I have argued, first logically and then practically, that it is impossible to accept this modest proposal for torture in its own neatly limited terms. We still have to imagine to what consequences such a principle would actually lead. But for the sake of argument let us look a little closer at the ways in which Bagaric and Clarke, amongst others, attempt to defend an entitlement to torture.^[49] Thus it is said that the prohibition against torture has only served, by its unnecessary absolutism, to drive it ‘beneath the radar screen of accountability’; legalisation ‘would reduce the instances of torture’.^[50] It is difficult to see why this assertion would be true. Our societies are not without experience of legal torture. Was there less of it then? Moreover, the emotions that lead to real torture — fear, crisis, hatred — will not be reduced by legality. In what sense will ‘accountability’ make a difference to the practice of torture except to provide a helpful framework in which it can be organised, carried out and defended?

I admit that the radar argument sounds plausible: illegality does not always work and sometimes seems to make matters worse.^[51] This is particularly the case, for example, in relation to so-called victimless crimes such as drug use. Indeed, Bagaric and Clarke are explicit in adopting for themselves a discourse of ‘harm minimisation’ drawn from that literature.^[52] But torture is hardly victimless. Let us look a little closer to see where the analogy falls down. With drug use or prostitution, the argument is that legalisation will clean up the secretive conditions in which they occur and therefore not lessen the incidence of them but instead make them safer. In general, the scholars of what is called ‘harm minimisation’ do not contend that a more open approach to drugs will lead to less use; only that it will dramatically improve the social and health conditions of users.^[53] But it is not the conditions under which torture is practised that are the problem. Danger and pain are not a by-product of torture (as they are, for example, to a considerable degree a by-product of the current regime of drug prohibition); they are the purpose of it. Were torture done in public, were it supervised by a qualified medical practitioner in an hygienic environment, were it made respectable — tell me, would any of this make torture better? Once again Voltaire comes to mind: ‘If we believe absurdities, we shall commit atrocities.’^[54]

The centre-piece of Bagaric and Clarke’s defence offers as obvious example of begging the question as I have seen. They argue by analogy to ‘the right to self‑defence, which extends to the right to defend another’.^[55] Just as we are entitled to respond with violence to a murderous attack, they say, we are entitled to protect others; if the only way to protect them is by torturing somebody for information, then torture must be legitimate too. But the analogy falls down in at least three ways. First, the principle of self-defence recognises a reality: when it is ‘him or me’ a law that said I could not respond to an attacker would be simply unenforceable.^[56] Here the violence of torture is a choice deliberately made and carried out, and not purely responsive.

Second, their analogy assumes the only point it needs to prove. One can legally defend oneself; one can even kill an attacker if necessary; but what legal system has ever authorised a case of torture ‘in self-defence’? Why do the authors assume that self-defence, which is strictly limited to a direct, minimal and reasonable response to threat,^[57] is in any way equivalent to torture, which is by its very nature indirect and maximal? In fact, our societies have, at least since the Enlightenment, feared pain more than death,^[58] believed that human dignity requires absolute protection under all circumstances, and for that very reason thought torture a more serious act than execution. Legal systems throughout the world outlawed torture long, long before capital punishment. In the US, torture has always been contrary to the 8th Amendment; it is the paradigmatic example of ‘cruel and unusual punishment’.^[59] Yet the death penalty continues to be applied — as painlessly as possible.^[60] So clearly in the US, throughout the world, and by most people, it is generally considered worse to torture than to kill.^[61] As Bagaric and Clarke go to some lengths to point out, we do indeed often ask many people, including civilians, to make the ultimate sacrifice for the good of others, for example in times of war.^[62] Bagaric and Clarke think it obvious that if we can kill someone in self-defence, or require them to die for us, therefore it must be all right to torture them. But this is precisely what the absolute prohibition of torture rejects. Bagaric and Clarke could certainly make an argument against this orthodoxy. But neither they nor, to the best of my knowledge, other apologists for torture have attempted to do so. They simply assert their position as self-evident. It is nothing of the kind.

