[UPDATE: I have in considerable measure conceded error on this subject, thanks to some very cogent arguments by Mark Kleiman; see here for my explanation.]

I am naturally daunted, as any thoughtful person would be, by the fact that my views on this run contrary to my nation's constitutional regime, contrary to what is seen by most as a worthy long-term trend in the civilization to which I belong, and the views of many people (both on the left and on the right) whom I admire. (My views are also largely pointless, since they can't be implemented in my country without a constitutional amendment that isn't going to happen.) Perhaps I am grievously mistaken, and have fallen victim to unsound emotion or the first flush of fatherhood.

Yet after reading the counterarguments, I confess that I continue to find them quite unpersuasive. I've gotten many more than I can possibly respond to, but I think I have an obligation to respond to some, so let me focus on those coming from Mark Kleiman, Matthew Yglesias, Maimon Schwarzschild, and Clayton Cramer. These are interesting and generally very thoughtful arguments (and I also thank their authors for framing the arguments not just civilly but quite generously, despite their disagreement with my views).

1. Clearing the underbrush. Let me first deal with a few general criticisms that I think are unhelpful here. Clayton quotes Gandhi's "An eye for an eye will blind the world," but while that might be relevant in some situations — for instance, as a warning against ethnic vendettas — it is about as relevant here as "Imprisoning kidnappers will leave the whole world locked up." It tells us nothing about the propriety of various punishments (whether prison, death, or deliberately painful death) for people who rape and kill 20 children.

Likewise, it seems to me that Maimon's analogy to lynch law is misplaced. Lynch law is bad for many reasons: Among other things, it doesn't provide adequate factfinding procedures, it leaves us at the mercy of our neighbors with no legal structure to channel and restrain the neighbors' actions, and it has often been used in racist ways. None of this tells us what punishments may properly be imposed by the legal system, or whether the legal system, in administering the punishment, can allow the victims' relatives to participate — not in deciding who's guilty, but in applying the legal penalty.

2. What about the Nazis? Maimon asks "if you 'execute' the serial killer of twenty children in this way, what do you do to criminals who are worse still? . . . What would Eugene wish the State of Israel to have done with Adolf Eichmann?" Yet this seems to me to support my original point rather than to undermine it. It seems to me an occasion for regret that Eichmann was executed by hanging. Such a decision was likely politically necessary; but I think it slighted the enormity of what he had done. He deserved a far worse death, and it would have been good had he received it.

Likewise, Clayton points to Hitler's having executed the von Stauffenberg coup plotters by hanging with piano wire, and to his having filmed the execution so he could enjoy watching it later. But what makes this bad is that the von Stauffenberg plotters were trying to do something very good. Had things been reversed, my regret would have been that hanging with piano wire didn't inflict enough pain on Hitler (though I would have been glad that he hadn't been turned over to a too-"civilized" government that would have dispatched him with less pain). Seriously, would most of us disagree? Maimon points to George Orwell's criticism of what he saw as the unduly painful hangings of some Nazis after World War II. I find much to admire in Orwell, but I don't share his generosity here (I speak here of the Nazi leaders generally, though recognizing the possibility that some lower-level military officials deserved to live, or even deserved to die painlessly).

Of course, as Matt and Clayton point out, these penalties are obviously inadequate. Like punishment generally, these punishments don't bring back the dead, or even inflict a fraction of the pain that the monsters have inflicted. But one should do what one can, and surely Eichmann et al. offer as strong examples as possible. (In fact, I didn't bring up Eichmann and Hitler in my original post because people could have plausibly argued that one can't really generalize from the abberational cases such as that of the Nazis; but if people bring up Eichmann, I have to acknowledge that my argument applies in spades to him.)

3. Practical effects: Matt argues that deliberately inflicting pain, even on the monsters, would cause bad effects on society: "The natural result of giving official sanction and encouragement to the desire to inflict suffering beyond the amount of suffering that serves a constructive purpose within the context of criminal law will be to encourage people to act on similar impulses (and, indeed, to have the impulses themselves) in non-criminal contexts as well. The result would, simply put, be a social disaster in which individuals are encouraged to nurse grudges, indulge spite and envy, and generally speak wreak havoc upon their fellow man." Maimon agrees.

If I agreed with this empirical speculation, I would come to a different result. But I just don't find it to be terribly plausible. People, it seems to me, have a natural desire to inflict pain on moral monsters. I doubt that the legal system's actions will much exacerbate this desire. If someone raped and killed your child, would your desire for revenge be much altered by what you know of the legal system's rules? I may be wrong, but I doubt it. (I agree that it might be altered by the legal system's threat of punishing you for the revenge, but that's a different matter.)

One can make an equally plausible claim, I think, that people will be less likely to seek private revenge if they think the legal system will impose accurate punishment: They'll both find private revenge less necessary, and will more generally trust and respect the legal system. This is utter speculation, I realize — but so is Matt's empirical argument. My sense is that one's empirical guesses on such things more often follow one's moral judgments rather than vice versa.

Nor have I seen evidence that harsh punishment generally makes society more brutal. The sharp increase in U.S. homicide rates in the 1960s and 1970s, for instance, followed a broad decline in the use of the death penalty. I'm not claiming that the decline in the death penalty caused more brutality; and I agree that death penalty calculated to inflict more pain (even if applied to a very few monsters) is different from the death penalty as such. But evidence such as this leads me to doubt that legal harshness in dealing with the guilty will translate into private harshness.

