Darius Rejali

The real significance of the Battle of Algiers, however, is the startling justification of torture by a democratic state. Algerian archives are now open, and many French torturers wrote their autobiographies in the 1990s. The story they tell will not comfort generals who tell self-serving stories of torture’s success. In fact, the battle shows the devastating consequences of torture for any democracy foolish enough to institutionalize it.

Torture by the French failed miserably in Vietnam, and the French could never entirely secure the Algerian countryside, so either torture really did not work or there was some additional factor that made the difference in Algiers.

Among many torture apologists, only Gen. Massu, with characteristic frankness, identified the additional factor. In Vietnam, Massu said, the French posts were riddled with informants. Whatever the French found by torture, the Vietnamese opposition knew immediately. And long distances separated the posts. In Algiers, the casbah was a small space that could be cordoned off, and a determined settler population backed the army. The army was not riddled with informants, and the FLN never knew what the army was doing.

And the French had an awesomely efficient informant system of their own. Massu took a census in the casbah and issued identity cards for the entire population. He ordered soldiers to paint numbers on each block of the casbah, and each block had a warden — usually a trustworthy Algerian — who reported all suspicious activities. Every morning, hooded informants controlled the exits to identify any suspects as they tried to leave. The FLN helped the French by calling a general strike, which revealed all its sympathizers. What made the difference for the French in Algiers was not torture, but the accurate intelligence obtained through public cooperation and informants.

In fact, no rank-and-file soldier has related a tale of how he personally, through timely interrogation, produced decisive information that stopped a ticking bomb. “As the pain of interrogation began,” observed torturer Jean-Pierre Vittori, “they talked abundantly, citing the names of the dead or militants on the run, indicating locations of old hiding places in which we didn’t find anything but some documents without interest.” Detainees also provided names of their enemies — true information, but without utility to the French.

The FLN military men had also been told, when forced to talk, to give up the names of their counterparts in the rival organization, the more accommodationist MNA (National Algerian Movement). Not very knowledgeable in the subtleties of Algerian nationalism, the French soldiers helped the FLN liquidate the infrastructure of the more cooperative organization and tortured MNA members, driving them into extreme opposition.

Unlike in the famous movie, which portrays the Algerian population as united behind the FLN and assumes that torture is why the French won the battle, the real Battle of Algiers was a story of collaboration and betrayal by the local population. It was, as Alistair Horne describes in “A Savage War of Peace: Algeria 1954-1962,” a population that was cowed beyond belief and blamed the FLN leadership for having brought them to this pass.

Gen. Massu’s strategy was not to go after the FLN bombers but to identify and disable anyone who was even remotely associated with the FLN. It was not a selective sweep. The smallest interrogation unit in Algiers possessed 100,000 files. Out of the casbah’s total population of 80,000 citizens, Massu arrested 30 to 40 percent of all males.

Torture forced “loyal” Algerians to cooperate, but after the battle, they either ended their loyalty to France or were assassinated. Torture forced a politics of extremes, destroying the middle that had cooperated with the French. In the end, there was no alternative to the FLN. As Paul Teitgin, the police prefect of Algiers, remarked, “Massu won the Battle of Algiers, but that meant losing the war.”

The judicial system also collapsed under the weight of torture. Judges and prefects found themselves unable to deny warrants to armed men who tortured and killed for a living. Police records show that Teitgin issued 800 detention orders (arrêtes d’assignation) for eight months before the battle, 700 for the first three months of the battle and then 4,000 a month for the remaining months. By the end of the battle, he had issued orders to detain 24,000, most of whom (80 percent of the men and 66 percent of the women) were routinely tortured.

And “what to do with these poor devils after their ‘use’?” asked a French soldier. Many torturers preferred to kill them, though, one soldier conceded, genocide was difficult. “There isn’t enough place in the prisons and one can’t kill everyone …, so one releases them and they’re going to tell others, and from mouth to mouth, the whole world knows.” Then, he observed, their relatives and friends “join the resistance.” By the end of the battle, about 13,000 Algerians (and some Frenchmen) were in detention camps and 3,000 “disappeared.”