There is a third, and to my mind even more important way in which the analogy between self-defence and torture fails. Self-defence is about individual action, torture is about government action: the limits are not necessarily the same. There is a profound difference between individual acts of violence and a system of government‑regulated torture. There is a difference between kidnapping and government sponsored ‘disappearance’. There is a difference between murder — even mass murder — and genocide.^[63] The difference is the government sanction and the government power that stands behind it in each case.^[64] Government action — law — carries a mark of legitimacy with it. Self-defence, which leads to murder, or even revenge, might elicit our sympathy. We might even in some way excuse it. But it is not the same thing as a government program, which establishes, institutionalises, administers, and authorises torture. No matter how limited, torture is thereby made right in a way that no act of personal self-defence ever makes murder right. We hold governments to higher standards for a good reason.

So too, the reach and mechanisms of government power make torture a weapon from which no member of the community will feel immune. If the state could torture any one of us — they probably would not but they could — what sort of a society would we live in? Now Bagaric and Clarke attempt to avoid this problem by implying that torture would only affect the very few that in some sense deserved it. Although they admit that there may be situations in which ‘torturing the innocent’ to extract information would be justifiable if enough lives were thereby saved, the purchase of their argument as to the moral justifiability of torture rests on distinguishing those who are in danger and ‘blameless’, from those who are ‘wrongdoers.’^[65] In the newspaper articles written at the same time, the authors make this distinction central to their construction of the moral basis for torture. There, they insist that it is ‘verging on moral indecency … to favour the interests of wrongdoers over those of the innocent’.^[66] One way or another, the notion of wrongdoing underlies their assumption that we have the right person; that their suffering can be used to save another’s life; and that even if the torture fails to elicit information the loss is not in the end so very terrible.

But who are these ‘wrongdoers’ who, according to the general assumptions that underscore this whole debate, can be legitimately made to suffer so that the innocent might live? Perhaps they are only the associates of terrorists, or family members; and in any case any torture that takes place will very probably precede a trial that might establish just how culpable they are. After all, as both Bagaric and Clarke along with Gonzales insist, the whole point of the argument in favour of torture is our need ‘to quickly obtain information from captured terrorists [sic] and their sponsors’.^[67] So much for the rule of law: another suspicion has been magically converted into something we just happen to know. In fact, Bagaric and Clarke’s own argument shows us exactly how, under the pressure of time, and the urge to get immediate results, even the most modest regime of torture will inexorably corrupt what one might have thought to be core values of this society and of our legal system. ‘The investigation and trial process is simply one way of distinguishing wrongdoers from the innocent,’ they cheerfully muse. ‘To that end, it does not seem a particularly effective process. There are other ways of forming such conclusions.’^[68] Torture, for example?

In the real world, which again I realise it might be considered bad form to bring up, there are many reasons why we might all live in fear of a government that had reserved to itself some kind of right to torture suspects who it has determined in some way are ‘wrongdoers’. Perhaps it might just be a case of mistaken identity, or maybe you happened to be born with a foreign sounding name, or maybe you look suspicious or are the wrong colour, or come from a country with a violent history, or are otherwise associated with the wrong people, or perhaps you were just known for holding unpopular opinions at one time or other. How much torture might it take to show that you were not really a ‘wrongdoer’ after all? And what effect would that endemic, nagging fear have on all our lives and our relationship to the state? Peter Faris, former head of the National Crime Authority, says it would be alright ‘to pull out a fingernail of a terrorist in order to save a couple of million lives’.^[69] But the government legitimisation of torture, whatever the reason, would ultimately serve only to cripple a few million lives and corrupt our understanding of law and of justice. Bagaric and Clarke try and avoid these profoundly serious consequences by insisting that ‘our decisions in extreme situations will be compartmentalised to desperate predicaments’.^[70] But it is precisely this effort to quarantine our thinking about torture that is inevitably doomed to failure.