The Dark Art Of Interrogation

The Most Effective Way To Gather Intelligence and Thwart Terrorism Can Also Be A Direct Route Into Morally Repugnant Terrain. A Survey Of The Landscape Of Persuasion.

Mark Bowden
October 2003
TheAtlantic

Rawalpindi, Pakistan

...Some say that Sheikh Mohammed was captured months before the March 1 date announced by Pakistan's Inter-Services Intelligence (ISI). Abdul Qadoos, a pale, white-bearded alderman in this well-heeled neighborhood, told me that Sheikh Mohammed was not there "then or ever."

The official video of the takedown appears to have been faked. But the details are of minor importance. Whenever, wherever, and however it happened, nearly everyone now agrees that Sheikh Mohammed is in U.S. custody, and has been for some time.

In the first hours of his captivity the hood came off and a picture was taken. It shows a bleary-eyed, heavy, hairy, swarthy man with a full black moustache, thick eyebrows, a dark outline of beard on a rounded, shaved face, three chins, long sideburns, and a full head of dense, long, wildly mussed black hair. He stands before a pale tan wall whose paint is chipped, leaning slightly forward, like a man with his hands bound behind him, the low cut of his loose-fitting white T-shirt exposing matted curls of hair on his chest, shoulders, and back. He is looking down and to the right of the camera. He appears dazed and glum.

Sheikh Mohammed is a smart man. There is an anxious, searching quality to his expression in that first post-arrest photo. It is the look of a man awakened into nightmare. Everything that has given his life meaning, his role as husband and father, his leadership, his stature, plans, and ambitions, is finished. His future is months, maybe years, of imprisonment and interrogation; a military tribunal; and almost certain execution. You can practically see the wheels turning in his head, processing his terminal predicament. How will he spend his last months and years? Will he maintain a dignified, defiant silence? Or will he succumb to his enemy and betray his friends, his cause, and his faith?

If Sheikh Mohammed felt despair in those first hours, it didn't show. According to a Pakistani officer who sat in on an initial ISI questioning, the al-Qaeda sub-boss seemed calm and stoic. For his first two days in custody he said nothing beyond confirming his name. A CIA official says that Sheikh Mohammed spent those days "sitting in a trancelike state and reciting verses from the Koran." On the third day he is said to have loosened up. Fluent in the local languages of Urdu, Pashto, and Baluchi, he tried to shame his Pakistani interrogators, lecturing them on their responsibilities as Muslims and upbraiding them for cooperating with infidels.

"Playing an American surrogate won't help you or your country," he said. "There are dozens of people like me who will give their lives but won't let the Americans live in peace anywhere in the world." Asked if Osama bin Laden was alive, he said, "Of course he is alive." He spoke of meeting with bin Laden in "a mountainous border region" in December. He seemed smug about U.S. and British preparations for war against Saddam Hussein. "Let the Iraq War begin," he said. "The U.S. forces will be targeted inside their bases in the Gulf. I don't have any specific information, but my sixth sense is telling me that you will get the news from Saudi Arabia, Qatar, and Kuwait." Indeed, in the following months al-Qaeda carried out a murderous attack in Saudi Arabia.

On that third day, once more hooded, Sheikh Mohammed was driven to Chaklala Air Force base, in Rawalpindi, and turned over to U.S. forces. From there he was flown to the CIA interrogation center in Bagram, Afghanistan, and from there, some days later, to an "undisclosed location" (a place the CIA calls "Hotel California")—presumably a facility in another cooperative nation, or perhaps a specially designed prison aboard an aircraft carrier.

It doesn't much matter where, because the place would not have been familiar or identifiable to him. Place and time, the anchors of sanity, were about to come unmoored. He might as well have been entering a new dimension, a strange new world where his every word, move, and sensation would be monitored and measured; where things might be as they seemed but might not; where there would be no such thing as day or night, or normal patterns of eating and drinking, wakefulness and sleep; where hot and cold, wet and dry, clean and dirty, truth and lies, would all be tangled and distorted.

Intelligence and military officials would talk about Sheikh Mohammed's state only indirectly, and conditionally. But by the time he arrived at a more permanent facility, he would already have been bone-tired, hungry, sore, uncomfortable, and afraid—if not for himself, then for his wife and children, who had been arrested either with him or some months before, depending on which story you believe. He would have been warned that lack of cooperation might mean being turned over to the more direct and brutal interrogators of some third nation.

He would most likely have been locked naked in a cell with no trace of daylight. The space would be filled night and day with harsh light and noise, and would be so small that he would be unable to stand upright, to sit comfortably, or to recline fully. He would be kept awake, cold, and probably wet. If he managed to doze, he would be roughly awakened. He would be fed infrequently and irregularly, and then only with thin, tasteless meals. Sometimes days would go by between periods of questioning, sometimes only hours or minutes. The human mind craves routine, and can adjust to almost anything in the presence of it, so his jailers would take care that no semblance of routine developed.

Questioning would be intense—sometimes loud and rough, sometimes quiet and friendly, with no apparent reason for either. He would be questioned sometimes by one person, sometimes by two or three. The session might last for days, with interrogators taking turns, or it might last only a few minutes.

He would be asked the same questions again and again, and then suddenly be presented with something completely unexpected—a detail or a secret that he would be shocked to find they knew.

He would be offered the opportunity to earn freedom or better treatment for his wife and children. Whenever he was helpful and the information he gave proved true, his harsh conditions would ease. If the information proved false, his treatment would worsen.

On occasion he might be given a drug to elevate his mood prior to interrogation; marijuana, heroin, and sodium pentothal have been shown to overcome a reluctance to speak, and methamphetamine can unleash a torrent of talk in the stubbornest subjects, the very urgency of the chatter making a complex lie impossible to sustain.

These drugs could be administered surreptitiously with food or drink, and given the bleakness of his existence, they might even offer a brief period of relief and pleasure, thereby creating a whole new category of longing—and new leverage for his interrogators.

Deprived of any outside information, Sheikh Mohammed would grow more and more vulnerable to manipulation.

For instance, intelligence gleaned after successful al-Qaeda attacks in Kuwait and Saudi Arabia might be fed to him, in bits and pieces, so as to suggest foiled operations. During questioning he would be startled regularly by details about his secret organization—details drawn from ongoing intelligence operations, new arrests, or the interrogation of other captive al-Qaeda members.

Some of the information fed to him would be true, some of it false. Key associates might be said to be cooperating, or to have completely recanted their allegiance to jihad. As time went by, his knowledge would decay while that of his questioners improved. He might come to see once-vital plans as insignificant, or already known. The importance of certain secrets would gradually erode.

Isolated, confused, weary, hungry, frightened, and tormented, Sheikh Mohammed would gradually be reduced to a seething collection of simple needs, all of them controlled by his interrogators.

The key to filling all those needs would be the same: to talk.

Smacky-Face

We hear a lot these days about America's overpowering military technology; about the professionalism of its warriors; about the sophistication of its weaponry, eavesdropping, and telemetry; but right now the most vital weapon in its arsenal may well be the art of interrogation. To counter an enemy who relies on stealth and surprise, the most valuable tool is information, and often the only source of that information is the enemy himself.

Men like Sheikh Mohammed who have been taken alive in this war are classic candidates for the most cunning practices of this dark art. Intellectual, sophisticated, deeply religious, and well trained, they present a perfect challenge for the interrogator. Getting at the information they possess could allow us to thwart major attacks, unravel their organization, and save thousands of lives.

They and their situation pose one of the strongest arguments in modern times for the use of torture.

Torture is repulsive. It is deliberate cruelty, a crude and ancient tool of political oppression. It is commonly used to terrorize people, or to wring confessions out of suspected criminals who may or may not be guilty. It is the classic shortcut for a lazy or incompetent investigator. Horrifying examples of torturers' handiwork are catalogued and publicized annually by Amnesty International, Human Rights Watch, and other organizations that battle such abuses worldwide. One cannot help sympathizing with the innocent, powerless victims showcased in their literature.

