2
It is torture when an investigation is conducted with torment and force. … Torture is a weak and dangerous thing that may fail the truth.
—ULPIAN, Justinian’s Digest
Marina Gonzàlez stood before the Toledan inquisitors Fernando de Mazuecos and Ferndando Rodríguez del Barco on April 29, 1494, and rejected their accusations of heresy. With this refusal, she was taken immediately to the torture chamber, stripped of her skirts, and bound tightly to the rack with cords, including her head. “They put a hood in front of her face, and with a jar that held three pints, more or less, they started to pour water down her nose and throat.” She refused to confess after the first pint so they poured the remaining two pints before removing the hood and cord from her head. She again refused to confess, was tied down again, and endured another three pints from the jar before the inquisitors ordered her brought down from the rack (Homza 2006, p. 45).
Juan de Salas stood before the inquisitor Moriz in Vallalodid on June 21, 1527, and once again denied he had blasphemed the evangelists. As a result, he was taken directly to the torture chamber, accompanied by the executioner Pedro Porras and the notary Henry Paz in addition to Moriz. Bound tightly to the escalera, a kind of grooved table that raises the legs above the head, a “fine wet cloth was put over his face, and about a pint of water was poured into his mouth and nostrils … nevertheless, Salas still persisted in denying the accusation” (emphasis in original). After several more rounds of questioning, denials, and water-pouring, Salas was taken down from the escalera (Llorente and Lovett 1967, pp. 121–122)).
John Clarke passed the “weeping and lamenting” Samuel Colson on his way into the examining room in the Dutch fort at Amboyna on February 16, 1622. Clarke was pulled two feet off the ground up onto a large door, his limbs outstretched and bound tightly with cords. A cloth was tied around his neck and folded up in a cone toward the top of his head. “That done, they poured water softly upon his head until the cloth was full, up to the mouth and nostrils, and somewhat higher; so that he could not draw breath, but he must withall suck-in the water: which being still continued to be poured in softly, forced all his inward parts, came out of his nose, ears, and eyes, and often as it were stifling and choking him, at length took away his breath and brought him to a swoon or fainting. Then they took him quickly down and made him vomit up the water.” When he had recovered, his Dutch captors repeated the procedure three more times. Clarke’s continued resistance, his refusal to confess to a plot to take over the castle for the English, led the examiners to believe that he was possessed by the devil for withstanding so much. After adding burning his skin to the water torture and “leading” him along with the necessary details, he confessed and he was carried out by four men to the castle dungeon (East India Company 1665, pp. 10–12)).
Jean Bourdil walked past a doctor and a priest and into the torture chamber of Toulouse on May 27, 1726. Surrounded by hooks, pulleys, ropes, weights, and other means of stretching human joints, he took an oath to tell the truth to his interrogators. A little while later, after those hooks, pulleys, ropes, and weights had failed to elicit his confession to the charge of killing two soldiers—just as they had two days prior—he was subjected to the question d’eau five times.
The question d’eau consisted of fastening the wrists of the accused to an iron ring bolted into the wall at waist height, and the feet, to another ring embedded in the floor, thus extending the prisoner’s body to its full length at a slant. Trestles of varying heights were then wedged under the prisoner’s back, forcing a further extension of the body. Finally, his face covered by a linen napkin, water was forced down his throat through a cow’s horn, as much as sixteen liters at a time (Silverman 2001, p. 46).
During each session, Bourdil was asked to tell the truth, to confess to his guilt in the crime, or his complicity in the crime, or his knowledge of those who were guilty. After the fifth session the linen rag was removed from his mouth, and he was unstrapped from the bench and likely set down by a fire and ministered to by the doctor before being returned to his cell.
In New Castle, Alabama, on January 6, 1881, F.H. Gafford, an employee of a coal company, gave testimony to the Joint Special Committee of the state legislature conducting an inquiry into the treatment of convicts in Alabama. Mr. Gafford’s company “rented” convicts from the state as part of its convict leasing program and was asked about the punishments employed for “wilful neglect of work.” In order to secure compliance, Mr. Gafford testified that the company generally employed the “strap” but that on three occasions in the three years of his employ, they had “punished with water.” “In water punishment the man is strapped to his back and the water is poured in his face on the upper lip, and effectually stops his breathing as long as there is a constant stream. It is a very dangerous punishment” (Alabama State Legislature 1881, p. 7).