Doctors, whose task it was to monitor torture, were themselves corrupted by the torture. “Our problem is,” remarked a doctor attached to a French torture unit, “should we heal this man who will again be tortured or let him die?” As oversight failed, the French military government arrested more people for flimsier reasons.

Use of torture also compromised the military. Lt. Col. Roger Trinquier, the famous French counterinsurgency expert, believed that torturers could act according to professional norms — applying only the pain necessary to get information and then stopping. But the stories of rank-and-file torturers confirm previous studies of the dynamics of torture. “I realized,” remarked a French soldier, “that torture could become a drug. I understood then that it was useless to claim to establish limits and forbidden practices, i.e. yes to the electrotorture but without abusing it, any further no. In this domain also, it was all or nothing.”

Torture drifted headlong into sadism, continuing long after valuable information could be retrieved. For example, soldiers arrested a locksmith and tortured him for three days. In his pocket, the locksmith had bomb blueprints with the address of an FLN bomb factory in Algiers. The locksmith bought time, the bombers relocated and the raid by the French three days later fell on open air. Had the soldiers been able to read Arabic, they would have found the bomb factory days earlier. But they were too busy torturing. As one would predict, engaging in torture prevented the use of ordinary — and more effective — policing skills. (Incidentally, the French could not believe that the most wanted man in the casbah had spent months only 200 yards from the headquarters of the army commandant.)

The French military also fragmented under the competition associated with torture. Parallel systems of administration emerged, and infighting occurred between the various intelligence agencies. Officers lost control of their charges, or the charges refused to follow higher command. And in the end, the soldiers blamed the generals for exposing them to torture, noting its pernicious effects on their lives, their families and their friends — a sense of betrayal that has not diminished with the years.

Yves Godard, Massu’s chief lieutenant, had insisted there was no need to torture. He suggested having the informant network identify operatives and then subject them to a simple draconian choice: Talk or die. This would have produced the same result as torture without damage to the army.

The British successfully used precisely this strategy with German spies during World War II. British counterespionage managed to identify almost every German spy without using torture — not just the 100 who hid among the 7,000 to 9,000 refugees coming to England to join their armies in exile each year, not just the 120 who arrived in similar fashion from friendly countries, but also the 70 sleeper cells that were in place before 1940. Only three agents eluded detection; five others refused to confess. Many Germans chose to become double agents rather than be tried and shot. They radioed incorrect coordinates for German V missiles, which landed harmlessly in farmers’ fields. But for this misdirection, British historian Keegan concludes, in October 1944 alone close to 1,300 people would have died, with 10,000 more injured and 23,000 houses destroyed.

The U.S. Army’s field manual for intelligence (FM34-52) notes that simple direct questioning of prisoners was 85 percent to 95 percent effective in World War II and 90 to 95 percent effective in the Vietnam War. What about those 5 percent at the margin? Couldn’t savage, unprofessional, hit-or-miss torture yield some valuable information from them? Actually, there was one case in the Battle of Algiers in which torture did reveal important information.

In September 1957, in the last days of the battle, French soldiers detained a messenger known as “Djamal.” Under torture, Djamal revealed where the last FLN leader in Algiers lay hidden. But that wasn’t so important; informants had identified this location months before. The important information Djamal revealed was that the French government had misled the military and was quietly negotiating a peace settlement with the FLN. This was shocking news. It deeply poisoned the military’s relationship with the civilian government, a legacy that played no small part in the collapse of the Fourth Republic in May 1958 and in the attempted coup by some French military officers against President De Gaulle in April 1961.

The French won the Battle of Algiers primarily through force, not by superior intelligence gathered through torture. Whoever authorized torture in Iraq undermined the prospect of good human intelligence. Even if the torture at Abu Ghraib served to produce more names (“actionable intelligence”) and recruit informants, torture in the end polarized the population, eliminating the middle that might cooperate. Dividing the world into “friends” and “enemies,” those who are with us or against us, meant that we lost the cooperation of those who wished to be neither or who were enemies of our enemies.

