A couple of years back Tim Scanlon did a blog post and comment-section discussion on PEA Soup. Here’s one bit:

In earlier work, including my Tanner Lectures on the significance of choice and Chapter 6 of What We Owe to Each Other, I rejected the idea of moral desert because I identified it with the idea that the fact that someone has behaved badly can make it a good thing that he or she should suffer some loss. I still find the latter view morally unacceptable. But it now seems to me that this rejection of desert is too quick. Desert should not be identified with this retributivist idea. There is, I believe, a distinct category of valid desert-based justifications. A desert-based justification for treating a person in a certain way claims that this form of treatment is made appropriate simply by facts about what that person is like, or what he or she has done. By simply, I mean without need to appeal to other factors such as the good consequences of treating the person in this way or to the fact that this treatment is called for by some institution or practice that is independently justified. Moral blame, gratitude, and some honors and distinctions can be justified in this way, and these justifications do not presuppose that the qualities that form the basis for justification are all under the person’s control. The responses are justified simply by what the person is like, or has done. By contrast, legal punishment, insofar as it involves forms of hard treatment such as fines or imprisonment, cannot be justified purely on the basis of desert, nor can significant differences in economic reward be justified in this way. I argue for these views in “Giving Desert Its Due,” which has just appeared in Philosophical Explorations.

I’m just today reading the article in Philosophical Explanations, and it has some interesting features that bear on some of my recent work with Daniel Levine. For one thing, he tries to argue that when we distinguish moral from legal blame–that is, when we distinguish blaming from punishment–we can start to justify withdrawing our personal willingness to have special relationships and obligations to a person based on what they are like or what they have done. For instance, if your male neighbor abuses his wife, it’s reasonable to find yourself less trusting of him, less willing to enter into friendship or shared projects, and less happy for him when things go well in his life… and thus less unhappy when things go badly. (page 11)

But what’s important is that these attitudes are all of the “special” designation: conditional attitudes which we cannot grant equally to all and thus appropriately deprive most people of–and now the neighbor as well. It is appropriate to “withdraw good will” towards a person if they act in certain ways. This is the distinction: everyone, regardless of their behavior, deserves certain unconditional kinds of respect. But conditional forms of respect are conditioned and thus winnable and loseable: esteem, deference, and honor; disesteem, disdain, and contempt.

So far, so good. Now, we know on Scanlon’s account that in some sense responsibility is merely a matter of attribution: so we blame and praise and engage in all the special interpersonal relationships because our actions are attributed to our character. We don’t just blame or praise an action, we blame or praise the person for being the kind of person who would engage in that action.

Now as it happens, I am not convinced that we need to so quickly conflate acting and being. Because what happens in punishment, at least in our society, is that everyone simultaneously withdraws good will towards the prisoner, while simultaneously we become willing to inflict suffering upon them. Scanlon focuses on this second element, and deplores it as not appropriate when we decide not just to acquiesce (by being less unhappy) to their suffering, but to actively visit it upon them, to act (often at our own expense, and obviously so in the criminal justice system) to make them worse off.

Interestingly, this rule-following punishment that Scanlon deplores is at the heart of the social production of norms in community. It’s at the heart of common pool resource management, including the management of the common pool of social reasons and thus our community and its mores, so there’s a strong practical sense in which Scanlon is probably wrong. More on this in moment.

Where Scanlon really seems to go wrong is in the special kind of deprivation in the widespread withdrawal of special relationships that Scanlon calls for in addressing the wrongdoer. It takes an odd kind of individualist contractualism to assume that the universal deprivation of good will and willingness to share projects is somehow unobjectionable. Shared projects are at the heart of human agency. They are the basis not just of the special respect of friendship and love, but of the shared practices that make dignity possible. We have ample evidence of this, that the conditional and unconditional interpersonal relationships are not as easily prised apart as Scanlon claims here.

I think we must probably accept that Scanlon has the wrong end of this problem. Probably it should work something like this: we deem it acceptable to visit suffering on another in a directed manner as a response to normative failures like wrongdoing. But we do so under the understanding that the punishment is a part of the restoration of the conditions of special interpersonal relationships–of good will. We punish so that we can go back to trusting and collaborating. Anything else is inappropriate. Thus we restore the priority of the conditional over the (allegedly) unconditional reactive attitudes, acknowledging as we do so that these never really were unconditional attitudes in the first place, that reactivity always trump unconditionality and honor always trumped dignity.

I need to think some more about what that means for the latter half of the paper, where Scanlon tries to tie these questions of punishment and moral blame to economic distributions and redistributions. I’m sympathetic to his conclusions there, but having undermined his foundations I’ll have to see if I can justify an alternative means to that end.