Ramon Navarro, a Filipino lawyer, had failed to provide the information sought by his Japanese interrogator after a morning of interrogation.1 It was World War II, and the Japanese occupiers of the Philippines were seeking information on guerilla activities. At around 2:00 pm, the interrogator, Major Chinsaku Yuki, ordered Navarro to remove all of his clothes and lie down on his back on a bench. Navarro complied with both orders. Yuki then tied Navarro’s feet, hands, and neck to the bench. Once Navarro was secured to the bench, Yuki put a cloth on Navarro’s face and began pouring water from a nearby faucet on Navarro’s face. The stream of water continued until Navarro lost consciousness. The process was then repeated four or five times over the next two hours. At that point, Navarro told his interrogator a lie in an attempt to stop the “treatment.” When Yuki discovered the lie, he repeated the procedure on Navarro another two or three times.
Captain Chase Nielsen, a B-25 navigator, had provided nothing but his name, rank, and serial number to his Japanese interrogator on the afternoon of April 24, 1942. After failing to elicit more information on what would later be called the Doolittle Raid, the interrogator had four soldiers pin his arms and legs to the floor. Another soldier wrapped his face with a wet towel and poured water on the towel.
They poured water on this towel until I was almost unconscious from strangulation, then they would let up until I’d get my breath, then they’d start all over again. I felt more or less like I was drowning, just gasping between life and death … (Nelson 2003, p. 245).
Nielsen was rescued at the end of the war, but returned to Shanghai to testify against his captors at a war crimes trial.
Thomas D. Harrison, a U.S. Air Force pilot, bailed out of his disabled jet on May 21, 1951, near Sinuiju, North Korea, and was captured. He refused to provide his captors any information, even after being starved for nine days. One day in November he was beaten with clubs and sticks before being subjected to the “water treatment” (Hansen 1953, p. 17):
They would bend my head back, put a towel over my face and pour water over the towel. I could not breathe. This went on hour after hour, day after day. It was freezing cold. When I would pass out, they would shake me and begin again. They would leave me tied to the chair with the water freezing on and around me (Hansen 1953, p. 17).
Henri Alleg refused to divulge to his French paratrooper interrogators where he had been hiding, despite a beating and two sessions of electrical torture shortly following his arrest on Wednesday, June 12, 1957, in Algiers. Still strapped by his wrists and ankles to a black plank, one of his interrogators asked him if he “knew how to swim” and the paratroopers picked up the plank and carried Alleg into the kitchen of the apartment in the unfinished building that served as the interrogation center the Algiers suburb of El-Biar (Alleg Calder and Sartre 2006, p. 48). They rested the plank on the edge of the sink and two paratroopers held up the other end. One of the troopers near the sink attached a rubber tube to the tap and then wrapped a rag around Alleg’s head. His mouth wedged open with a small piece of wood, they told him to move his fingers when he wanted to talk. Then they
turned on the tap. The rag was soaked rapidly. Water flowed everywhere: in my mouth, in my nose, all over my face. But for a while I could still breathe in some small gulps of air. I tried, by contracting my throat, to take in as little water as possible and to resist suffocation by keeping air in my lungs for as long as I could. But I couldn’t hold on for more than a few moments. I had the impression of drowning; and a terrible agony, that of death itself, took possession of me. In spite of myself, all the muscles of my body struggled uselessly to save myself from suffocation. In spite of myself, the fingers of my two hands shook uncontrollably. “That’s it! He’s going to talk,” said a voice (Alleg Calder and Sartre 2006, p. 49).
When Alleg again fell silent after he had caught his breath, the torturers yelled, “He’s playing games with us! Put his head under again!”
This time I clenched my fists, forcing the nails into my palm. I had decided I was not going to move my fingers again. It was better to die of asphyxiation right away. I feared to undergo again that terrible moment where I felt myself losing consciousness, while at the same time fighting with all my power not to die. I did not move my hands, but three times I knew again this insupportable agony. In extremis, they let me get my breath back while I threw up the water. The last time I lost consciousness (Alleg Calder and Sartre 2006, pp. 49–50)).