Whoever authorized torture in Afghanistan and Iraq also destroyed the soldiers who were ordered to perform it. Studies of torturers show that they would rather work as killers on death squads, where the work is easier. Torture is hard, stressful work. Many torturers develop emotional problems, become alcoholics, beat their families and harbor a deep sense of betrayal toward the military brass that hangs them out to twist in the wind. The soldiers at Abu Ghraib had dreams, dreams that democracy promised to fulfill, dreams that now may never be fulfilled thanks to the arrogance of their superiors.

Those who authorize torture need to remember that it isn’t something that simply happens in some other country. Soldiers trained in stealthy techniques of torture take these techniques back into civilian life as policemen and private security guards. It takes years to discover the effects of having tortured. Americans’ use of electric torture in Vietnam appeared in Arkansas prisons in the 1960s and in Chicago squad rooms in the 1970s and 1980s.

Likewise, the excruciating water tortures U.S. soldiers used in the Spanish-American War appeared in American policing in the next two decades. For those who had been tortured, it was small comfort when, on Memorial Day 1902, President Roosevelt regretted the “few acts of cruelty” American troops had performed.

Some believe that judges can issue selective torture warrants to security officers in important cases. But the rapid increase in the number of torture warrants issued during the Battle of Algiers is evidence enough that civil servants can exercise little selective control once they have licensed unlimited power.

Others believe that torture occasionally is necessary and that when it is, one should have to answer for one’s actions before the law. But “morally justified” torture does not resemble morally justified civil disobedience. Civil rights protesters break the law in public and then submit their behavior to juries and courts. But I know of no modern torturer who voluntarily submitted to public scrutiny and took the heat. Like boasts of bravery, this opinion is too easy to hold when there is no danger of it being tested. Modern torturers operate in secrecy and specialize in techniques that leave no marks. What would we really know of Abu Ghraib in the absence of the photographs?

And once soldiers get away with torture, they repeat it. Few things predict future torture as much as past impunity.

It is easy to criticize the leaders and torture apologists who misled us and continue to do so. What is harder is to determine how to repair the damage. One crazy man can block the well, but it takes the whole village to remove the stone, an Iranian proverb says.

We can learn from the mistakes of other democracies that have tortured. These democracies lost their wars because the brutality they licensed reduced their intelligence, compromised their allies and corrupted their military and government, and they could not come to terms with that.

When the politicians first heard of the torture, they denied it happened, minimized the violence and called it ill treatment. When the evidence mounted, they tried a few bad apples, disparaged the prisoners and observed that terrorists had done worse things. They justified the torture as effective and necessary for the extreme circumstances and countercharged that critics were aiding the enemy. As time passed, they offered apologies but accepted no consequences and argued that there was no point in dwelling on past events.

The torture continued because these democrats could not institutionally recommit themselves to limited power at home or abroad. The torture interrogations yielded the predictable results, and the democracies remained mired in their wars despite overwhelming military superiority against a far smaller enemy. Soon the politicians had to choose between losing their democracy and losing their war. That is how democracies lose wars.

Choking with water: For centuries, interrogators used water to cause pain in two ways, “pumping” and “choking.” Pumping involves forcibly filling the stomach and intestines with water. A garden hose in the mouth is sufficient. This technique (the “Tormento de Toca” in Spanish) was one of the most fearful tortures of the Inquisition. Choking pushes pumping one step further by preventing breathing.

Americans were the first to use pumping for interrogation in the 20th century. In 1902, during the Spanish-American War, U.S. soldiers put funnels in the mouths of Filipinos to force water into their organs. In pumping, victims’ organs stretch and convulse, causing the most intense pain that visceral tissue can experience. William Howard Taft, governor of the Philippines, carelessly conceded to the Senate that pumping was the policy in some cases.