But professional terrorists pose a harder question. They are lockboxes containing potentially life-saving information. Sheikh Mohammed has his own political and religious reasons for plotting mass murder, and there are those who would applaud his principled defiance in captivity. But we pay for his silence in blood.

The word "torture" comes from the Latin verb torquere, "to twist." Webster's New World Dictionary offers the following primary definition: "The inflicting of severe pain to force information and confession, get revenge, etc."

Note the adjective "severe," which summons up images of the rack, thumbscrews, gouges, branding irons, burning pits, impaling devices, electric shock, and all the other devilish tools devised by human beings to mutilate and inflict pain on others. All manner of innovative cruelty is still commonplace, particularly in Central and South America, Africa, and the Middle East.

...Then there are methods that, some people argue, fall short of torture. Called "torture lite," these include sleep deprivation, exposure to heat or cold, the use of drugs to cause confusion, rough treatment (slapping, shoving, or shaking), forcing a prisoner to stand for days at a time or to sit in uncomfortable positions, and playing on his fears for himself and his family.

Although excruciating for the victim, these tactics generally leave no permanent marks and do no lasting physical harm.

The Geneva Convention makes no distinction: it bans any mistreatment of prisoners. But some nations that are otherwise committed to ending brutality have employed torture lite under what they feel are justifiable circumstances.

In 1987 Israel attempted to codify a distinction between torture, which was banned, and "moderate physical pressure," which was permitted in special cases. Indeed, some police officers, soldiers, and intelligence agents who abhor "severe" methods believe that banning all forms of physical pressure would be dangerously naive.

Few support the use of physical pressure to extract confessions, especially because victims will often say anything (to the point of falsely incriminating themselves) to put an end to pain.

But many veteran interrogators believe that the use of such methods to extract information is justified if it could save lives—whether by forcing an enemy soldier to reveal his army's battlefield positions or forcing terrorists to betray the details of ongoing plots.

As these interrogators see it, the well-being of the captive must be weighed against the lives that might be saved by forcing him to talk. A method that produces life-saving information without doing lasting harm to anyone is not just preferable; it appears to be morally sound.

Hereafter I will use "torture" to mean the more severe traditional outrages, and "coercion" to refer to torture lite, or moderate physical pressure.

...It is likely that some captured terrorists' names and arrests have not yet been revealed; people may be held for months before their "arrests" are staged.

Once a top-level suspect is publicly known to be in custody, his intelligence value falls. His organization scatters, altering its plans, disguises, cover stories, codes, tactics, and communication methods.

The maximum opportunity for intelligence gathering comes in the first hours after an arrest, before others in a group can possibly know that their walls have been breached. Keeping an arrest quiet for days or weeks prolongs this opportunity.

If March 1 was in fact the day of Sheikh Mohammed's capture, then the cameras and the headlines were an important intelligence failure.

The arrest of the senior al-Qaeda figure Abu Anas Liby, in Sudan in February of 2002, was not made public until a month later, when U.S. efforts to have him transferred to custody in Egypt were leaked to the Sunday Times of London.

...All these suspects are questioned rigorously, but those in the top ranks get the full coercive treatment. And if official and unofficial government reports are to be believed, the methods work.

In report after report hard-core terrorist leaders are said to be either cooperating or, at the very least, providing some information—not just vague statements but detailed, verifiable, useful intelligence.

...How much of this can be believed? Are such reports wishful thinking, or deliberate misinformation?

...It would make sense to claim that top al-Qaeda leaders had caved under questioning even if they had not...Word that they had been broken would demoralize their followers, and would encourage lower-ranking members of their organization to talk; if their leaders had given in, why should they hold out?

To some, all this jailhouse cooperation smells concocted. "I doubt we're getting very much out of them, despite what you read in the press," says a former CIA agent with experience in South America...

Bill Cowan, a retired Marine lieutenant colonel who conducted interrogations in Vietnam, says, "I don't see the proof in the pudding. If you had a top leader like Mohammed talking, someone who could presumably lay out the whole organization for you, I think we'd be seeing sweeping arrests in several different countries at the same time. Instead what we see is an arrest here, then a few months later an arrest there."

These complaints are all from people who have no qualms about using torture to get information from men like Sheikh Mohammed. Their concern is that merely using coercion amounts to handling terrorists with kid gloves.

...Is the United States torturing prisoners? Three inmates have died in U.S. custody in Afghanistan, and reportedly eighteen prisoners at Guantánamo have attempted suicide; one prisoner there survived after hanging himself but remains unconscious and is not expected to revive.

Shah Muhammad, a twenty-year-old Pakistani who was held at Camp X-Ray for eighteen months, told me that he repeatedly tried to kill himself in despair. "They were driving me crazy," he said.

Public comments by Administration officials have fueled further suspicion. An unnamed intelligence official told The Wall Street Journal, "What's needed is a little bit of smacky-face. Some al-Qaeda just need some extra encouragement."

Then there was the bravado of Cofer Black, the counterterrorism coordinator, in his congressional testimony last year. A pudgy, balding, round-faced man with glasses, who had served with the CIA before taking the State Department position...Describing the clandestine war, Black said, "This is a highly classified area. All I want to say is that there was 'before 9/11' and 'after 9/11.' After 9/11 the gloves came off." He was referring to the overall counterterrorism effort, but in the context of detained captives the line was suggestive.

A story in December of 2002 by the Washington Post reporters Dana Priest and Barton Gellman described the use of "stress and duress" techniques at Bagram, and an article in The New York Times in March described the mistreatment of prisoners there.

..The treatment alleged falls clearly within the category of torture and other cruel, inhuman or degrading treatment or punishment which is absolutely prohibited under international law ... [We] urge the US government to instigate a full, impartial inquiry into the treatment of detainees at the Bagram base and to make the findings public. We further urge the government to make a clear public statement that torture and other cruel, inhuman or degrading treatment of suspects in its custody will not be tolerated under any circumstances, and that anyone found to have engaged in abuses will be brought to justice.

In June, at the urging of Amnesty and other groups, President Bush reaffirmed America's opposition to torture, saying, "I call on all governments to join with the United States and the community of law-abiding nations in prohibiting, investigating, and prosecuting all acts of torture ... and we are leading this fight by example."

A slightly more detailed response had been prepared two months earlier by the Pentagon's top lawyer, William J. Haynes II, in a letter to Kenneth Roth, the executive director of Human Rights Watch. (My requests for interviews on this subject with the Pentagon, the White House, and the State Department were declined.) Haynes wrote,

The United States questions enemy combatants to elicit information they may possess that could help the coalition win the war and forestall further terrorist attacks upon the citizens of the United States and other countries. As the President reaffirmed recently to the United Nations High Commissioner for Human Rights, United States policy condemns and prohibits torture. When questioning enemy combatants, US personnel are required to follow this policy and applicable laws prohibiting torture.

As we will see, Haynes's choice of words was careful—and telling. The human-rights groups and the Administration are defining terms differently. Yet few would argue that getting Sheikh Mohammed to talk doesn't serve the larger interests of mankind. So before tackling the moral and legal questions raised by interrogation, perhaps the first question should be, What works?

Acid Tests and Monkey Orgasms

The quest for surefire methods in the art of interrogation has been long, ugly, and generally fruitless. Nazi scientists experimented on concentration-camp inmates, subjecting them to extremes of hot and cold, to drugs, and to raw pain in an effort to see what combination of horrors would induce cooperation. The effort produced a long list of dead and maimed, but no reliable ways of getting people to talk.

In 1953 John Lilly, of the National Institute of Mental Health, discovered that by placing electrodes inside the brain of a monkey, he could stimulate pain, anger, fear—and pleasure. He placed one inside the brain of a male monkey and gave the monkey a switch that would trigger an immediate erection and orgasm. (The monkey hit the switch roughly every three minutes, thus confirming the gender stereotype.)

The idea of manipulating a brain from the inside promptly attracted the interest of the CIA, which foresaw, among other things, the possibility of sidestepping a reluctant informant's self-defenses. But Lilly dropped the line of research, pointing out that merely inserting the electrodes caused brain damage.

