We are pleased to present the latest installment of our partnership with Ethics, in which we host a discussion on one, or in this case more than one, article from each issue of the journal. The articles selected from Volume 122, issue 1, are John Gardner and François Tanguay-Renaud's "Desert and Avoidability in Self-Defense" and Jeff McMahan's "Duty, Obedience, Desert, and Proportionality in War: A Response" (particularly section IV). We are very grateful that Victor Tadros has agreed to provide the critical précis, which appears below the fold.
In their rich, imaginative and complex paper, John Gardner and Francois Tanguay-Renaud (GTR) explore the idea that desert is relevant to self-defence. In his careful and typically pellucid response McMahan continues to defend the view that desert is irrelevant to self-defence. Obviously, there is a great deal in these papers to discuss, and any summary will be inadequate. I hope both to provide the main issues of debate and raise issues and questions to help advance the debate further. Apologies for any misunderstandings, which may be many and serious.
GTR think that desert provides an explanation of the asymmetry in self-defence cases – that we hope that the person defending herself succeeds against a culpable attacker, and will typically wish to support her in her defensive aims. McMahan denies that desert is important in self-defence.
What is desert? GTR don’t identify clearly what they mean by desert. They endorse the following: if D deserves x there is always a reason for x to be given to D. McMahan believes that ‘D deserves x’ implies that it is intrinsically good that D gets x. GTR reject McMahan’s idea. GTR may be right. We could replace the words ‘intrinsically good’ with ‘impersonally valuable’, leaving open the possibility that there are ways in which it can be valuable that D gets x which are not grounded in the intrinsic goodness of D getting x. In this GTR only clarify a possibility. They leave untouched the difficult question how it can be impersonally valuable that a person is harmed. Unlike GTR, I find this idea both implausible and barbaric (let’s call it the ‘barbaric thesis’ for short).
Why does McMahan think Desert is Irrelevant? McMahan seems to accept the barbaric thesis. Here is a question for him. If the barbaric thesis is valid, why rule out the idea that desert plays a role in self-defence? Why can’t it provide some reason to harm the aggressor in self-defence, at least in circumstances where it is clear that D will not later be punished for what he has done? McMahan claims that harming the person would have nothing to do with self-defence. But this evades the issue only by terminological fiat.
Even if the impersonal value of the person suffering were not a ‘defensive reason’, it would still have a role in determining the permissibility of self-defence. It seems hard to believe that it is intrinsically good that bad people suffer and yet that this plays no role in determining whether defending oneself against a culpable person in a way that makes them suffer is permissible. This conjunction of views seems difficult to motivate.
Furthermore, McMahan thinks that culpability is relevant to proportionality. Desert would provide an explanation why this is so. Of course, there are other explanations available grounded in the significance of responsibility and the importance of kinds of responsibility to questions of liability. But desert is at least a contender for an explanation why culpability makes a difference. More could be done by both GTR and McMahan to outline the range of possible explanations why culpability matters, and to defend one contender in the range. McMahan’s claim that culpability matters because it is a special kind of responsibility seems attractive to me given my antipathy to desert, but needs more defence.
Is desert sufficient reason to harm others? McMahan notes that it is intuitively wrong to harm a person in self-defence where no instrumental good is served. GTR reject the following claim that drives McMahan to reject the significance of desert to self-defence on this basis:
Desert excludes instrumental considerations (DEIC): If D deserves to be harmed it is permissible to harm D even if no further goal is served.
McMahan suggests that self-defence is instrumental in ambition and that desert is non-instrumental, hence rejecting the significance of desert. In rejecting DEIC, GTR note that defenders of plural theories of punishment sometimes claim that punishment is justified only if both D deserves to be punished and some further goal is served. Here is one explanation GTR offer: the impersonal value of D being harmed may never provide sufficient reason to harm D. This is because, in cases of deserved harm, there are always conflicting reasons – reasons of justice to give D what he deserves are always counterbalanced by reasons of humanity against doing so.
Whilst not incoherent, this view is hard to motivate and defend. GTR appeal to accounts of punishment by HLA Hart and John Rawls. This is not ideal. For brevity, let’s focus on Hart. Hart believed that deterrence was the general justifying aim of punishment, but that it was to be distributed on retributivist grounds. This seems to imply that desert had a role in Hart’s theory. But it is not obvious that it did – Hart was unclear about what he meant by retributivism, and probably rejected the idea that deserved suffering of offenders was intrinsically good or impersonally valuable, or even not bad (recall this excellent claim: retributivists appear to believe in ‘a mysterious piece of moral alchemy in which the combination of the two evils of moral wickedness and suffering are transmuted into good’ Punishment and Responsibility 234-5). One question for GTR, then, is how to clarify and motivate plural theories of punishment in a way that can provide support for the role of desert in self-defence. It seems unsatisfactory to defend a highly controversial account of self-defence by appealing to Hart’s highly controversial, unclear and inadequately defended account of punishment.
