient condition: it doesn’t tell us when we should punish, just when we should not. And the utilitarian necessary condition is what gives rise to the problem of unfairness. Finally, it’s not obvious that there is a coherent story to be told about how these two elements could go together. That said, it’s better than the pure alternatives.