These experiments and others are recorded in detail in John Marks's somewhat overheated book The Search for the "Manchurian Candidate": The CIA and Mind Control (1979) and in George Andrews's book mkULTRA: The CIA's Top Secret Program in Human Experimentation and Behavior Modification (2001). Andrews summarized information revealed in congressional probes of CIA excesses. Marks was more sensational. In the spirit of the times, he tended to interpret the Agency's interest in behavioral science, hypnosis, and mind-altering drugs as a scheme to create zombie-like secret agents, although it appears that the real goal was to make people talk.

There was a lot of hope for LSD. Discovered by accident in a Swiss pharmaceutical lab in 1943, it produced powerful mind-altering effects in very small doses. It was more powerful than mescaline, which had its own adherents, and could easily be administered without the victim's knowledge, slipped into food or drink. The hope was that an informant in such an artificially open-minded state would lose sight of his goals and sense of loyalty and become putty in the hands of a skilled interrogator.

Studies on LSD began at a number of big universities, and as word of the drug's properties spread, it started to attract a broad range of interest. Theologians, scholars, and mental-health workers visited the Maryland Psychiatric Research Institute, just outside Baltimore, to turn on and tune in, and similar programs began in Boston, New York, Chicago, and other cities. Almost twenty years ago I interviewed a number of those who took part in these experiments; all of them were apparently motivated only by professional curiosity. The CIA's role was kept quiet.

But the most notorious of its efforts at LSD experimentation involved Frank Olson, an Army scientist who was dosed without his knowledge and subsequently committed suicide. The U.S. Army conducted field tests of LSD as an interrogation tool in 1961 (Operation Third Chance), dosing nine foreigners and an American soldier named James Thornwell, who had been accused of stealing classified documents. Thornwell subsequently sued the government and was awarded $650,000. Most of these efforts led to little more than scandal and embarrassment. The effects of the drug were too wildly unpredictable to make it useful in interrogation. It tended to amplify the sorts of feelings that inhibit cooperation.

Fear and anxiety turned into terrifying hallucinations and fantasies, which made it more difficult to elicit secrets, and added a tinge of unreality to whatever information was divulged. LSD may have unlocked the mind in some esoteric sense, but secrets tended to ride out the trip intact.

Experiments were also conducted with heroin and psychedelic mushrooms, neither of which reliably delivered up the secrets of men's souls. Indeed, drugs seemed to enhance some people's ability to be deceptive. Scopolamine held out some early hope, but it often induced hallucinations. Barbiturates were promising, and were already used effectively by psychiatrists to help with therapy. Some researchers advocated electroshock treatments, to, as it were, blast information from a subject's brain. Drugs such as marijuana, alcohol, and sodium pentothal can lower inhibitions, but they do not erase deep-seated convictions. And the more powerful the drug, the less reliable the testimony.

According to my intelligence sources, drugs are today sometimes used to assist in critical interrogations, and the preferred ones are methamphetamines tempered with barbiturates and cannabis. These tools can help, but they are only as effective as the interrogator.

Better results seemed to come from sensory deprivation and solitary confinement.

For most people severe sensory deprivation quickly becomes misery; the effects were documented in the notorious 1963 CIA manual on interrogation, called the Kubark Manual. It remains the most comprehensive and detailed explanation in print of coercive methods of questioning—given the official reluctance to discuss these matters or put them in writing, because such things tend to be both politically embarrassing and secret.

...Unearthed in 1997, through the Freedom of Information Act, by the Baltimore Sun reporters Gary Cohn, Ginger Thompson, and Mark Matthews, the Kubark Manual reveals the CIA's insights into the tougher methods employed by the military and intelligence agencies.

Much of the practice and theory it details is also found unchanged in the 1983 Human Resource Exploitation Training Manual, usually known as the Honduras Manual—which the CIA had tried to soften with a hasty edit prior to releasing it. The manual was shaken loose at the same time by Cohn and Thompson. And the more summary discussions of technique in later U.S. Army manuals on interrogation, including the most recent, also clearly echo Kubark.

If there is a bible of interrogation, it is the Kubark Manual.

The manual cites a 1954 study at the National Institute of Mental Health (again led by John Lilly) in which two volunteers attempted to see how long they could stay suspended in water wearing blackout masks and hearing only the sound of their own breathing and "some faint sounds of water from the piping." Neither lasted more than three hours.

According to the study, "Both passed quickly from normally directed thinking through a tension resulting from unsatisfied hunger for sensory stimuli and concentration upon the few available sensations to provide reveries and fantasies and eventually to visual imagery somewhat resembling hallucinations."

John Marks reported in his book that in a similar experiment a volunteer kicked his way out of a sensory-deprivation box after an hour of tearful pleas for release had been ignored.

The Summary Of Another Experiment Concluded, The Results Confirmed Earlier Findings:

1) The deprivation of sensory stimuli induces stress;
2) the stress becomes unbearable for most subjects;
3) the subject has a growing need for physical and social stimuli; and
4) some subjects progressively lose touch with reality, focus inwardly, and produce delusions, hallucinations, and other pathological effects.

But these effects didn't trouble everyone. One man's misery is another man's mind-altering experience. Some people found they liked sensory-deprivation tanks; indeed, in later years people would pay for a session in one. Lilly was fond of injecting himself with LSD and then closing himself off in his tank—a series of experiments made famous in the 1980 film Altered States.

In Canada a scientist put a fifty-two-year-old woman identified only as Mary C. in a sensory-deprivation chamber for thirty-five days. She never asked to be let out.

One thing all these experiments made clear was that no matter what drugs or methods were applied, the results varied from person to person. So another major area of inquiry involved trying to define certain broad personality types and discover what methods would work best for each.

The groups were ridiculously general—the Kubark Manual lists "The Orderly-Obstinate Character," "The Greedy-Demanding Character," "The Anxious, Self-Centered Character"—and the prescriptions for questioning them tended to vary little and were sometimes silly (the advice for questioning an Orderly-Obstinate Character recommends doing so in a room that is especially neat). The categories were useless. Everyone, and every situation, is different; some people begin a day greedy and demanding and end it orderly and obstinate.

The one constant in effective interrogation, it seems, is the interrogator. And some interrogators are just better at it than others.

"You want a good interrogator?" Jerry Giorgio, the New York Police Department's legendary third-degree man, asks. "Give me somebody who people like, and who likes people. Give me somebody who knows how to put people at ease. Because the more comfortable they are, the more they talk, and the more they talk, the more trouble they're in—the harder it is to sustain a lie."

Though science has made contributions, interrogation remains more art than science. Like any other subject, Sheikh Mohammed presented his interrogators with a unique problem. The critical hub of a worldwide secret network, he had a potential road map in his head to the whole shadow world of jihad. If he could be made to talk, to reveal even a few secrets, what an intelligence bonanza that would be! Here was a man who lived to further his cause by whatever means, who saw himself as morally, spiritually, and intellectually superior to the entire infidel Western world, a man for whom capitulation meant betraying not just his friends and his cherished cause but his very soul.

What Would Make A man Like That Decide To Talk?

Alligator Clips

Bill Cowan spent three and a half years fighting the war in Vietnam. He was a young Marine captain assigned to the Rung Sat Special Zone, a putrid swamp that begins just south of Saigon...
So when he captured a Vietcong soldier who could warn of ambushes and lead them to hidden troops but who refused to speak, wires were attached to the man's scrotum with alligator clips and electricity was cranked out of a 110-volt generator.

"It worked like a charm," Cowan told me. "The minute the crank started to turn, he was ready to talk. We never had to do more than make it clear we could deliver a jolt. It was the fear more than the pain that made them talk."

Fear works. It is more effective than any drug, tactic, or torture device.