Because water torture left few marks, it was easy to dismiss. President Theodore Roosevelt privately called pumping a mild torture, observing that no one had been seriously damaged. Italian dictator Mussolini’s police adopted pumping in the 1920s, as did the Japanese military police (Kempeitai) in the 1930s. In 1939, British police used pumping in Mandatory Palestine.

In 1943, “Masuy,” a Belgian who ran Gestapo black market operations in France, made choking his signature technique. Masuy’s henchmen held the victim’s head under water while Masuy offered cognac and questioned the victim. Masuy maintained at his trial in 1947 that choking was “more humane” than plucking nails. The Gestapo, not known for its humanity, authorized this method for Norway and Czechoslovakia in 1945 as the resistance struggles there intensified.

After the war, French police interrogators used choking to produce false confessions in Madagascar (1947) and France (1947-1948). The technique also appeared in Vietnam, where it was known as the “submarine.” By the late 1950s, French military intelligence used both pumping and choking in interrogation. They also compressed bloated stomachs violently, forcing water back out through all the orifices, including the eyes.

By the 1960s, Francisco Franco’s police in Spain and many Latin American torturers were using choking, which has become increasingly common in the past 20 years as a supplement to electrotorture in countries that torture routinely. (In a technique reportedly used by the CIA after 9/11, the victim is strapped to a board and then forcibly pushed under water till he nearly chokes to death, in a method known as “waterboarding.”)

Forced standing: Stress positions include Japanese kneeling positions, Soviet “sitting” positions and the Israeli shabeh techniques. But the American form draws on the oldest technique of all, forced standing (called the planton in Latin America and the stoika in the Soviet Union). The hooded man in the famous photo from Abu Ghraib was kept standing on a box for a whole evening. Like water torture, forced standing leaves no marks.

Two experts commissioned by the CIA in 1956, Harold Wolff and Lawrence Hinkle, described the effects of forced standing: The ankles and feet swell to twice their size within 24 hours, and moving becomes agonizing. Large blisters develop. The heart rate increases, and some people faint. The kidneys eventually shut down.

In the early 20th century, forced standing was a prescribed field punishment for soldiers in West European armies. Soldiers in the British Army referred to it as the “crucifixion,” and French legionnaires called it the silo.

By the 1920s, forced standing was a routine police torture in America. In 1931, the National Commission on Lawless Enforcement of the Law found numerous American police departments using forced standing to coerce confessions. In the 1930s, Joseph Stalin’s NKVD used forced standing to coerce seemingly voluntary confessions for show trials.

The British used forced standing as early as 1937-1939 on Arab and Jewish prisoners in Mandatory Palestine. The Gestapo routinely used it as a punishment in concentration camps, sometimes creating tiny, narrow cells where prisoners had to stand all night.

In the mid-20th century, torturers learned how to use the swelling and blistering to cause more pain. The Chinese wrapped the feet in gauze, which squeezed painfully as the feet swelled. Brazilian police made prisoners stand on the edges of cans or bricks, causing excruciating pain to the feet. During apartheid in South Africa, forced standing was the third most common torture technique after beating and electricity.

South African and Brazilian police also commonly used hooding, a variation on forced standing. In the 1970s, Brazilians added electroshock — which they called the “Vietnam,” combining techniques used by the Communists and non-Communist forces during the Vietnam War. When hooded victims began to collapse in exhaustion from standing, jolts of electricity would make their feet stick to the cans and force them to stand up straight.

Sweating and sweatboxes: During the Civil War, the sweatbox was a cell near a very hot stove that produced intense heat. By the early 20th century, sweatboxes were dark, solitary and very cold, hot or wet. A related police practice in Europe and the United States in the 1920s was known as sweating, in which detectives took turns interrogating suspects for hours or days (in a stifling interrogation room, hence the term). They kept them awake by shining bright lights into their eyes, shaking them or making constant noise. Stalin’s police also used forms of the sweatbox and sweating, using them, like U.S. police, to generate false confessions for public trials.

