05.11.11

Torture Can be Wrong and Still Work

German authorities used beatings and threats of torture to get Magnus Gäfgen to confess to killing a child. Stephen Carter on why, if the goal is to obtain information, torture can work.

German authorities used beatings and threats of torture to get Magnus Gäfgen to confess to killing a child. Stephen Carter on why, if the goal is to obtain information, torture can work. Plus, full coverage of Osama bin Laden's death.

Unless you happen to follow the German press—or unless, as I did, you happen to have taught a German law student interested in the subject of torture—you might have missed the tale of Magnus Gäfgen, the convicted child murderer currently suing the Hesse police for beating him and threatening worse in order to extract a confession. The Gäfgen story seems quite apropos now that, in the wake of the killing of Osama bin Laden, our national debate over the use of torture has taken a bizarre turn, from whether torture is right or wrong to whether it ever works.

Back in 2002, Gäfgen kidnapped an 11-year-old boy, Jakob von Metzler, whom he then murdered. Without disclosing that Jakob was dead, Gäfgen demanded a ransom of €1 million from the child’s wealthy parents. He collected the ransom, and was arrested soon after. The police, who thought Jakob was still alive, demanded to know where he was hidden. Gäfgen refused to say. According to Gäfgen’s lawsuit, they beat him, then told him a torture specialist was being flown in, a man whose training would enable him to “inflict more pain on me than I had ever experienced.” At that point he confessed, telling the police that the boy was dead and where his body could be found.

Now, I am not endorsing what the authorities did in interrogating Gäfgen, but I do think it provides evidence—which should hardly be necessary—that, if the goal is to obtain information, torture sometimes works. Let me repeat that: If the goal is to obtain information, torture sometimes works.

We know that information gained through enhanced interrogation has at times proved correct, and even useful.

I am not saying that I like torture. I am not saying we should do it—I think we shouldn’t. But to rest the moral argument against torture on the proposition that it doesn’t work eventually starts to sound silly. We have created a peculiar cognitive dissonance, where we want all good things to be true at once: so our military forces and intelligence analysts are to be congratulated for their exemplary work in discovering the whereabouts of Osama bin Laden and dispatching him, while, at the same time, we insist that none of the so-called enhanced interrogation techniques had anything to do with producing any of the information that helped lead to his hiding place. But there is so much evidence to the contrary (including the words of Leon Panetta, the director of the Central Intelligence Agency) that this position is no longer seriously sustainable.

Torture is wrong for all sorts of reasons, from its affront to basic human dignity to its violation of fundamental human rights. Making a moral case is not difficult. The puzzling part is that so many people insist on joining the moral case (torture is wrong) to the empirical case (torture never works) even though the empirical case is unpersuasive. We ought to be adult enough to accept the possibility that a tool might exist that is wrong despite the fact that it is useful.

The claim that torture never works is a popular corruption of a more serious claim, made by many professional interrogators, that torture produces unreliable results. The unreliability argument rests on a simple, and surely accurate, assessment of probability: if the pain or horror is bad enough, the victim will say anything the torturer wants, just to make it stop. The victim will sign a confession, make up a story, do whatever it takes. “Yes, I robbed the bank,” the innocent man admits after half a day of beating, and he fills in whatever details the torturer supplies: the motive, the make of the getaway car, what have you. If the purpose of the torture is simply to extract a confession, and if the victim can be punished on the basis of the confession alone, there is no check on the accuracy of the story. The torturer will thus extract plenty of guilty pleas, many or most of them from the innocent.

On the other hand, when the torture—the enhanced interrogation—is intended not to adjudicate guilt but to extract information to lead to the whereabouts of a kidnapped child, there is indeed an implicit check: If the suspect is lying to make the torture stop, then the child will not be where he tells the police to look. Whatever else may be said about the threat by the Hesse police to torture Gäfgen, his confession was truthful: The threat did work.

Surely the same is true when enhanced techniques are used to extract information that might lead to the uncovering of a terrorist plot—or the nom de guerre of the courier who carries messages back and forth to the leader of al Qaeda. If no courier of that nom de guerre is ever found, or if the safe houses the suspect describes turn out not to exist, one can go back and try another way. The check on the confession comes in the real world.

Again, let me emphasize, I am not defending the morality of torture, or of other methods of enhanced interrogation. I am simply suggesting that others who, like me, oppose the techniques should stop pretending that nothing useful ever comes from them. One does not have to endorse what the police did to Magnus Gäfgen to recognize that had he been threatened with torture a little earlier, his 11-year-old victim might be alive to this day; and one does not have to endorse whatever our own interrogators may have done to recognize that had they not done it, Osama bin Laden might still be safe in Abbottabad, plotting another attack. Maybe had the police used gentler methods, Gäfgen would still have confessed. Maybe information could have been obtained from terror suspects through less forceful means: We should all hope so, but, as Panetta has said, we will never know.

What we do know is that information that has come from repugnant methods has at times proved correct, and even useful. Torture can be morally wrong and still sometimes work, a proposition that those of us who are against it should be mature enough to admit.

Stephen L. Carter is the William Nelson Cromwell Professor of Law at Yale, where he has taught since 1982. His seven nonfiction books include God’s Name in Vain: The Wrongs. His first novel, The Emperor of Ocean Park (2002), spent 11 weeks on The New York Times bestseller list. His twelfth book, The Violence of Peace: America’s Wars in the Age of Obama, was published by Beast Books in January.