Nguyen Cong arrived at LZ English on the southeastern coast of Vietnam in the fall of 1968. Suspected of belonging to the Viet Cong, he was interrogated on the morning of August 21 by Staff Sgt. David Carmon, of the 172nd Military Intelligence detachment. Carmon later admitted to using the “water rag” in his interrogations: “I held the suspect down, placed a cloth over his face and then poured water over the cloth, thus forcing water into his mouth. The suspect, after becoming choked on the water, confessed that he was a VC and stated he was a propaganda man” (Nelson and Turse 2006). In the Cong case, Carmon “poured water on his face from a five gallon can” (Criminal Investigation Division 1971, August 23, p. 10). Carmon and other interrogators also slapped, kicked, and beat Cong (Nelson 2008, pp. 62–63)). Carmon later said that “our intentions were never to hurt anyone, we simply wanted the information …. This is the reason that we primarily used water. Water poured over a cloth gave a sensation of drowning that generally scared the PW into talking” (Nelson 2008, p. 65). A little before 11:00 am and still under interrogation, Cong went into convulsions and died from a ruptured spleen, as a result of either external trauma or malaria.
Kevin Coffman was arrested in Coldspring, Texas, in the fall of 1979 for drug possession and put in the old San Jacinto County jail.2 Having been sentenced to six months in the jail for a misdemeanor offense, he was made a “trusty” after a couple of weeks.3 One morning after about two months in jail, he was taken along with another inmate, Craig Punch, to the new county jail still under construction. Each prisoner was placed in his own cell; and Deputy Sheriff Floyd Baker handcuffed Coffman to a brace in the wall, securing him to the chair he was sitting in. Coffman was left alone for a time before deputies Baker, Carl Lee, and John Glover entered his cell.
Deputy Baker folded a towel along its longer axis twice, wrapped it tightly around Coffman’s face, and pulled his head back. Deputy Glover “took a bucket of water and slowly poured it over the towel, asking [Coffman] questions every once in a while” (Coffman 1983, p. 211, lines 1–6). Coffman was being questioned over a theft of money from the property room (as a trusty he had access to it). The deputies divided their labor neatly, with Glover pouring the water, Lee refilling the bucket, and both Glover and Baker asking questions. Coffman struggled so much that even with the handcuffs the deputies had difficulty holding him down, so they tied his feet together with belts and secured his feet to his hands with handcuffs. When asked at trial about why he was struggling, Coffman said he was frightened. He was “afraid of drowning; it was hard to breathe. … I was just trying to fight that bucket of water” (Coffman 1983, p. 212, lines 10, 12). “It seemed like it lasted forever, but I guess maybe two hours” (Coffman 1983, p. 213, lines 8–9). Coffman testified that he knew nothing about the money, but told two lies to stop the torture, and eventually the deputies stopped.
About one year later, on a late September evening in 1980, James Hicks was transferred from the Montgomery County jail to the new San Jacinto County jail.4 Accused of taking San Jacincto County Sheriff James Parker’s tractor, Hicks denied knowing anything about it. Parker responded that Hicks would be “begging to tell them where the tractor was” and Hicks was taken to a small holding cell in the back of the jail. After half an hour, Hicks was made to change into white coveralls and then, after about another half hour, he was taken by Deputy Baker to a detox cell. The cell was about ten feet by eight feet with a “little concrete ledge built on the back part of the cell” six or eight inches off the floor, which sloped to a drain in the middle (Hicks 1983, p. 23, lines 21–22). The only piece of furniture in the room was a wooden armchair.
Baker kept asking him where “our” tractor was. Other deputies and a trusty came in and out. At one point, Hicks was blindfolded. His hands were handcuffed behind his back and he was sat in the chair. Deputy Baker gave leg irons to a trusty and instructed him to shackle Hicks’s legs to the legs of the chair. Hicks stood up and kicked the trusty. A struggle ensued with deputies Baker and Lee struggling to shackle Hicks to the chair. Lee would hit Hicks with a blackjack everytime Hicks kicked him, at least six times, so that his eyes began to swell nearly shut very quickly.
Once Hicks was subdued and shackled to the chair, the deputies tied a rope around his midsection and Deputy Baker covered Hicks’s nose and mouth with a towel. They set the chair backside down on the floor with the legs resting on the concrete ledge and Hicks’s head down on the floor, sloping toward the drain. “Then they poured water over my face through the towel and drowned me. … I was drowning. … I thought I was going to drown” (Hicks 1983, p. 29, lines 20–24; p. 30, line 5). This went on for “probably” five to ten minutes, though “it seemed like a lot longer” (Hicks 1983, p. 30, lines 1–3).
After “stomping” on Hicks’s stomach, the deputies set the chair back up, removed the blindfold, and began questioning him again about the location of the tractor. Eventually, Hicks gave them a location, though he knew there was no tractor there. When asked why he told the deputies that, Hicks said, “I didn’t want them to do it again. I would have took them anywhere, just about” (Hicks 1983, p. 33, lines 4–5).
