Steven Sverdlik

This is a presentation of the utilitarian approach to punishment. It is meant for students. A note added in July, 2022 advises the reader about the author's current views on some topics in the paper. The first section discusses Bentham's psychological hedonism. The second briefly criticizes it. The third section explains abstractly how utilitarianism would determine of the right amount of punishment. The fourth section applies the theory to some cases, and brings out how utilitarianism could favor punishments more or less severe than the lex talionis.

The main previous analyses of punishment by Hart, Feinberg and Wasserstrom are considered and criticized. One persistent fault is the neglect of the idea that in punishment the person subjected to it is represented as having no valid excuse for wrongdoing. A new analysis is proposed which attempts to specify in what sense punishment by its very nature is retributive, as Wasserstrom has asserted. Certain problematic cases such as strict liability offenses and pre-trial detention are considered in light of the new analysis.

Retributivist approaches to the philosophy of punishment are usually based on certain claims related to moral desert. I focus on one such principle:Censuring Principle : There is a moral reason to censure guilty wrongdoers aversively.Principles like CP are often supported by the construction of examples similar to Kant’s ‘desert island’. These are meant to show that there is a reason for state officials to punish deserving wrongdoers, even if none of the familiar goals of punishment, such as deterrence, will be achieved. When suitable variants of such examples are presented, however, it is evident that there cannot be much reason to punish such wrongdoers, even if there is some. The same problem besets claims that there is intrinsic value in the suffering of wrongdoers, or that wrongdoers deserve to suffer. All such claims are relatively weak normatively.

Contemporary Kantians who defend Kant''s view of the superiority of the sense of duty as a form of motivation appeal to various ideas. Some say, if only implicitly, that the sense of duty is always ``available'''' to an agent, when she has a moral obligation. Some, like Barbara Herman, say that the sense of duty provides a ``nonaccidental'''' connection between an agent''s motivation and the act''s rightness. In this paper I show that the ``availability'''' and ``nonaccidentalness'''' arguments are in tension with one another. And the ``availability'''' idea, although certainly supported by some passages in Kant himself, is also clearly denied in other passages. My conclusion is that Kantians will need to abandon either availability or nonaccidentalness if they wish to have a consistent set of views about the sense of duty.

It is often argued that Kantian and consequentialist approaches to the philosophy of punishment differ on the question of whether using punishment to achieve deterrence is morally acceptable. I show that this is false: both theories judge it to be acceptable. Showing this requires attention to what the Formula of Humanity in Kant requires agents to do. If we use the correct interpretation of this formula we can also see that an anti-consequentialist moral principle used by Victor Tadros to criticize consequentialism is implausible. I go on to examine the version of John Rawls' theory that is used by Sharon Dolovich to develop a Kantian theory of legal punishment. This makes clear why punishment to achieve deterrence in the 'circumstances of justice' is morally acceptable. However, in at least one respect consequentialism gives us a more convincing understanding of the limits on the pursuit of deterrence than the Kantian theory does.