Anscombe on Punishment
I remember we had a good discussion a while back on the nature and purposes of punishment. I argued that defense against the criminal is the fundamental (and often only) purpose of punishment, and others pointed to the importance of retribution. I thought it would be useful here to reference the thought of Elizabeth Anscombe, as laid out in her insightful essay from 1978, “On the Source of the Authority of the State” (published in G.E.M. Anscombe, 1981, Ethics, Religion, and Politics: Collected Philosophical Papers Volume III, Basil Blackwell, Oxford.
Anscombe begins by asking a question that seems almost too simple: how can the state, with its monopoly on violence, be justified? She shows that a lot of the functions of the state in this regard are common to civil society and voluntary organizations. If you start from the premise that any person may justly intervene with violence “to resist violence committed without right”, then you are forced to admit that the state goes far beyond this, by investigating past actions, trying and punishing people, and forcing rules on them. Clearly, a voluntary society that used armed guards for such purposes would not be legitimate. What, then, makes the state legitimate? She dismisses the Hobbesian solution immediately, by arguing against the notion that civil authority arises from the transfer of rights already possessed by people without a state. For there is no private right of punishment in a state of nature. As she puts it” “civil society is the bearer of rights of coercion not possibly existent among men without government”. This has obvious implications in the domain of punishment.
To start, Anscombe asks if those subject to punishment are wronged by it. She answers this question in two senses. In the first sense, they are not wronged if they are truly guilty, as they are not “getting what they don’t deserve”. But in a second sense, they may be wronged, by “getting one’s deserts at the hands of someone who has no right so to inflict them”. Legitimacy is essential.
Anscombe concludes that punishment must be based on need, not on prescription. In other words, “the justification of the institutions of law, charge, trial, and sentence can only be the protection of people”. The other purposes of punishment flow from this concept, and do not exist independently of it. On the crucial issue of retribution, Anscombe is crystal clear:
”A retributive theory of punishment is merely a punishment theory of punishment; therefore correct, in that it declares that nothing can be properly called punishment if it is not offered as affliction deserved by ill-doing; but incorrect, in that it says that punishment of wrongdoing is eo ipso justified and needs no further reason.”
This is crucial. In her interpretation, “retribution” is simply another word for punishment. Retribution is not something added to punishment. When punishment is not strictly based on the need to protect people, then it becomes an abuse of state power. When the state abuses power, force becomes violence, and violence is sinful. (I wonder is Scalia understands this?). She goes on:
“For even if (which one may doubt) there is something intrinsically good about an evil-doer’s suffering, what is one man or some set of men that they should bring this about? Are they so good themselves? And are they in charge of the order of things, to see to it that such a good is brought about? It is obvious nonsense.”
Ancombe goes through a similar exercise to show that the other purported purposes of punishment– deterrence and reformation — also cannot stand in isolation from its primary purpose. As for deterrence, it is simply “putting affliction on people for something they have done, and therefore, if it is to be just at all, the something must be a known offense and the punishment deserved.” If that is the case, the deterrent theory is correct. But in isolation, it becomes an abuse.
As for the reformatory idea, Anscombe likens it to assimilating the state to a parent, which she dubs a “monstrous impertinence”. Why is it that some people simply because they exercise civil power should be able “dispose of others in this way against their wills for their own good”?. Even if the aim is noble, Anscombe denies that such a right exists. No, for the key question of legitimacy, necessity is key. All other purposes of punishment flow from this, and are distinctly secondary considerations.
Comments are closed.
This would seem to lead to the idea that the purpose of the criminal law is the need to protect society. What are the purposes of the civil laws, then?
If I recall correctly, Anscombe thought that the purpose of the civil law was to enforce prescriptive (i.e. customary) rights.
I’m guessing, then, that she would include contract under the term “prescriptive.” Interesting.
MM — I do not see that it follows from “retribution is not something added to punishment” (a position that I am happy to endorse; this is, in fact, what those of us who insist on a retributive theory of punishment maintain, i.e., that “punishment” done only in order to deter others from anti-social behavior is not really “punishment”) that “when punishment is not strictly based on the need to protect people, . . . it becomes an abuse of state power.” In any event, this last statement seems mistaken. That is, it is not the case (and I am not aware — though I could be wrong — that Anscombe asserted otherwise) that punishment is justified / legitimate only when its purpose is to “protect people,” if by “protect people” one means “protect third parties from harms likely to be caused in the future by the person being punished”.
