I first became acquainted with David Boonin (who is an associate professor of philosophy at the University of Colorado, Boulder) when I used the text, What's Wrong?, (which he co-authored with Graham Oddie) in one of my classes. His next book, A Defense of Abortion, is a carefully argued treatise in which he respectfully dismantles the extreme conservative position point by point. I say "respectfully" because he makes every possible accommodation to the conservative position, rejecting easier rebuttals that contain premises unacceptable to the conservative. Having read this book, Boonin was placed on my "must read" list, so I was overjoyed when The Problem of Punishment appeared in 2008, and even more pleased when I saw that within he questioned the validity of punishment.

I'm not going to spoil a great read with detailed argument summaries; rather, I will focus on a few major points and make a few comments on his methodology, for the book is not only a summary and criticism of current work on punishment, it is also a textbook example of how to argue a position. Boonin takes the time not only to carefully explain a position, but also describes exactly how the position can be criticized, burdens of proof, etc. In this part (one) of the review I will focus on the problem of punishment itself and discuss Boonin's definition of 'punishment'. In part two (which I will post in a followup) I will talk about pure restitution theory and how it should be approached based on Boonin's methodology.

Definitions

Boonin's definition of 'punishment', which is carefully crafted in the first chapter, is that a person, P, inflicts an act, a, of legal punishment on person Q for an offense committed by Q if and only if:

1. P is a legally authorized official acting in his or her official capacity, and
2. P does a because P correctly believes that Q has committed the offense, and
3. P does a with the intent of harming Q, and
4. P's doing a does in fact harm Q, and
5. P's doing a expresses official disapproval of Q for having committed the offense.

I don't usually like extended definition discussions, but Boonin uses the process to explore some of the various and subtle features of punishment. For example, point 2 raises the question of whether the innocent can be punished. Nothing important hangs on the answer (Boonin offers a weaker definition in which this correct belief is changed to "P believes, perhaps mistakenly"), and yet, in arguing that they cannot, Boonin makes an appeal to the structural symmetry of the two concepts of punishment and reward. Consider:

Suppose that I have lost my beloved dog and have offered a $500 reward for her safe return. As you walk into my front yard for a visit one afternoon, she comes bounding up the steps, and I mistakenly conclude that you have found and returned her. So, I give you the $500. Clearly, what you have received from me is a benefit, and you have received it from me for just this reason. (18)

But surely I didn't reward you! Rather, I think I rewarded you, though you know better. So also with punishment. If we fine or imprison you for something you did not do, then we think we are punishing you, but we are mistaken. The analysis with respect to the concept of reward is enlightening in this and other regards.

Point 4 raises a similar question with what Boonin calls the "masochist objection." What if a person actually enjoys the pain we inflict upon him? Is he still being punished? The objection itself is weak, for masochists generally like pain inflicted in very controlled ways, but the it raises a host of interesting issues concerning what it means to harm a person. I'm reminded here of John Locke's famous example concerning free will in which he imagines a person locked in a room with someone he always wanted to have a conversation with. Locke claims the person is not free because, although he enjoys speaking with the person and has no desire to leave, he couldn't leave if he wanted to--he is externally constrained. So also with the masochist. Even if he enjoys certain aspects of his situation, he has no control. He cannot end it whenever he wishes, and therein lies the harm.

Again, it is useful to appeal to the punishment-reward symmetry: if I give you something to repay you for an act of kindness, and it turns out that this unintentionally causes you harm, then I did not in fact reward you. I attempted to; I intended to; but I did not. So also with punishment. If my action against you does not really harm you then you are not being punished.

A final definitional point. Punishment must involve intentional harm. When we harm someone unintentionally, as in a car accident, we are not punishing them. But though necessary, intentional harm is not sufficient for punishment, for we are intentionally harmed by the state in many ways we do not consider punishment. This discussion sheds light on an important distinction. Consider the following cases:

1. Larry marries Laverne and is charged a fee for processing his marriage license.
2. Moe marries both Betty and Veronica and is charged a fine for violating antipolygamy laws.
3. Curly is found not guilty of murder by reason of insanity and is confined against his will for the rest of his life.
4. Shemp is found guilty of murder and is confined against his will for the rest of his life. (12)