4. Assuming the Conclusion: There is, however, a deeper objection to Matt's point. Matt argues that it's proper to punish criminals but only to the extent that it "serves a constructive purpose." Presumably he'd think that incapacitation, deterrence, and rehabilitation are three such constructive purposes; a deliberately painful death penalty will add nothing to incapacitation or rehabilitation, and I'll also assume that it adds little to deterrence.

But in my view retribution is also a constructive purpose. This is most easily seen if we for a moment set aside deliberate infliction of physical pain, and even the death penalty. Consider a scenario where punishment will do little to prevent future crime: For instance, the imprisonment of Nazis who committed their crimes decades ago and are now in their 60s or 70s. There's little need to incapacitate them as a means of preventing future crimes and little likelihood of rehabilitating them. Nor will it do much to deter future atrocities, I think. If people are deciding whether to participate in a future Nazi regime, they'll probably be much more worried that they'll just get killed in the war, or killed shortly after the war by people seeking revenge. I doubt that many would-be Nazi war criminals in 1941 would have been deterred by the risk that some decades later, when they're old men, they'll be tracked down. No, the real reason it was right to punish them was retribution (as Mark points out).

In my view, painful death for certain monstrous acts is the proper level of retribution — anything less is inadequate, just as a slap on the wrist would be inadequate for an armed robber, or a short jail term would be inadequate for a rapist. Therefore, such a punishment does serve a constructive purpose — the purpose of retribution. Matt may disagree that retribution is a constructive purpose, or he may disagree that painful death is the proper level of retribution (he may think it's too much). But his argument doesn't demonstrate any of these points. Rather, it rests on the assumption that a painful death penalty for monsters doesn't serve the constructive purpose of retribution or that retribution isn't a constructive purpose, which are the very things he was trying to prove.

5. Humanity: Likewise, I think, with Mark's argument that deliberate infliction of pain, even on monsters, "makes the person who engages in it a little bit more of a beast, and a little bit less of a human being, than he would otherwise be." First, we should recognize that this is a metaphor; I may be mistaken, but my sense is that most literal beasts (i.e., animals) don't actually try to inflict pain as punishment for wrongs. Literally speaking, this desire is quite characteristic of human beings (though perhaps some other higher primates might be included; I'm not sure). This doesn't make Mark's argument wrong, but only shows that we need to look behind the metaphor.

So what's behind the metaphor? It could be a judgment that it's beastly, less-than-human, and thus morally improper to succumb to our visceral emotional impulses. But I don't think that's what Mark literally means. Love, empathy, the desire to pick a mate, the desire to have children, and other worthy emotions are also visceral emotional impulses; while we should certainly indulge in them with rational caution and care, there's nothing wrong in following emotions, and it's sometimes bad to resist them.

I take it, then, Mark's point is that it's beastly, less-than-human, and improper to indulge this particular emotion. But that too, I think, assumes the conclusion. When someone rapes and murders twenty children, why is it a "beastly" impulse as opposed to a worthy one to try to exact a harsh retribution? Mark acknowledges that retribution in general is a proper goal of punishment — but his argument doesn't, I think, explain why this particular sort of retribution is not. (To be fair, he does say "in my eyes, at least" — here we may be returning to a point I mentioned in my original post, which is that a lot in this debate rests on people's visceral moral intuitions.)

6. Risk of Error: Mark also points to the risk that we might be wrong, a risk I briefly discussed in my earlier post. His Torquemada analogy doesn't work for me — I think that we'd have contempt for Torquemada even if he had simply painlessly executed insincere coverts, rather than burning them at the stake, or even if he had locked them up for life. Conversely, had he burned at the stake people who had raped and murdered 20 children, we probably would barely remember him. Our condemnation of him is based on disagreement with his substantive moral judgment about the crime; and I think we'd say that the risk of such moral error about what's guilty conduct is indeed different from the risk of factual error about who's guilty of it.

Nonetheless, I admit that all human institutions have a capacity for error, and wrongfully inflicting deliberately painful death is indeed a more serious error than wrongfully inflicting painless death, or wrongfully imprisoning someone for life. The question is how this risk of error balances against the moral imperative for retribution. This is a question that defenders of the death penalty must ask themselves. (I doubt that the death penalty as currently administered has much of a deterrent effect; I think it's justified because some people deserve to die, and it's unfair to their victims and the victims' families not to execute those people.) It's likewise the question with regard to deliberately painful death penalty.

One can certainly reach a different judgment than I do: Even if one thinks there's some moral benefit to executing the Eichmanns or even the serial rapist-killers, one might say that the benefit is small enough that it's exceeded by the risk of error, and the very serious moral cost of that error. As I mentioned at the outset, I am keenly aware that I may be wrong on this general question, and the matter that causes me the most trouble is precisely this one. Yet my tentative current sense is that for a small number of extraordinarily monstrous crimes, the need for retribution is so strong — and the risk of error can be made so low — that not just death but deliberately painful death is the proper punishment.

* * *

In any event, I have gone on at ridiculous length. Yet, as I mentioned above, I think respect for those particular people who disagree with me on this, and broader respect for the weight of moral authority against which I'm pushing, required me to provide some response. I hope even those who disagree with me have found these arguments to be candid, clear, and fair to those whose arguments I'm trying to rebut. And perhaps the arguments may be helpful even to much more pragmatic debates, such as those about the death penalty generally, and about retribution still more generally.