According to unnamed scientific studies cited by the Kubark Manual (it is frightening to think what these experiments might have been), most people cope with pain better than they think they will. As people become more familiar with pain, they become conditioned to it. Those who have suffered more physical pain than others—from being beaten frequently as a child, for example, or suffering a painful illness—may adapt to it and come to fear it less. So once interrogators resort to actual torture, they are apt to lose ground.

"The threat of coercion usually weakens or destroys resistance more effectively than coercion itself," the manual says.

The threat to inflict pain, for example, can trigger fears more damaging than the immediate sensation of pain ... Sustained long enough, a strong fear of anything vague or unknown induces regression, whereas the materialization of the fear, the infliction of some form of punishment, is likely to come as a relief. The subject finds that he can hold out, and his resistances are strengthened.

Furthermore, if a prisoner is subjected to pain after other methods have failed, it is a signal that the interrogation process may be nearing an end. "He may then decide that if he can just hold out against this final assault, he will win the struggle and his freedom," the manual concludes.

Even if severe pain does elicit information, it can be false, which is particularly troublesome to interrogators seeking intelligence rather than a confession. Much useful information is time-sensitive, and running down false leads or arresting innocents wastes time.

By similar logic, the manual discourages threatening a prisoner with death. As a tactic "it is often found to be worse than useless," the manual says, because the sense of despair it induces can make the prisoner withdraw into depression—or, in some cases, see an honorable way out of his predicament.

Others Disagree.

"I'll tell you how to make a man talk," a retired Special Forces officer says. "You shoot the man to his left and the man to his right. Then you can't shut him up."

John Dunn found the truth to be a little more complicated. In his case the threat of execution forced him to bend but not break.

He was a U.S. Army intelligence officer in the Lam Dong Province of Vietnam, in March of 1968, when he was captured by the Vietcong. He and other captives were marched for weeks to a prison camp in the jungle, where initially he was treated quite well. The gentle treatment lulled him, Dunn says, and contributed to his shock when, in his first interrogation session, he was calmly told, "We don't need you. We did not sign the Geneva Convention, and you are not considered a prisoner of war anyway. You are a war criminal. If you don't cooperate with us, you will be executed."

...His training for captivity had been basic. He had been instructed to tell his captors only his name, rank, and serial number. Anything beyond that was considered a breach of duty—a betrayal of his country, his role as a soldier, and his personal honor.

Faced with death, Dunn weighed his devotion to this simple code. He felt it was unrealistic. He wrestled to come up with a solution that would keep him alive without completely compromising his dignity. He figured there were certain details about his life and service that were not worth dying to protect. Some things needed to be kept secret, and others did not.

Struggling with shame, he decided to answer any questions that did not intrude on that closed center of secrecy. He would not tell them he was an intelligence officer. ("Not out of patriotism," he says. "Out of fear, strictly self-preservation.") He would not reveal accurate details about fortifications around his company's headquarters, in Di Linh. He would not tell them about upcoming plans, such as the Phoenix Program (an assassination program targeting Vietcong village leaders), and above all, he would not make any public statements. But he would talk. The threat of execution in his case was not "worse than useless." It shook Dunn to his core.

In a subsequent session he talked, but not enough to satisfy his captors. Again and again he refused to make a public statement. Starved, sore, and still frightened, Dunn was told, "You will be executed. After dark."

When the sun set, the interrogator, his aide, and the camp commander came for Dunn with a group of soldiers. They unlocked his chain, and he carried it as they led him away from the encampment into the jungle. They stopped in front of a pit they had dug for his grave and put a gun to his head. The interrogator gave him one more chance to agree to make a statement.

"No," Dunn said. He had gone as far as he was willing to go.

"Why do you want to die?" he was asked.

"If I must, I must," Dunn said. He felt resigned. He waited to be killed.

"You will not be executed," the camp commander said abruptly, and that was that.

Judging by Dunn's experience, the threat of death may be valuable to an interrogator as a way of loosening up a determined subject. But, as with pain, the most important factor is fear. An unfrightened prisoner makes an unlikely informer.

If there is an archetype of the modern interrogator, it is Michael Koubi. The former chief interrogator for Israel's General Security Services, or Shabak, Koubi probably has more experience than anyone else in the world in the interrogation of hostile Arab prisoners, some of them confirmed terrorists and religious fanatics—men, he says, "whose hatred of the Jews is unbridgeable."

...For decades he has been experimenting with captive human beings, cajoling, tricking, hurting, threatening, and spying on them, steadily upping the pressure, looking for cracks at the seams.

...There are still many things he is not free to discuss, but he is happy to talk about his methods. He is very proud of his skills, among them an ability to speak Arabic so fluently that he can adopt a multitude of colloquial flavors.

Koubi came to his career as an interrogator through his love of language. He grew up speaking Hebrew, Yiddish, and Arabic, and he studied Arabic in high school, working to master its idiom and slang.

He also had a knack for reading the body language and facial expressions of his subjects, and for sensing a lie.

He is a skilled actor who could alternately befriend or intimidate a subject, sometimes turning on a dime.

Blending these skills with the tricks he had learned over the years for manipulating people, Koubi didn't just question his subjects, he orchestrated their emotional surrender.

...Charm has always been as important to interrogation work as toughness or cruelty—perhaps more important. Koubi says that only in rare instances did he use force to extract information from his subjects; in most cases it wasn't necessary.

"People change when they get to prison," Koubi says. "They may be heroes outside, but inside they change. The conditions are different. People are afraid of the unknown. They are afraid of being tortured, of being held for a long time. Try to see what it is like to sit with a hood over your head for four hours, when you are hungry and tired and afraid, when you are isolated from everything and have no clue what is going on."

When the captive believes that anything could happen—torture, execution, indefinite imprisonment, even the persecution of his loved ones—the interrogator can go to work.

Under pressure, he says, nearly everyone looks out first and foremost for No. 1. What's more, a very large part of who a man is depends on his circumstances. No matter who he is before his arrest, his sense of self will blur in custody. Isolation, fear, and deprivation force a man to retreat, to reorient himself, and to reorder his priorities.

For most men, Koubi says, the hierarchy of loyalty under stress is:
1) self,
2) group,
3) family,
4) friends.

In other words, even the most dedicated terrorist (with very rare exceptions), when pushed hard enough, will act to preserve and protect himself at the expense of anyone or anything else. "There's an old Arab saying," Koubi says. "'Let one hundred mothers cry, but not my mother—but better my mother than me.'"

With older men the priorities shift slightly. In middle age the family often overtakes the group (the cause) to become the second most important loyalty. Young men tend to be fiercely committed and ambitious, but older men—even men with deeply held convictions, men admired and emulated by their followers—tend to have loves and obligations that count for more.

Age frays idealism, slackens zeal, and cools ferocity. Abstractions lose ground to wife, children, and grandchildren.

"Notice that the leaders of Hamas do not send their own sons and daughters, and their own grandchildren, to blow themselves up," Koubi says.

So it is often the top-level men...who are easier to crack. Koubi believes that having the al-Qaeda leader's wife and children in custody gives his interrogators powerful leverage.

The key is to find a man's weak point and exploit it.

For Koubi the three critical ingredients of that process are preparation, investigation, and theater.

Preparing a subject for interrogation means softening him up. Ideally, he has been pulled from his sleep—like Sheikh Mohammed—early in the morning, roughly handled, bound, hooded (a coarse, dirty, smelly sack serves the purpose perfectly), and kept waiting in discomfort, perhaps naked in a cold, wet room, forced to stand or to sit in an uncomfortable position. He may be kept awake for days prior to questioning, isolated and ill-fed. He may be unsure where he is, what time of day it is, how long he has been or will be held. If he is wounded, as Abu Zubaydah was, pain medication may be withheld; it is one thing to cause pain, another to refuse to relieve it.

Mousa Khoury, a Palestinian businessman, knows the drill all too well. A slender thirty-four-year-old man with a black goatee and thinning hair, he is bitter about the Israeli occupation and his experiences in custody. He has been arrested and interrogated six times by Israeli forces. He was once held for seventy-one days.