Intentional harming and Non-comparative Justice. GTR then explore the idea that desert might play a similar role in self-defence – as a matter of non-comparative justice, it is permissible to harm a person intentionally only if that person deserves to be harmed. Only wrongdoers deserve to be harmed. Here are their principles, which are general principles of harming:
(NCJ1) It is morally permissible intentionally to inflict suffering or deprivation (we will say ‘harm’ for short) only on those who deserve such an infliction and only to the extent that they deserve it.
(NCJ2) Those who deserve such an infliction are all and only guilty wrongdoers, to the extent and only to the extent that the infliction is proportionate to their guilty wrongs.
As McMahan suggests, taken literally, these principles seem implausible even if one accepts the barbaric thesis. They would at least need to be revised to accommodate cases where a person consents to be harmed. Here is a further reason in support of the idea that these principles are implausible. Suppose that D culpably attacks E with lethal force. E shoots D in the head. Shooting D in the head was necessary to avert the threat. E is an attempted murderer. E would deserve to die only if attempted murderers deserve to die. That seems implausible. If intentional harming is permissible only if it satisfies GTR’s principles of non-comparative justice, this would seem to imply that it is wrong for E to shoot D in the head. That also seems implausible.
Perhaps GTR will appeal to considerations of comparative justice here. NCJ (1 and 2) might be understood as exclusively governing the all things considered permissibility of intentional harming. If so, the example in the previous paragraph seems decisive against this view. But GTR leave open the possibility that there might be other grounds for acting in self-defence, as a matter of comparative justice, for example in mistaken attacker cases. In such cases GTR believe that the person harmed is wronged, but wronging her may be permissible. This suggests that ‘permissible’, in NCJ (1 and 2) is not to be understood as ‘all things considered’ permissible. It is about whether the conduct is prima facie wrong, leaving open whether that conduct may nevertheless be justified all things considered. It is not clear whether GTR believe that intentional harming may be all things considered permissible on a comparative justice basis.
If NCJ (1 and 2) are to be understood as governing intentional harming exclusively, much will depend on how to understand intentional harming. This is how I interpret the dispute between McMahan and GTR about intentional harming. McMahan reports that GTR believe that enforcing compensatory duties does not involve intentional harming. It is somewhat unclear whether GTR believe that harming in self-defence involves intentional harming, given that harm is only imposed to avert the threat faced, and if non-harmful means were available they would be used. If they do not, NCJ (1 and 2) seem less relevant for self-defence, as McMahan implies.
I agree with McMahan that harming, in compensation or in self-defence, is typically best understood as intentional harming, though there are difficult problems to solve here: specifically, if D intends x and x is very close (in some sense to be specified) to y, does D also intend y.
It is worth noting that in the case of punishment, unlike the case of self-defence, harm itself is typically aimed at. A person does not necessarily fail to defend himself if he fails to harm his attacker. He may succeed simply by disarming him. In contrast it is plausible that a person necessarily fails to punish another if he fails to harm him (the failure test of intent may not be perfect, but it will do here).
Perhaps GTR mean NCJ (1 and 2) only to govern intentional harming in the very strict sense that harm itself is aimed at. Even on the strict reading of intentional harming, the claim that intentional harming is permissible only if the person harmed deserves it seems implausibly restrictive. Consider:
Deterring the Mistaken Attacker: A reasonably mistaken attacker will repeatedly attack me unless I deter him. I can deter him only by harming him.
Surely I can harm him to some degree in order to deter him. Yet he is not a guilty wrongdoer. I doubt I necessarily wrong him if I do this, even prima facie. He has no right that I not harm him in this way to at least some degree – that he is liable to be harmed.
Here’s an explanation. He would have a duty to accept being harmed, or even to harm himself, to some degree for this end had he been, by some magical twist of fate able, in advance, to cause this to happen. It seems odd to say that this person has (or would have, were he to know the facts) a duty to cause himself to be harmed and yet a right that he not be harmed. It follows that this person is not wronged if an equivalent harm is imposed on him.
Conclusion. GTR could do more to clarify and defend their version of the barbaric thesis. They could also do more to explain the relationship between principles of non-comparative and comparative justice in self-defence claims and show that this relationship has plausible implications. And they could clarify whether NCJ (1 and 2) are intended exclusively to govern the permissibility of intentional harming, and if so what ‘intentional harming’ means. McMahan could do more to explain why he endorses the barbaric thesis if he does, and why he thinks it plays no role in self-defence. He could also defend his alternative account of why culpability matters to proportionality more deeply to demonstrate that desert is not doing the work.
Nothing that I say here should cast doubt on my admiration for both pieces. They do much to advance our understanding of a neglected issue in the philosophy of self-defence.