Feinberg’s question is what distinguishes punishment from mere penalties. His answer is that punishment has an expressive function that penalties lack.
As evidence, he gives several cases of punishment that fit neither the utilitarian nor the retributive theories. In one kind of case, for example, it matters who inflicts punishment beyond what the punishment itself involves. That is not covered by either of the two theories.
Feinberg then uses the expressive theory to smoke out sneaky attempts to punish people while avoiding the normal legal constraints on doing so, to explain why strict liability is tolerated when fines are the punishments but not when the punishment involves imprisonment, and to solve a problem with the justification of punishment that the retributive theory fails to solve.
We talked a bit about whether Feinberg’s theory could be useful in discussing the confiscation of property in drug cases, known as asset forfeiture. What happens is that law enforcement agencies confiscate property of people accused of violating drug laws and don’t return it even if they are not actually convicted.** Read more here and here.
It seems to me that it depends on what is happening in these cases. If confiscation of property is meant to express disapproval of people who are accused, but not convicted, of drug crimes, then we could use Feinberg’s analysis. That is, we could say that the expression of disapproval shows that this is punishment. Since it is meted out without passing through the normal procedures for punishment, it is illegitimate.
But I suspect that what’s going on is more simple than that. It’s that this is a corrupt practice in which the law has been bent to enable public agencies to steal from innocent people. If so, you don’t need anything fancy about punishment to say what’s wrong with the practice.
We ended with Feinberg’s arguments that the expressive theory does a better job of justifying punishment than retributivism does. I said that I thought Feinberg scored one point with his argument that punishment always hurts the innocent as well as the guilty. If so, there’s no way of meeting the retributivist requirement that punishment hurt all and only the guilty. The expressive theory, by contrast, can avoid this problem. The innocent are hurt when the guilty are punished, but only the guilty have society’s disapproval expressed towards them.
We were less certain about Feinberg’s other arguments. We didn’t see why it would be easier to obtain the information needed to put the expressive theory into practice than it is to obtain the information required to put the retributive theory into practice.