As for what “Scalia understands”, I’m confident he understands that wrongful violence is sinful. (He would deny, I suppose that all “violence is sinful”.)
On deterrence and reformation, Anscombe is, again, entirely retributivist, as punishment theorists use the term. Those of us who insist on a retributive theory of punishment agree entirely — indeed, it is our main claim — that it is “monstrous” for the state to simply set about coercively improving people. This is not to say, though, that punishment — once justified — should not aim at, among other things, the moral improvement of those being punished.
You end with “necessity is key” and “all other purposes of punishment . . . are distinctly secondary considerations.” But, none of this refutes the retributive theory of punishment. That theory holds, at bottom, that what makes an otherwise unjustified imposition by the state on the liberty of an offender justifiable “punishment”, as opposed to mere state-thuggery, is the offender’s desert. Desert, on the offender’s part, is what justifies punishment. It is not the case that “necessity” (in the sense of “necessary to prevent future wrongs by this offender”) is required to justify punishment, even though it is certainly the case that punishment should be structured in such a way that it, consistent with its justification, deters, incapacitates, and reforms.
The “other reason” Anscombe meant is that the punishing public authority is legitimate. She did not mean to suggest that, unless incapacitation is necessary to protect third parties, it is immoral for legitimate public authority to punish wrongdoers in accord with their deserts.
Rick– thanks for responding (I was hoping you would!).
What I take away from Anscombe is that punishment can only be based on need, which means the protection of people. If there is no such need, then punishing somebody who may deserve it is not legitimate, because the person has no right to do the punishing. She says quite specifically that punishment of wrongdoing by its very act is not legitimate in isolation.
So, it would appear from what you are arguing, MM, that the state’s punishment of a spouse under murder or manslaughter laws would be illegitimate for a person who kills his / her spouse and spouse’s extramarital lover when catching them in flagrante delicto, because in this case, the killer is unlikely to kill again (being driven by a very specific set of circumstances), and the individuals needing protection, the spouse and paramour, are already dead.
Your interpretation of Anscombe’s logic might lead to the conclusion that such an individual – even if he/she stalked the spouse and paramour, tortured them, and killed them, driven entirely by his/her awareness of the extramarital relationship – could not be legitimately punished by the state.
I’m not sure my answer was entirely satisfactory. I think we have an issue of semantics here. The offender’s desert is virtually the same as the necessity for punishment. A thief deserves punishment because he stole, and because we need to protect the property of others. It is important to avoid notions of nominalism, the notion that each word refers to a distinct entity. In reality, words overlap, and often what they refer to are entities intrinsically related almost to the point of overlap.
Let me use my example above of the offender who will be unlikely, if ever, to repeat his / her offense. I would argue that the only proper reason that such an individual should be punished is that the individual has transgressed a law laid down by society for protection of it’s citizens. In other words, the people have said, “This far – no farther,” which is why legitimacy of a government is important, for it enacts the rules for virtuous behavior as established by society. Having transgressed that rule, then punishment, purely for the action taken, is merited. The level of punishment merited is based, or perhaps ought to be based, on the interior state of the individual offender.
I am not sure Anscombe is arguing something other than what I have said above in terms of the purposes of punishment. Your thoughts?
I love this stuff! MM, you’ve looked at the relevant texts more recently than I have, and so I have to concede the possibility that I am mis-remembering, but . . . I would be very surprised if Anscombe’s view was that the coercive restraint, by legitimate public authority, on the liberty of a wrongdoer is justified only in those cases where the restraint is necessary in order to prevent future harm, from the wrongdoer, to third parties.
We agree, though — and this is an important point — that the “is X punishment justified” question is not answered simply by observing that the person to be punished did wrong and deserves punishment. The question of the public authority’s legitimacy to impose X punishment is, it seems to me — and, it appears, to you — also a crucial one.