According to one of the now infamous “torture memos” from the Justice Department’s Office of Legal Counsel (OLC), in waterboarding
the individual is bound securely to an inclined bench, which is approximately four feet by seven feet. The individual’s feet are generally elevated. A cloth is placed over the forehead and eyes. Water is then applied to the cloth in a controlled manner. As this is done, the cloth is lowered until it covers both the nose and mouth. Once the cloth is saturated and completely covers the mouth and nose, air now is slightly restricted for 20 to 40 seconds due to the presence of the cloth. This causes an increase in carbon dioxide level in the individual’s blood. This increase in the carbon dioxide level stimulates increased effort to breathe. This effort plus the cloth produces the perception of “suffocation and incipient panic,” i.e., the perception of drowning. The individual does not breathe any water into his lungs. During those 20 to 40 seconds, water is continuously applied from a height of twelve to twenty-four inches. After this period, the cloth is lifted, and the individual is allowed to breathe unimpeded for three or four full breaths. The sensation of drowning is immediately relieved by the removal of the cloth. The procedure may then be repeated. The water is usually applied from a canteen cup or small watering can with a spout. You have orally informed us that this procedure triggers an automatic physiological sensation of drowning that the individual cannot control even though he may be aware that he is in fact not drowning (Bybee 2002, p. 3–4).
KSM was led by a guard from his cell into another room in March 2003, in a secret prison in Stare Kiejkuty, Poland, one of the CIA’s “black sites.” The room, its essential contents—a tilted gurney or bench with arm and leg restraints, a couple of cloths, and a pitcher of water—and its purpose were all well known to Mohammed. Over the course of that month, he had spent a lot of time in this room, though he probably couldn’t say exactly how much. What happened now was more or less what had happened over 180 times since the beginning of March.
He was strapped to the “special bed,” with his head on the downward side of the incline (International Committee of the Red Cross 2007, p. 10). With his head held immobile by one of his captors, one of the cloths was placed over his face, covering his nose and mouth. With a doctor and a psychologist looking on, water was poured onto the cloth. KSM ingested enough water to “distend” his abdomen, and water gushed out when his stomach was pressed. To prevent him from ingesting the water, interrogators cupped their hands around his mouth in order to maintain a continuous “pool” of water. At one point, contractors Jessen and Mitchell seized an opportunity to pour water into his mouth when he attempted to speak (United States Senate 2014a, pp. 112, 114).
Now, suppose I tell you I am writing a history of the Inquisition and its trials and I write the following sentence about what happened to Marina Gonzàlez in 1494 and to Juan de Salas in 1527: “Upon the refusal of Gonzàlez and Salas to provide any information, the inquisitors employed enhanced interrogation techniques.” Or what if I’m writing about imperial competition and I write the following sentence about the Amboyna case of 1622: “At a time of increasing tensions with the British and the possibility of a plot to seize the strongest of the Dutch fortifications, the commander ordered enhanced interrogation techniques be used upon John Clarke.” Or suppose I’m writing about police interrogation methods in the United States and I use the following to describe what happened in San Jacinto County in 1983: “In a departure from usual police practice, Sheriff Parker used enhanced interrogation techniques.”
Do these descriptions ring hollow to you? Do they sound anachronistic? Do they sound euphemistic? Now suppose it is the year 3157. You are writing a history of interrogations and torture and cover the same cases we just considered above. Would you use the word “torture” to describe the Gonzàlez, Salas, Bourdil, the Alabama prisoner, Navarro, Nielsen, Harrison, Alleg, Nguyen, Hicks, and Coffman cases, but use the phrase “enhanced interrogation techniques” to describe the KSM case? Would you claim that pouring water over a cloth into someone’s mouth is torture in 1494, 1527, 1622, 1726, 1881, 1942, 1958, 1969, and 1983, but not in 2003? Probably not. But this is what the Bush administration’s Justice Department attempted to claim.
The U.S. Department of Justice has not always held this view. Here’s the phrase used repeatedly by the U.S. district attorney in the trial of the San Jacinto County sheriff and his deputies in 1983: “water torture” (Coffman 1983, p. 213; Hicks 1983, p. 5). Even the lawyers for the defendants used the term “water torture” to describe the events (Coffman 1983, pp. 225, 226, 230).