Sleep deprivation reduces a body’s tolerance for physical pain, causing deep aches first in the lower part of the body, followed by similar pains in the upper body. Sleep-deprived people are also highly suggestible (a condition not unlike drunkenness or hypnosis), making sleep deprivation ideal for inducing false confessions.

The Gestapo was the first to use sleep deprivation to gather information. In 1942, Gestapo chief Heinrich Mueller authorized “sharpened interrogation” for terrorists, approving sleep deprivation, starvation, exhaustion exercises, regulated beating and confinement in dark cells — but only to gather intelligence on those who had “plans hostile to the state,” not to get confessions of guilt.

Gestapo torture actually went beyond these methods, but Mueller’s order was the only one the Allies found. When confronted with the order at Nuremberg, Werner Best, the Nazi governor of Denmark, objected to what he considered the hypocritical criticism of an American investigator. “Similar methods were used in other countries,” Best said.

Indeed, American police commonly used sweating on suspects in the early 20th century. In 1941 Tennessee police subjected one suspect, Ashcraft, to sleep deprivation and interrogation for 36 hours until he confessed he had killed his wife. In Ashcraft v. Tennessee (1944), the Supreme Court did not simply toss out the confession as unacceptable in any democratic society; it linked sweating directly to the practices of “certain foreign nations dedicated to an opposite policy,” namely, “physical or mental torture.”

Yet sweating remains a common technique in many countries. After 9/11 the U.S. military authorized sleep deprivation for up to 72 hours, far longer than what Ashcraft was subjected to. The U.S. interrogators and military police who applied these methods at Abu Ghraib may not have known it, but they were acting in a tradition that includes colonial imperialists, Stalin’s secret police and the Gestapo.

The abuse and torture at Abu Ghraib prison were originally blamed on a few American soldiers. Various investigations into the exact chain of command are underway, but they already point to policy decisions made at the highest levels of the U.S. government. Indeed, the recently revealed memos written by Justice Department lawyers in August 2002, at the request of the CIA and the White House, concerning treatment of al-Qaida suspects at Guantánamo Bay, Cuba, and by Pentagon lawyers in March 2003 (in which it was argued that the president and those he has empowered to conduct torture of foreign prisoners are immune from prosecution under international law) are evidence that the government was seeking ways to legally circumvent the Geneva Conventions. “The question put to lawyers was how the president and the others could commit war crimes and get away with it” is how Anne Applebaum put it in the Washington Post Wednesday. It turns out that many of the severest interrogation techniques used in Iraq were sanctioned by top military officials, including Secretary of Defense Donald Rumsfeld and the former commander in Iraq, Lt. Gen. Ricardo Sanchez. And this week, we learned that Rumsfeld had ordered a “high-value” detainee in Iraq held in secret, in part to keep him from being seen by the International Red Cross.

I learned how torture fit into modern life while growing up in Iran under Mohammad Reza Shah Pahlavi, whose government relied on Savak, a secret intelligence agency formed with the help of the CIA in 1957. Savak arrested and detained indefinitely people suspected of opposing the shah and tortured and executed thousands of political prisoners during his rule.

In the course of 20 years of research on modern torture and the bureaucracies that sponsored and practiced it in Germany, Japan, France and Britain, I have studied the “stealthy” methods, those that leave few visible marks (i.e., blood or scars) on the victim. I noticed that stealthy techniques appeared more often in the wars of democracies than in those of dictatorships. Democratic states that use torture to gain intelligence or as punishment obviously prefer methods — such as electroshock, torture by water and ice, tying victims in agonizing postures, sonic devices and drugs — that cause pain but do not result in lasting injury, so that the torture cannot be verified by journalists, human rights monitors or congressional committees. The advantage of stealth torture is that it reduces adverse publicity and finesses democratic oversight.