"My hands were cuffed behind my back, and a potato sack was over my head," he says. "My legs were cuffed to a tiny chair. The chair's base is ten centimeters by twenty centimeters. The back is ten centimeters by ten centimeters. It is hard wood. The front legs are shorter than the back ones, so you are forced to slide forward in it, only your hands are bound in the back. If you sit back, the back of the chair digs into the small of your back. If you slump forward, you are forced to hang by your hands. It is painful. They will take you to the toilet only after screaming a request one hundred times." He could think about only one thing: how to make the treatment stop. "Your thoughts go back and forth and back and forth, and you can no longer have a normal stream of consciousness," he says.

Preparing an interrogator means arming him beforehand with every scrap of information about his subject.

U.S. Army interrogation manuals suggest preparing a thick "dummy file" when little is known, to make it appear that the interrogator knows more than he does.

Nothing rattles a captive more than to be confronted with a fact he thought was secret or obscure. It makes the interrogator seem powerful, all-knowing. A man's sense of importance is wounded, and he is slower to lie, because he thinks he might be caught at it.

There are many ways that scraps of information—gathered by old-fashioned legwork or the interrogation of a subject's associates—can be leveraged by a clever interrogator into something new. Those scraps might be as simple as knowing the names of a man's siblings or key associates, the name of his girlfriend, or a word or phrase that has special meaning to his group.

Uncovering privileged details diminishes the aura of a secret society, whether it is a social club, a terrorist cell, or a military unit. Joining such a group makes an individual feel distinct, important, and superior, and invests even the most mundane of his activities with meaning.

An interrogator who penetrates that secret society, unraveling its shared language, culture, history, customs, plans, and pecking order, can diminish its hold on even the staunchest believer.

Suspicion that a trusted comrade has betrayed the group—or the subject himself—undermines the sense of a secretly shared purpose and destiny.

Armed with a few critical details, a skilled interrogator can make a subject doubt the value of information he has been determined to withhold.

It is one thing to suffer in order to protect a secret, quite another to cling to a secret that is already out.

This is how a well-briefed interrogator breaches a group's defenses.

Koubi believes that the most important skill for an interrogator is to know the prisoner's language. Working through interpreters is at best a necessary evil. Language is at the root of all social connections, and plays a critical role in secret societies like Hamas and al-Qaeda. A shared vocabulary or verbal shorthand helps to cement the group.

"I try to create the impression that I use his mother tongue even better than he does," Koubi says. "No accent, no mistaken syntax. I speak to him like his best friend speaks to him. I might ask him a question about a certain word or sentence or expression, how it is used in his culture, and then demonstrate that I know more about it than he does. This embarrasses him very much."

Once a prisoner starts to talk, rapid follow-up is needed to sort fact from fiction, so that the interrogator knows whether his subject is being cooperative or evasive, and can respond accordingly.

Interrogation sessions should be closely observed (many rooms designed for this purpose have one-way mirrors), and in a well-run unit a subject's words can sometimes be checked out before the session is over.

Being caught so quickly in a lie demonstrates the futility of playing games with the interrogator, and strengthens his hand. It shames and rattles the subject.

When information checks out, the interrogator can home in for more details and open up new avenues of exploration.

Religious extremists are the hardest cases. They ponder in their own private space, performing a kind of self-hypnosis. They are usually well educated. Their lives are financially and emotionally tidy. They tend to live in an ascetic manner, and to look down on nonbelievers. They tend to be physically and mentally strong, and not to be influenced by material things—by either the incentives or the disincentives available in prison. Often the rightness of their cause trumps all else, so they can commit any outrage—lie, cheat, steal, betray, kill—without remorse. Yet under suffi-cient duress, Koubi says, most men of even this kind will eventually break—most, but not all. Some cannot be broken.

"They are very rare," he says, "but in some cases the more aggressive you get, and the worse things get, the more these men will withdraw into their own world, until you cannot reach them."

Mousa Khoury, the Palestinian businessman who has been interrogated six times, claims that he never once gave in to his jailers. Koubi has no particular knowledge of Khoury's case, but he smiles his crooked, knowing smile and says, "If someone you meet says he was held by our forces and did not cooperate at all, you can bet he is lying. In some cases men who are quite famous for their toughness were the most helpful to us in captivity."

Interrogation Is Also Highly Theatrical. The Kubark Manual Is Very Particular About Setting The Stage:

The room in which the interrogation is to be conducted should be free of distractions. The colors of the walls, ceiling, rugs, and furniture should not be startling. Pictures should be missing or dull. Whether the furniture should include a desk depends not upon the interrogator's convenience but rather upon the subject's anticipated reaction to the connotations of superiority and officialdom. A plain table may be preferable. An overstuffed chair for the use of the interrogatee is sometimes preferable to a straight-backed, wooden chair because if he is made to stand for a lengthy period or is otherwise deprived of physical comfort, the contrast is intensified and increased disorientation results.

The manual goes on to recommend lighting that shines brightly in the face of the subject and leaves the interrogator in shadow. There should be no phone or any other means of contact with those outside the room, to enhance concentration and the subject's feeling of confinement.

In Koubi's experience it was sometimes helpful to have associates loudly stage a torture or beating session in the next room.

In old CIA interrogation training, according to Bill Wagner, a retired agent, it was recommended that mock executions take place outside the interrogation room.

A good interrogator is a deceiver. One of Koubi's tricks was to walk into a hallway lined with twenty recently arrested, hooded, uncomfortable, hungry, and fearful men, all primed for interrogation, and shout commandingly,

"Okay, who wants to cooperate with me?" Even if no hands, or only one hand, went up, he would say to the hooded men, "Okay, good. Eight of you. I'll start with you, and the others will have to wait."

Believing that others have capitulated makes doing so oneself much easier. Often, after this trick, many of the men in the hall would cooperate.

Men are herd animals, and prefer to go with the flow, especially when moving in the other direction is harsh.

In one case Koubi had information suggesting that two men he was questioning were secretly members of a terrorist cell, and knew of an impending attack. They were tough men, rural farmers, very difficult to intimidate or pressure, and so far neither man had admitted anything under questioning. Koubi worked them over individually for hours. With each man he would start off by asking friendly questions and then grow angrier and angrier, accusing the subject of withholding something. He would slap him, knock him off his chair, set guards on him, and then intervene to pull them off. Then he would put the subject back in the chair and offer him a cigarette, lightening the mood.

"Let him see the difference between the two atmospheres, the hostile one and the friendly one," Koubi says. Neither man budged.

Finally Koubi set his trap. He announced to one of the men that his interrogation was over. The man's associate, hooded, was seated in the hallway outside the room. "We are going to release you," Koubi said.

"We are pleased with your cooperation. But first you must do something for me. I am going to ask you a series of questions, just a formality, and I need you to answer 'Yes' in a loud, clear voice for the recorder."

Then, in a voice loud enough for the hooded man outside in the hall to hear, but soft enough so that he couldn't make out exactly what was being said, Koubi read off a long list of questions, reviewing the prisoner's name, age, marital status, date of capture, length of detainment, and so forth. These were regularly punctuated by the prisoner's loud and cooperative "Yes."

The charade was enough to convince the man in the hall that his friend had capitulated.

Koubi dismissed the first man and brought in the second. "There's no more need for me to question you," Koubi said. "Your friend has confessed the whole thing." He offered the second prisoner a cigarette and gave him a good meal. He told him that the information provided by his friend virtually ensured that they would both be in prison for the rest of their lives ... unless, he said, the second prisoner could offer him something, anything, that would dispose the court to leniency in his case.

Convinced that his friend had already betrayed them both, the second prisoner acted promptly to save himself. "If you want to save Israeli lives, go immediately," he told Koubi. "My friends went with a car to Yeshiva Nehalim [a religious school]. They are going to kidnap a group of students ..." The men were found in Erez, and the operation was foiled.

There are other methods of keeping a prisoner confused and off balance, such as rapidly firing questions at him, cutting off his responses in mid-sentence, asking the same questions over and over in different order, and what the manual calls the "Silent" technique, in which the interrogator "says nothing to the source, but looks him squarely in the eye, preferably with a slight smile on his face."