In all four cases an individual is being intentionally harmed by the state, and yet, while 2 & 4 are clearly cases of punishment, 1 & 3 are clearly not. Boonin's explanation hangs on a more precise look at the intentions behind each of the actions, for in 2 & 4 the explicit intention of the state is to harm the person. In imposing a licensing fee, the intention is to cover various costs involved in the licensing process, and in committing a person found not guilty by reason of insanity the intention is to protect society from the individual. This distinction is further illustrated by a supreme court decision with regard to Megan's law, which authorizes state's to post pictures and information of convicted sex offenders on the internet. The law was challenged on the basis that it imposed an additional punishment on sex offenders over and above the punishment meted out by the courts, but the supreme court deemed the law constitutional on the grounds that, though the offenders were in fact harmed by the law, this was not its intention.

The distinction used here by Boonin is the same distinction used by Gay-Williams and others in the euthanasia debate. Both use it in order to block an implication, the acceptance of which would render a too-broad definition. Boonin needs "intentional harm" in his definition of 'punishment', but must exclude specific cases (licensing fees, etc.). Williams needs "intentional killing" in his definition of 'euthanasia' (which he argues is immoral because of this) but must exclude certain cases (such as cases in which treatment is withheld) which he thinks are moral. The problem is that the use of the distinction is controversial in the euthanasia debate, which means it might also be problematic with respect to punishment. Boonin is silent on this, but what about those who might want to question the distinction on the one hand and allow it on the other?

The answer lies in the fact that Boonin never denies that we intentionally harm Larry and Curly in the above examples; his claim is merely that harming them is not our primary intention. Williams, though, in order to make his point, must deny that we intend to kill by withholding treatment. This is clear in his claim that we no more intend the death of a person by withholding treatment than we intend to influence the GNP by buying a pencil, and it is this that makes his argument disingenuous, for in fact, Boonin's analysis should also be applied to withholding treatment: we do intend to kill the patient, but it is not our primary intention. Indeed, as with harming Larry and Curly, if we could bring about the result in a way in which they were not harmed we would surely do so.

The Problem of Punishment

Inflicting intentional (and not merely foreseen) harm upon individuals, then, is a necessary condition for punishment, and it is this very condition which underlies the "problem" of punishment. Punishment theory has to show how the distinction between offender and nonoffender is "important enough to justify not merely harming those on one side of the line but intentionally harming them." (28-9) Punishment theory allows the state to harm people in ways that would be unacceptable for individuals.

The psychologist Jonathan Haidt has given us the interesting term "moral dumbfounding," to describe a process in which we intuitively label a practice moral or immoral, without knowing exactly why. One of his examples concerns a brother and sister (both on birth control) who decide to have sex one night while on vacation in Europe. Though they enjoy the experience they agree not to repeat it. The reaction to this story is almost always negative, but when asked "Why?" the answers are rather shallow. In other words, there is often a disconnect between our emotional reactions and our deliberations.

Reactions to the concept of punishment are similarly emotional. We humans seem to have a built-in "tit for tat" mechanism. If someone does something either to us or for us, we feel an instinctive obligation to repay in kind. This instinct forms the basis of penal codes dating back to Hammurabi and before, a foundation guaranteeing some sort of lex talionis approach. This knee-jerk reaction forms the basis of many of the justifications of punishment, and is linked to another common justification, the idea that we must punish offenders in order to prevent total anarchy. This brings up pictures of countless movies in which thugs are picking on the helpless who cannot defend themselves until the "hero" arrives and punishes the wrongdoers. We get a huge endorphin rush once the baddies are properly bested and the world is again set right.

All this is pure vengeance: the intentional infliction of harm on another because they did something we think they shouldn't have, or didn't do something we think they should have. Grant that the desire for vengeance is inherent within our nature; is it really the best way to proceed? To paraphrase Hepburn in African Queen, "Our nature, in this case, is what we are meant to rise above."

The two major justifications of punishment theory are derived from the two major ethical systems: consequentialism (utilitarianism) and formalism (retributivism). Boonin spends over 180 pages carefully analyzing these and other proposed solutions to the problem of punishment, but ultimately finds them wanting. If we accept his arguments, and they are compelling, what are we to do with those who break the laws of society and endanger our lives and lifestyles? Boonin thinks restitution is the answer and it is his defense of this theory as an alternative to punishment that I will examine in part two of this review. Stay tuned...