After 9/11, the warning signs of what the United States was up to were there, but then the events from Abu Ghraib brought all of it to the surface, revealing that the U.S. military was employing some of these stealthy techniques in Afghanistan and Iraq.

My research shows, however, that torture during interrogations rarely yields better information than traditional human intelligence, partly because no one has figured out a precise, reliable way to break human beings or any adequate method to evaluate whether what prisoners say when they do talk is true. Nor can torture be done in a professional way — anyone who tortures is necessarily corrupted by the experience and is often turned into a sadist. The psychic damage to the soldiers who conducted the torture at Abu Ghraib is likely to be permanent.

What’s more, a democracy that legalizes the use of torture in its desperation to gain information loses something more important — the trust of its people, the foundation of a democracy. In Iraq, the United States was desperate as it sought to find and stop those responsible for the insurgency. When “intelligence” was not forthcoming from prisoners, senior U.S. Army officials decided to turn over interrogation to military intelligence personnel, who were instructed to use aggressive, even brutal techniques. These methods were rationalized as necessary in the overall global war on terrorism, but as my research has shown, institutionalizing torture in such a manner only ends up destroying all the individuals involved — and the military and political goals of the government in whose name torture is carried out.

Aside from its devastating effects and the wasted time and resources, does torture actually work? Organizations can certainly use torture to intimidate prisoners and to produce confessions (many of which turn out to be false). But the real question is whether organizations can apply torture scientifically and professionally to produce true information. Does this method yield better results than others at an army’s disposal? The history of torture demonstrates that it does not — whether it is stealthy or not.

Advocates of torture believe that more physical pain stimulates more compliance, but this view is not based on science; it is medical nonsense. Pain, as noted clinical psychologist Ron Melzack has shown, is far more complex than that. Injury does not always produce pain. In one study, 37 percent of people who arrived at an emergency ward with injuries such as amputated fingers, major skin lacerations and fractured bones did not feel any pain until many minutes, even hours, after the injury. Similarly, soldiers with massive wounds sometimes do not feel their pain for a long time.

In addition, human beings differ widely in their ability to endure extreme pain. Clinical psychologists and some torturers in colonized nations agree that past experiences and cultural beliefs (for example, “suffering is divine”) enable some human beings to endure pain others could not. People also vary in their ability to use psychological states like distraction or anxiety to reduce pain.

Moreover, pain, unlike heat, is not a single sensation but, as Melzack observes, can variously feel like burning, throbbing or cutting. Victims can play these different sensations against each other, using one pain to distract themselves from another, much like a person might bite his hand as someone extracts a thorn.

Last, pain is not constant. As the body is damaged, its ability to sense pain declines. Torturers run out of places where they can apply pain effectively.

Unlike the physics of boiling water, in which one knows how much heat to apply, there is no way to calculate in advance how much torture is needed to obtain compliance from a prisoner. Science and technology can help with the conduct of torture: Modern instruments reduce the hard labor of torture, helping ensure that it is not lethal; and they guarantee that few marks will be left as evidence. But science and technology cannot predict the precise amount or kind of torture that will work with each human being. Opportunistic use of technology does not make torturers scientific any more than wearing a white lab coat makes torturers scientists.

If torture can’t be scientific, can it at least be done professionally? To think that professionalism is a guard against causing excessive pain is an illusion. Instead, torture induces a dynamic that breaks down professionalism. Yale psychologist Stanley Milgram has shown that professionalism can serve to excuse ever more violent behavior. The myth of the professional torturer is also shattered in “Violence Workers: Police Torturers and Murderers Reconstruct Brazilian Atrocities,” by Martha K. Huggins, Philip G. Zimbardo and Mika Haritos-Fatouros.

As a victim feels less pain, torturers have to push harder, using more severe methods to overtake the victim’s maximal pain threshold. And because victims experience different types of pain, torturers have to use a scattershot approach. No matter how professional torturers may think they are, they have no choice but behaving like sadists. Even though many of the interrogators at Abu Ghraib were using techniques approved by their superiors, it is no surprise that they went far beyond these techniques, trying anything that worked.