The manual advises forcing the subject to break eye contact first. "The source will become nervous, begin to shift around in his chair, cross and recross his legs, and look away," the manual says. "When the interrogator is ready to break silence, he may do so with some quite nonchalant questions such as 'You planned this operation a long time, didn't you?

Was it your idea?'"

Then There Is "Alice In Wonderland."

The aim of the Alice in Wonderland or confusion technique is to confound the expectations and conditioned reactions of the interrogatee ... The confusion technique is designed not only to obliterate the familiar but to replace it with the weird ... Sometimes two or more questions are asked simultaneously. Pitch, tone, and volume of the interrogators' voices are unrelated to the import of the questions. No pattern of questions and answers is permitted to develop, nor do the questions themselves relate logically to each other.

If this technique is pursued patiently, the manual says, the subject will start to talk "just to stop the flow of babble which assails him."

Easily the most famous routine is "Good Cop/Bad Cop," in which one interrogator becomes the captive's persecutor and the other his friend.

A lesser-known but equally effective technique is "Pride and Ego," "Ego Up/Ego Down," or (as the more pretentious Kubark Manual puts it) "Spinoza and Mortimer Snerd," in which the "Ego Down" part involves repeatedly asking questions that the interrogator knows the subject cannot answer.

The subject is continually berated or threatened ("How could you not know the answer to that?") and accused of withholding, until, at long last, he is asked a simple question that he can answer.

An American POW subjected to this technique has said, "I know it seems strange now, but I was positively grateful to them when they switched to a topic I knew something about."

CIA psychologists have tried to develop an underlying theory for interrogation—namely, that the coercive methods induce a gradual "regression" of personality. But the theory is not convincing.

Interrogation simply backs a man into a corner. It forces difficult choices, and dangles illusory avenues of escape.

A skillful interrogator knows which approach will best suit his subject; and just as he expertly applies stress, he continually opens up these avenues of escape or release. This means understanding what, at heart, is stopping a subject from cooperating.

If it is ego, that calls for one method. If it is fear of reprisal or of getting into deeper trouble, another method might work best.

For most captives a major incentive to keep quiet is simply pride. Their manhood is being tested, not just their loyalty and conviction. Allowing the subject to save face lowers the cost of capitulation, so an artful interrogator will offer persuasive rationales for giving in: others already have, or the information is already known.

Drugs, if administered with the subject's knowledge, are helpful in this regard. If a subject believes that a particular drug or "truth serum" renders him helpless, he is off the hook. He cannot be held accountable for giving in.

A study cited in George Andrews's book MKULTRA found that a placebo—a simple sugar pill—was as effective as an actual drug up to half of the time.

Koubi layered his deception so thick that his subjects never knew exactly when their interrogation ended.

After questioning, captives usually spent time in a regular prison. The Israelis had bugged the prison with a system that was disguised well enough to appear hidden but not well enough to avoid discovery. In this way prisoners were led to believe that only certain parts of the prison were bugged. In fact, all of the prison was bugged. Conversations between prisoners could be overheard anywhere, and were closely monitored. They were an invaluable source of intelligence.

Prisoners who could hold out through the most intense interrogation often let their guard down later when talking to comrades in jail.

To help such inadvertent confessions along, Koubi had yet another card to play. Whenever an interrogated subject was released to the general prison, after weeks of often grueling questioning, he was received with open arms by fellow Palestinians who befriended him and congratulated him for having endured interrogation. He was treated like a hero.

He was fed, nursed, even celebrated. What he didn't know was that his happy new comrades were working for Koubi.

Koubi calls them "birdies." They were Palestinians who, offered an incentive such as an opportunity to settle with their families in another country, had agreed to cooperate with Shabak.

Some days or weeks after welcoming the new prisoner into their ranks, easing his transition into the prison, they would begin to ask questions. They would debrief the prisoner on his interrogation sessions. They would say, "It is very important for those on the outside to know what you told the Israelis and what you didn't tell them. Tell us, and we will get the information to those on the outside who need to know."

Even prisoners who had managed to keep important secrets from Koubi spilled them to his birdies.

"The amazing thing is that by now the existence of the birdies is well known," Koubi says, "and yet the system still works. People come out of interrogation, go into the regular prison, and then tell their darkest secrets. I don't know why it still works, but it does."

Big Daddy Uptown

Most professional interrogators work without the latitude given the CIA, the FBI, or the military in the war on terror. A policeman's subjects all have to be read their Miranda rights, and cops who physically threaten or abuse suspects—at least nowadays—may find themselves in jail.

Jerry Giorgio, the legendary NYPD interrogator, has operated within these rules for nearly forty years. He may not know all the names of the CIA and military techniques, but he has probably seen most of them at work. Known as "Big Daddy Uptown," Giorgio now works for the New York County district attorney in a cramped office in Lower Manhattan that he shares with two others. He is a big man with a big voice, thinning gray hair, a broad belly, and wide, searching greenish-brown eyes. He is considered a wizard by his former colleagues in the NYPD. "All of us of a certain generation came out of the Jerry Giorgio school of interrogation," says John Bourges, a recently retired Manhattan homicide detective.

"Everybody knows the Good Cop/Bad Cop routine, right?" Giorgio says. "Well, I'm always the Good Cop. I don't work with a Bad Cop, either. Don't need it. You want to know the truth? The truth is—and this is important—everybody down deep wants to tell his or her story. It's true. No matter how damaging it is to them, no matter how important it is for them to keep quiet, they want to tell their story. If they feel guilty, they want to get it off their chest. If they feel justified in what they did, they want to explain themselves. I tell them, 'Hey, I know what you did and I can prove it. Now what are you going to do about it? If you show remorse, if you help me out, I'll go to bat for you.' I tell them that. And if you give them half a reason to do it, they'll tell you everything."

The most important thing is to get them talking. The toughest suspects are those who clam up and demand a lawyer right at the start. Giorgio believes that once he gets a suspect talking, the stream of words will eventually flow right to the truth...

Here is how Giorgio summarizes what turned out to be a very long and fruitful conversation:

"I was at home last night," Martinez said. "She did call me."

"Really, why?"

"She wanted me to pick her up. I told her, 'I'm watching the Mets game; I can't pick you up.'"

That was it. Giorgio acted very pleased with this statement, thanked Martinez, wrote it up, and asked the young man to sign it. Martinez did.

Then Giorgio stared at the statement and gave Martinez a quizzical look.

"You know, Carlos, something about this statement doesn't look right to me. You two had been going out for, what? Seven years? She calls you and asks you to pick her up at night where she's just gotten off work. It's not a safe neighborhood, and you tell her no? You mean a ball game on TV was more important to you?"

The question was cunning. The detective knew that Martinez was trying to make a good impression; he definitely didn't want to leave Giorgio with any unresolved issues to play in his mind. So it concerned him that his first statement didn't sound right.

Giorgio's question also touched Martinez's sense of chivalry, an important quality for many Hispanic men. It wouldn't do to be seen as ungentlemanly. Here was a young woman who had just been brutally killed. How would it look to her family and friends if he admitted that she had called and asked him for a ride and he had left her to her fate—for a ball game on TV?

The question also subtly suggested an out: The neighborhood wasn't safe. People got hurt or killed in that neighborhood all the time. Maybe Martinez could admit that he had seen Cheryl on the night of the murder without directly implicating himself. No one ever accused the former footballer of being especially bright. He rose to Giorgio's bait immediately.

He said, "Jerry, let me tell you what really happened." ("Note," Giorgio says proudly, "already I'm Jerry!") Martinez now said that he had left his place to pick Wright up after work, but they had gotten into an argument. "She got mad at me and told me she didn't need a ride, so I waited until she got on the bus, and then I left." ("Look, now he's the picture of chivalry!" Giorgio says happily.)

"Let me take that down," Giorgio said, again acting pleased with the statement. He wrote it out neatly and asked Martinez to look it over and sign it. Martinez did.

