Does torture work?

Military analysis.
Sept. 14 2004 5:01 PM

Does Torture Work?

Seymour Hersh evades the question.

Book cover

Seymour Hersh's new book, Chain of Command: The Road From 9/11 to Abu Ghraib, reveals our most intrepid investigative reporter working near the top of his game. Basically a compilation of the pieces that Hersh wrote for The New Yorker over the past few years—expanded, updated, and re-edited, in some cases significantly so—the book holds up as a cohesive tale and a searing indictment of the Bush administration: its chicanery with intelligence in the months leading up to the Iraq war, its inadequate planning for the war's aftermath, and its muffing of all the wars—in Iraq, Afghanistan, and the broader war against terrorism—ever since.

There is, however, one gnawing equivocation in Hersh's otherwise forthright account. It comes in the first section, called "Torture at Abu Ghraib," which takes up over 70 pages of this 370-page book. Hersh blew the lid off the Abu Ghraib scandal last spring—the photographs, the Taguba report, the cover-ups, the links up the chain of command (which, in his book, he extends all the way up to the Oval Office). But he has always skirted a vital question: Does torture work?


Hersh is not alone in his evasiveness. Liberals have a tendency to accept, all too eagerly, the argument that torture is ineffective, that it doesn't yield useful information, that a tortured detainee will tell his inquisitors whatever they want to hear. This is an appealing argument. If it's true, we don't have to wrestle with any moral or legal dilemmas. If torture simply doesn't work, all those difficult questions are moot.

But it is, in fact, very likely that, under some circumstances, with some detainees, torture does produce, in the parlance of the trade, "actionable intelligence." Torture to produce a confession ("Yes, I am a terrorist") almost certainly is useless; at some point of pain, many people would confess to anything. But torture to elicit specific information (Who told you to do this? Where did the meeting take place? Who else is in your cell? What are they planning to blow up tomorrow?) sometimes will do—clearly, has done—the job. If it hasn't, many times over the centuries, then why do so many regimes engage in it? Some no doubt do it for the kicks, but they're not all purely sadists.

I do not mean to advocate torture. I mean only to suggest that it's time to start wrestling with those moral and legal dilemmas, to face them straightforwardly. If al-Qaida strikes the United States again, our leaders—whoever they are—will be tempted to resort to torture as a method of getting vital intelligence quickly, and we or they or someone should have mapped out crucial distinctions ahead of time: What is acceptable, what isn't; who should engage in it, who shouldn't; for what purposes is it legitimate, for what purposes isn't it; or whether we should decide, after an honest appraisal of its costs and benefits, that the whole business of torture—however you define it—is irredeemably beyond the pale.

It should be noted that the torture at Abu Ghraib appears to be utterly unjustified by any standards. Hersh clearly shows—and the Schlesinger report has confirmed—that the vast majority of the inmates at Abu Ghraib were common criminals or total innocents rounded up in random sweeps who were subjected to no screening before their horrendous ordeals began.

But what about the inmates elsewhere, many of whom really were, and are, al-Qaida operatives? Hersh refers to a highly classified "special-access program"—approved by President Bush and carried out by Secretary of Defense Donald Rumsfeld—that involved, as he puts it, "snatching or strong-arming suspected terrorists and questioning them in secret prison facilities in Singapore, Thailand, and Pakistan, among other sites." What about the torture—presumably there's torture of one sort or another—that went on there? For the moment, forget about whether such techniques are proper. That's a separate though no less important matter, to be dealt with after this question is answered: Did they produce useful intelligence?

At one point, Hersh suggests that they did. He writes that, early on in the Iraqi insurgency, detainees weren't giving their American interrogators any substantive information. Hersh quotes a "former intelligence official" on what Stephen Cambone, the assistant secretary of defense in charge of the operation, did in response in mid-2003:

Cambone says, I've got to crack this thing and I'm tired of working through the normal chain of command. I've got this apparatus set up—the black special-access program—and I'm going in hot. So he pulls the switch, and the electricity begins flowing last summer. And it's working. We're getting a picture of the insurgency in Iraq and the intelligence is flowing into the white world. We're getting good stuff.

Things went awry, Hersh's source goes on, because, when the order went out, too few soldiers were trained in what to do, and too many of their commanders looked the other way. But notice what the source said about the initial results: "[I]t's working. … We're getting good stuff." So, is the problem Cambone's orders or the fact that the U.S. military didn't have enough people with the brains or training to carry them out with restraint?