Competition among torturers also drives brutality. As one torturer put it, each interrogator “thinks he is going to get the information at any minute and takes good care not to let the bird go to the next chap after he’s softened him up nicely, when of course the other chap would get the honor and glory of it.” Torture, as New York University economist Leonard Wantchekon has said, is a zero-sum game.

What’s more, coercive interrogation undermines other professional policing skills: Why do fingerprinting when you’ve got a bat? Which means investigators rely on even more torture to get information, increasing the degree of brutality.

Finally, competition between intelligence agencies that conduct torture tears bureaucracies apart. Competition between intelligence agencies is a normal phenomenon, and usually a good one, producing multiple sources of information. But when agencies turn to torture, the competition to get first crack at the victim leads to unprofessional behavior and bureaucratic fragmentation. Brazilian police raided each other’s prisons. The Nazi intelligence machine fragmented under the intense competition among the Kripo, Sipo, and regional Gestapos.

Is this way of applying pain more effective than other investigative methods? Torture is definitely inferior. The interrogation manual of Japanese fascists put it this way: “Care must be exercised when making use of rebukes, invectives or torture as it will result in his telling falsehoods and making a fool of you.” Torture “is only to be used when everything else has failed as it is the most clumsy [method].”

Since the 1970s, a large body of research has shown that unless the public specifically identifies suspects to the police, the chances that a crime will be solved falls to about 10 percent. Only a small percentage of crimes are discovered or solved through surveillance, fingerprinting, DNA sampling and offender profiling.

Police in long-term dictatorships like China and the Soviet Union also know the importance of public cooperation for solving crimes. Where they can’t get public cooperation for certain crimes (such as against state property), they create an alternative human intelligence system — informants. Although such police states use torture for intimidation and false confessions, they also know that good intelligence requires humans willing to trust government enough to work with it.

Even guerrillas know this truth. An internal report from Iraq, quoted by Seymour Hersh in the May 24 New Yorker, states that the insurgents have depended mainly on “painstaking surveillance and reconnaissance” by the Iraqi police force, “which is rife with sympathy for the insurgents” and “pro-insurgent individuals working within the [Coalition Provisional Authority's] so-called Green Zone.” Not surprisingly, the insurgents’ “strategic and operational intelligence has proven to be quite good.”

Torture is a sign that a government either does not enjoy the trust of the people it governs or cannot recruit informers for a surveillance system. In both cases, torture to obtain information is a sign of institutional decay and desperation — as Saddam Hussein’s Iraq clearly demonstrates. And torture accelerates this process, destroying the bonds of loyalty, respect and trust that keep information flowing. As any remaining sources of intelligence dry up, governments have to torture even more.

But perhaps torture for something, anything, is better than sitting on one’s hands. Maybe, somehow, one can retrieve a nugget of true information.

The problem is that “anything” needs to be verified, and as the Vietnam-era CIA Kubark manual explains, “a time-consuming delay results.” In the meantime, the prisoner can think of new, more complex falsehoods. Intelligence gathering is especially vulnerable to this. In police work, the crime is already known; all one wants is the confession. But in intelligence, one must gather information about things that one does not know.

What’s more, even prisoners who tell the truth under torture normally provide less detailed information than that obtainable through noncoercive interrogation. Damaged, sleep-deprived bodies remember information inaccurately, unable to make fine distinctions. Consider the case of a prisoner who wanted to tell the Chilean police an address: “Although I knew the street name, I had no idea of the number. Still furious, they realized that in truth I could not tell them where to go and once more they untied me.”

Sometimes when prisoners provide true information, interrogators refuse to recognize it, since they assume most victims lie. So they continue torturing until they are satisfied. The notion that one will stop when one hears the right information presupposes that one has gathered circumstantial information that allows one to know the truth when one hears it. But that is precisely what does not happen with torture. Torturers spend so much time on torture that they have no time to gather supplementary evidence.