Again Giorgio squinted at the paper. "You know, Carlos, something is still not right here. Cheryl was a strikingly beautiful girl. People who saw her remembered her. She's taken that bus home from work many nights, and people on that bus know who she is. And you know what? Nobody who rode that bus saw her on it last night."

(This was, in Giorgio's words, "pure bullshit." He hadn't talked to anybody who rode that bus. "Sometimes you have to just take a chance," he says.)

Again Martinez looked troubled. He had not allayed the detective's suspicions. So he tried again. "Okay, okay," he said. "This is really it. Let me tell you what really happened. Cheryl called, and I left to pick her up, but I ran into a friend of mine—I can't tell you his name—and we picked her up together. Then Cheryl and I got in this argument, a big fight. My friend got fed up. So we drove away, up Broadway to 181st Street, and stopped at the McDonald's there. He pulled out a gun, my friend, and he told me to get out of the car. 'Wait here,' he told me. 'I'm going to get rid of your problem.' Then he left. I waited. Then he came back. He said he had gotten rid of my problem."

Giorgio nodded happily and started to write up statement No. 3. He acted troubled over the fact that Martinez refused to name the friend, and the young man quickly coughed up a name. Giorgio's lieutenant, who had been watching the session through a one-way mirror, immediately got to work tracking down Martinez's friend. By the time the third statement had been written up, signed, and nestled neatly on top of the other two, Giorgio had a new problem to pose to Martinez: it seemed that his friend was in South Carolina, and had been for some time.
"We never did get to finish the fourth statement," Giorgio says. "Martinez's family had hired a lawyer, and he called the station forbidding us to further question his client." It was, of course, too late.

Captain Crunch Versus The Tree Huggers

..."It isn't about getting mad, or payback," says Bill Cowan, the Vietnam interrogator. "It's strictly business. Torturing people doesn't fit my moral compass at all. But I don't think there's much of a gray area. Either the guy has information you need or not. Either it's vital or it's not. You know which guys you need to twist."

The official statements by President Bush and William Haynes reaffirming the U.S. government's opposition to torture have been applauded by human-rights groups—but again, the language in them is carefully chosen.

What does the Bush Administration mean by "torture"? Does it really share the activists' all-inclusive definition of the word?

In his letter to the director of Human Rights Watch, Haynes used the term "enemy combatants" to describe those in custody. Calling detainees "prisoners of war" would entitle them to the protections of the Geneva Convention, which prohibits the "physical or mental torture" of POWs, and "any other form of coercion," even to the extent of "unpleasant or disadvantageous treatment of any kind."

(In the contemptuous words of one military man, they "prohibit everything except three square meals, a warm bed, and access to a Harvard education.")

Detainees who are American citizens have the advantage of constitutional protections against being held without charges, and have the right to legal counsel. They would also be protected from the worst abuses by the Eighth Amendment, which prohibits "cruel and unusual punishment."

The one detainee at Guantánamo who was discovered to have been born in the United States has been transferred to a different facility, and legal battles rage over his status. But if the rest of the thousands of detainees are neither POWs (even though the bulk of them were captured during the fighting in Afghanistan) nor American citizens, they are fair game. They are protected only by this country's international promises—which are, in effect, unenforceable.

What are those promises?

The most venerable are those in the Geneva Convention, but the United States has sidestepped this agreement in the case of those captured in the war on terror.

The next most important would be those in the Universal Declaration of Human Rights, which asserts, in Article 5, "No one shall be subjected to torture or to cruel, inhuman or degrading treatment or punishment."

There is also the Convention Against Torture, the agreement cited by Bush in June, which would seem to rule out any of the more aggressive methods of interrogation. It states, in Article I, "For the purposes of this Convention, torture means any act by which severe pain or suffering, whether physical or mental, is intentionally inflicted on a person." Again, note the word "severe." The United States is avoiding the brand "torturer" only by sleight of word.

The history of interrogation by U.S. armed forces and spy agencies is one of giving lip service to international agreements while vigorously using coercion whenever circumstances seem to warrant it.

However, both the Army and the CIA have been frank in their publications about the use of coercive methods.

The Kubark Manual offers only a few nods in its 128 pages to qualms over what are referred to, in a rare euphemism, as "external techniques": "Moral considerations aside, the imposition of external techniques of manipulating people carries with it the grave risk of later lawsuits, adverse publicity, or other attempts to strike back."

The use of the term "strike back" here is significant; it implies that criticism of such unseemly methods, whether legal, moral, or journalistic, would have no inherent validity but would be viewed as an enemy counterattack.

Bill Wagner, the former CIA agent, remembers going to the Agency's three-week interrogation course at "The Farm," in Williamsburg, Virginia, in 1970. Until it was shut down, a few years later, it was considered the Agency's "premier course," Wagner says, and only the best recruits were invited to take it. "To say you had been through it was a real feather in your cap."

Volunteers played the role of captives in return for guaranteed space in a future session of the coveted course. They were deprived of sleep, kept doused with water in cold rooms, forced to sit or stand in uncomfortable positions for long periods, isolated from sunlight and social contacts, given food deliberately made unappetizing (oversalted, for instance, or tainted with a green dye), and subjected to mock executions.

At least 10 percent of the volunteers dropped out, even though they knew it was just a training exercise. Wagner says that many of those who had served as victims later refused to take the course and victimize others. "They lost their stomach for it," he says.

Several years after Wagner took the course, he says, the Agency dropped it entirely. The scandals of the Nixon years put the CIA under unprecedented scrutiny. Over the next three decades spying schools and most human-intelligence networks were gradually dismantled. The United States itself was losing its stomach for hands-on intelligence gathering—and with it, interrogation.

Nobody experienced the effects of this shift more dramatically than Keith Hall, who earned the nickname Captain Crunch before he lost his job as a CIA agent. Now he describes himself as "a poster child for political correctness." He is a pugnacious brick of a man, who at age fifty-two is just a thicker (especially in the middle) version of the young man who joined the Marines thirty years ago. After his discharge he earned a master's degree in history and international relations; he took a job as a police officer, because he craved a more physical brand of excitement than academia had to offer. His nickname comes from this craving.

The CIA hired Hall immediately after he applied, in 1979, because of his relatively rare combination of academic and real-world credentials. He was routed into the Investigation and Analysis Directorate, where he became one of the Agency's covert operators, a relatively small group ("about forty-eight guys, total," Hall says) known as the "knuckle-draggers." Most CIA agents, especially by the 1980s, were just deskmen.

Hall preferred traveling, training, and blowing things up, even though he felt that the rest of the Agency looked down its patrician nose at guys like him. When the U.S. Embassy in Beirut was bombed, on April 18, 1983, eight of the seventeen Americans killed were CIA employees. There were going to be plenty of official investigations, but the Agency wanted one of its own. Hall was selected to carry it out.

"They flew me to Langley on one of their private planes, and delivered me to the seventh floor," he says. "They told me, 'We want you to go to Beirut and find out who blew up the embassy and how they did it. The President himself is going to be reading your cables. There is going to be some retribution here.'"

Hall was honored, and excited. This was a mission of singular purpose, of the highest priority, and he knew he was expected to get results. Having been a police officer and a Marine, he knew that the official investigations had to build a case that might someday stand up in court. His goal was not to build a case but just to find out who did it.

He slept on rooftops in Beirut, changing the location every two nights. It was a dangerous time to be an American—especially a CIA officer—there, and Hall kept moving. He worked with the Lebanese Special Security Force, and set up a computer in the police building.

Hall says he took part without hesitation in brutal questioning by the Lebanese, during which suspects were beaten with clubs and rubber hoses or wired up to electrical generators and doused with water. Such methods eventually led him to the suspected "paymaster" of the embassy bombing, a man named Elias Nimr. "He was our biggest catch," Hall says—a man with powerful connections. "When I told the Lebanese Minister of Defense, I watched the blood drain out of his face."

