The serious moral condemnation and legal penalization of blackmail has often been considered paradoxical. This practice, after all, often simply combines two fairly innocuous elements; asking for money or other favors, and threatening to do something which one is "allowed" to do. Michael Clark has recently argued that previous discussions of this issue were fundamentally mistaken, and that there is no paradox about blackmail. The relation between the two elements, Clark argues, brings forth something new, and thus there is nothing paradoxical about the fact that "in themselves" the elements which make up the practice of blackmail are permissible. I argue for the paradoxality of blackmail in a different way, which considers the practice as a whole, and is not based only on the permissibility of the elements of ordinary blackmail when taken separately.

On the assumption that we are able to justify the institution of punishment, when people may be punished? Christopher New has recently argued that, despite our intuitions to the contrary, there is in principle nothing which forbids punishment' before the offense has been committed, i.e., prepunishment'. The issue, he argues, is only epistemological. I explore the challenge presented by New, and argue that prepunishment is deeply ethically unacceptable. The problem with prepunishment derives, in the end, from the widely recognized need to respect persons and from the unacceptability of the punishment' of the innocent.

Widespread and deep intuitions about the basic content of any satisfactory theory for justifying punishment, together with some plausible empirical assumptions, are seen to yield two closely related paradoxes about justice and the severity of punishment. Considerations of desert point in the opposite direction than do considerations of deterrence with regard to the severity of punishment of the underprivileged. And this leads to a situation in which, if considerations of the desert of the underprivileged are taken seriously, the convicted from privileged backgrounds are to be more severely punished, unnecessarily'.

Anti-utilitarians have argued about the issue of utilitarianism and the 'punishment' of the innocent by presenting extreme and peripheral examples, allowing utilitarians to disclaim the reality of the problem. Both sides have shared the opinion that in daily life utilitarianism will not entail much (if any) 'punishment' of the innocent. I argue that this is a mistake. If we consider not specific detailed examples (or the total transformation of society) but the utilitarian attractions of a limited relaxation of the rigorous criteria within the current judicial framework, the opposition between utility and justice is clear.