Finally, even when torturers think they know what they are looking for, they sometimes can’t believe true information. One prisoner in Chile broke down several days into torture and revealed the names of the nuns and priests who had sheltered her. But the conservative and devout interrogators could not believe they were involved and continued torturing her.

What if time is short, as with a “ticking bomb”? Does torture offer a shortcut? Real torture — not the stuff of television — takes days, if not weeks. Even torturers know this. There are three things that limit torture’s value in this context.

First, there is the medical limit. Physical methods, like psychological methods, take time. In the face of extreme pain, human beings faint and, as one French resistance fighter said, this “gives you a reprieve between blows” and delays interrogation. As the interrogation proceeds, victims become less sensitive to pain. After undergoing four torture sessions, a Norwegian resistance fighter concluded that “pain had reached its limit — when it could hurt no more, what did it matter how it was inflicted?” In addition, as torturers push harder, they sometimes cause inadvertent death. And dead men, like unconscious men, don’t talk.

Second, there is the resource limit. For decades, guerrilla organizations have had “torture contracts” with their members: If you get arrested, keep the interrogators busy for 24 hours and let us change the passwords and locations. Give them false information mixed with half-truths. Make them waste their time and resources, and then after a day say whatever you want, since it will be useless then. Remember that you will become unconscious when the pain is extreme, and consider feigning unconsciousness.

Last, there’s the psychological limit. The CIA Kubark manual notes that coercive investigation requires compiling a psychological profile, which can take days to write. Without a psychological profile, the manual says, torture is a “hit or miss” practice and “a waste of time and energy.” Shot-in-the-dark torturing brings to mind the torturer’s paradox. If he tortures first, he may be unable to get information by gentler means later. But if he tortures at the end, the prisoner may conclude that he is getting desperate and hold out longer.

Hardcore believers, including presumably the common terrorist, don’t break quickly. Torturing them just gives them an excuse to prove that they’re stronger than you. Even the Gestapo discovered that with members of the resistance in World War II: Few resistance fighters gave accurate information.

In fact, as George Browder explains in his powerful book “Hitler’s Enforcers,” “the Gestapo, like police anywhere, could not do its work without public support.” The Gestapo’s enormous success against the resistance, first in Germany and then elsewhere, depended heavily on bureaucratic files, police informants (G-men or V-men) and collaborators in foreign countries. “Increased reliance on interrogation through torture during the war years reflects the declining professionalism of an overextended staff much watered down with neophytes,” Browder writes.

The priority in America’s war on terror should be on developing human intelligence. Working one’s way into a terror cell is not unlike working one’s way into organized crime in the United States. One has to turn potential terrorists into double agents and to win the confidence and cooperation of the communities that shelter them. Technology is no substitute for this. Nor is torture.

Abu Ghraib should teach us what America’s founders would have told us: that we are our own worst enemy. Leaders of dictatorships sign on to the Geneva Conventions only out of prudential fear of what other states might do to their POWs. Leaders of democracies sign on to them because they understand the evil that lurks in the heart of all human beings. Those who choose to abide by the rules do so not simply to restrain others but to restrain themselves.

Unrestrained power leaves behind a legacy of destruction that takes generations to undo. Torture, like incest, is the gift that keeps on giving. Democratic societies that legalized torture or tried to constrain its use have come to two ends. Some, like the Greeks and Romans, created tiered societies where authorities could torture whole classes of people (slaves or lesser citizens) and those who were beyond torture. Others, like the Italian city-states, were unable to prevent the executive branch from torturing more and more citizens and in the end fell to its dictatorial power.

The first result is hardly a model for modern democracies, and the second serves as a warning. In modern times, France routinized torture in Algeria, producing a racist, tiered society and an aggressive military government that almost overthrew French democracy. Proponents of torture would argue that destroying democratic institutions — and the individuals involved — is worth it if torture, as for the French in Algeria, succeeds in defeating terrorism.

Read Part 2: The example of military victory against terrorism by a democracy that used torture