Nimr was a fat, pampered-looking twenty-eight-year-old, used to living the good life, a young man of wealth, leisure, and power. He came to the police building wearing slacks, a shiny sport shirt, and Gucci shoes. He had a small, well-trimmed moustache at the center of his soft, round face, and wore gold on his neck, wrists, and fingers. When he was marched into the building, Hall says, some of the officers "tried to melt into the shadows" for fear of eventual retribution. Nimr was nonchalant and smirking in his initial interview, convinced that when word got back to his family and connections, he would promptly be released.

When Hall got a chance to talk to him, he set out to disabuse Nimr. "I'm an American intelligence officer," he said. "You really didn't think that you were going to blow up our embassy and we wouldn't do anything about it, did you? You really should be looking inside yourself and telling yourself that it's a good idea to talk to me. The best way to go is to be civilized ... I know you think you are going to walk right out of here in a few minutes. That's not going to happen. You're mine. I'm the one who will make the decisions about what happens to you. The only thing that will save your ass is to cooperate." Nimr smiled at him dismissively.

The next time they met, Nimr wasn't in such good shape. In this case his connections were failing him. No one had roughed him up, but he had been kept standing for two days. Hall placed him in a straight-backed metal chair, with hot floodlights in his face. The agent sat behind the light, so that Nimr couldn't see him. Nimr wasn't as cocky, but he was still silent.

At the third interrogation session, Hall says, he kicked Nimr out of his chair. It was the first time anyone had physically abused him, and he seemed stunned. He just stared at Hall. He hadn't eaten since his arrest, four days earlier. But he still had nothing to say.

"I sent him back to his cell, had water poured over him again and again while he sat under a big fan, kept him freezing for about twenty-four hours. He comes back after this, and you can see his mood is changing. He hasn't walked out of jail, and it's beginning to dawn on him that no one is going to spring him."

Over the next ten days Hall kept up the pressure. During the questioning sessions he again kicked Nimr out of his chair, and both he and the Lebanese captain involved cracked him occasionally across the shins with a wooden bat. Finally Nimr broke. According to Hall, he explained his role in the bombing, and in the assassination of Lebanon's President. He explained that Syrian intelligence agents had been behind the plan. (Not everyone in the CIA agrees with Hall's interpretation.)

Soon afterwards Nimr died in his cell. Hall was back in Washington when he heard the news. He assumed that Nimr had been killed to prevent him from testifying and naming others involved in the plot. Armed with tapes of Nimr's confession, Hall felt he had accomplished his mission; but several months after finishing his report he was fired. As he understood it, word had leaked out about torture sessions conducted by a CIA agent, and the U.S. government was embarrassed.

...The perfect model of an interrogation center would be a place where prisoners lived in fear and uncertainty, a place where they could be isolated or allowed to mingle freely, as the jailer wished, and where conversations anywhere could be overheard.

Interrogators would be able to control the experience of their subjects completely, shutting down access to other people, or even to normal sensation and experience, or opening that access up. Subjects' lives could be made a misery of discomfort and confusion, or restored to an almost normal level of comfort and social interaction within the limitations of confinement. Hope could be dangled or removed. Cooperation would be rewarded, stubbornness punished.

Interrogators would have ever growing files on their subjects, with each new fact or revelation yielding new leads and more information—drawn from field investigations (agents in the real world verifying and exploring facts gathered on the inside), the testimony of other subjects, collaborators spying inside the prison, and surreptitious recordings. The interrogators in this center would have the experience and the intuition of a Jerry Giorgio or a Michael Koubi.

Serious interrogation is clearly being reserved for only the most dangerous men, like Sheikh Mohammed. So why not lift the fig leaf covering the use of coercion? Why not eschew hypocrisy, clearly define what is meant by the word "severe," and amend bans on torture to allow interrogators to coerce information from would-be terrorists?

This is the crux of the problem. It may be clear that coercion is sometimes the right choice, but how does one allow it yet still control it?

In considering a change in national policy, one is obliged to anticipate the practical consequences. So if we formally lift the ban on torture, even if only partially and in rare, specific cases (the attorney and author Alan Dershowitz has proposed issuing "torture warrants"), the question will be, How can we ensure that the practice does not become commonplace—not just a tool for extracting vital, life-saving information in rare cases but a routine tool of oppression?

As it happens, a pertinent case study exists. Israel has been a target of terror attacks for many years, and has wrestled openly with the dilemmas they pose for a democracy.

In 1987 a commission led by the retired Israeli Supreme Court justice Moshe Landau wrote a series of recommendations for Michael Koubi and his agents, allowing them to use "moderate physical pressure" and "nonviolent psychological pressure" in interrogating prisoners who had information that could prevent impending terror attacks. The commission sought to allow such coercion only in "ticking-bomb scenarios"—that is, in cases like the kidnapping of Jakob von Metzler, when the information withheld by the suspect could save lives.

Twelve years later the Israeli Supreme Court effectively revoked this permission, banning the use of any and all forms of torture. In the years following the Landau Commission recommendations, the use of coercive methods had become widespread in the Occupied Territories.

It was estimated that more than two thirds of the Palestinians taken into custody were subjected to them. Koubi says that only in rare instances, and with court permission, did he slap, pinch, or shake a prisoner—but he happens to be an especially gifted interrogator.

What about the hundreds of men who worked for him? Koubi could not be present for all those interrogations.

Every effort to regulate coercion failed. In the abstract it was easy to imagine a ticking-bomb situation, and a suspect who clearly warranted rough treatment. But in real life where was the line to be drawn?

Should coercive methods be applied only to someone who knows of an immediately pending attack? What about one who might know of attacks planned for months or years in the future?

"Assuming you get useful information from torture, then why not always use torture?" asks Jessica Montell, the executive director of B'Tselem, a human-rights advocacy group in Jerusalem.

"Why stop at the bomb that's already been planted and at people who know where the explosives are? Why not people who are building the explosives, or people who are donating money, or transferring the funds for the explosives? Why stop at the victim himself? Why not torture the victims' families, their relatives, their neighbors? If the end justifies the means, then where would you draw the line?"

And how does one define "coercion," as opposed to "torture"?

If making a man sit in a tiny chair that forces him to hang painfully by his bound hands when he slides forward is okay, then what about applying a little pressure to the base of his neck to aggravate that pain? When does shaking or pushing a prisoner, which can become violent enough to kill or seriously injure a man, cross the line from coercion to torture?

...She knows that the use of coercion in interrogation did not end completely when the Israeli Supreme Court banned it in 1999. The difference is that when interrogators use "aggressive methods" now, they know they are breaking the law and could potentially be held responsible for doing so. This acts as a deterrent, and tends to limit the use of coercion to only the most defensible situations.

"If I as an interrogator feel that the person in front of me has information that can prevent a catastrophe from happening," she says, "I imagine that I would do what I would have to do in order to prevent that catastrophe from happening. The state's obligation is then to put me on trial, for breaking the law. Then I come and say these are the facts that I had at my disposal. This is what I believed at the time. This is what I thought necessary to do. I can evoke the defense of necessity, and then the court decides whether or not it's reasonable that I broke the law in order to avert this catastrophe. But it has to be that I broke the law. It can't be that there's some prior license for me to abuse people."

In other words, when the ban is lifted, there is no restraining lazy, incompetent, or sadistic interrogators. As long as it remains illegal to torture, the interrogator who employs coercion must accept the risk. He must be prepared to stand up in court, if necessary, and defend his actions. Interrogators will still use coercion because in some cases they will deem it worth the consequences. This does not mean they will necessarily be punished...

The Bush Administration has adopted exactly the right posture on the matter. Candor and consistency are not always public virtues.

Torture is a crime against humanity, but coercion is an issue that is rightly handled with a wink, or even a touch of hypocrisy; it should be banned but also quietly practiced. Those who protest coercive methods will exaggerate their horrors, which is good: it generates a useful climate of fear. It is wise of the President to reiterate U.S. support for international agreements banning torture, and it is wise for American interrogators to employ whatever coercive methods work. It is also smart not to discuss the matter with anyone.