The only difference between the techniques used on James Hicks and KSM is two decades. A cloth covers the mouth, water is poured over the cloth to produce a drowning sensation—the act is the same. If the act itself does not change, then it is the same thing, whether we use a euphemism or whether or not the laws written around it have changed. Waterboarding is torture. It was torture when it was used by the Inquisition in 1494 and it was torture when it was used by the CIA in 2003. It meets the everyday “interocular trauma” test.5
What about the other enhanced techniques, individually and in combination? Do they constitute torture? How effective is torture in eliciting information and intelligence? The next section takes up the first two questions before turning to the effectiveness question.
ARE “ENHANCED INTERROGATION TECHNIQUES” TORTURE?
When most of us think of the word torture, the first thing that comes to mind are images of medieval dungeons with racks and red-hot pokers or perhaps the more modern electrocution. These techniques, however, can cause permanent physical injury or at least permanent physical scars, signs that the body had been tortured at one time. Such evidence is inconvenient in democratic regimes which are more subject to public monitoring and which make explicit claims to protecting human rights. As a result, democracies invented what Darius Rejali in his masterful study has called “clean” or “stealthy” methods of torture (Rejali 2007, p. 8). Such methods are designed to accomplish exactly the same goals as the previous methods (confessions, information, intimidation), but leave no marks.
There may be a further advantage of such techniques for democratic regimes if the methods are ever revealed to the public, as they were under the Bush administration: They do not seem “as bad” as those earlier “medieval” methods, making them more palatable for at least some people. But to say that sleep deprivation or stress positions are not “as bad” as the rack or electrocution or splinters under the fingernails or even waterboarding is both trivially true and beside the point. The fact that Andy is a bigger jerk than Billy does not at all mean that Billy is not a jerk. This is, though, a common “defense” of torturers (Conroy 2000, p. 112).
The Bush interrogation program was full of such so-called torture lite techniques (Bowden 2003, p. 53). The military employed the following measures: yelling, light deprivation, stress of a female interrogator, long (20 hours) interrogations, removal of comfort items (includes religious items), forced grooming, stress positions, isolation, hooding, nudity, and inducing fears—for example, with dogs (Phifer 2002, pp. 1–3)). The CIA, however, went much further.
One of the Justice Department memos by Jay Bybee to the CIA approved the following methods: attention grasp, walling, facial hold, insult (facial) slap, cramped confinement (perhaps with insects), stress positions (including wall standing), and sleep deprivation (Bybee 2002, pp. 1–4)). A later internal CIA memo defined “standard techniques” as ‘isolation, reduced caloric intake, use of loud music and white noise, and diapering (Central Intelligence Agency 2003c, p. 1). The same memo added “abdominal slap” to the roster of “enhanced techniques” (Central Intelligence Agency 2003c, p. 2). A CIA Office of Medical Services memo from December 2004 listed shaving, stripping (nudity), hooding, and isolation as the least intensive of the techniques, with continuous light or darkness, uncomfortably cool environment, and shackling in various positions further up the scale (Central Intelligence Agency 2004b, p. 9). The secret prison or “black site” in Bucharest, Romania, apparently even had cells specially constructed on “springs to keep the floor shifting and the prisoners constantly destabilized.”6 Then, of course, there were the standard conditions of confinement (white noise/loud sounds and constant light) and “conditioning techniques” (nudity, dietary manipulation, and sleep deprivation) (Central Intelligence Agency 2004a, pp. 5–6)). Finally, the procedures during the capture and transport (rendition) of the detainee to the black site, to which many if not all the detainees in the program were subjected, included cutting clothes from the body, stripping, nude photographs, shackling, sensory deprivation in the form of blindfolds, hoods, and earmuffs/headphones, a rectal exam, and, in some cases, forcible sodomy in the form of sedation by suppository (Carle 2011; Marty 2006, p. 24).
In fact, the Bush administration wasn’t even original in the name they gave the program. The Nazis beat them to it. The Gestapo euphemism for its interrogational torture program was verschärfte Vernehmung—which translates as “sharpened” or “enhanced” interrogations (Sullivan 2007). Some of the techniques called for under verschärfte Vernehmung are identical to the EIT program, including food manipulation (reduced food), darkness, and sleep deprivation. Like the CIA, the Nazis also imposed limits on who “qualified” to receive the treatment and which techniques required the presence of a doctor.
As difficult as any of these techniques were when used in isolation, their combined use was more powerful than the sum of their individual parts. Indeed, the Office of Legal Counsel (OLC) in the Department of Justice had become concerned by the end of 2004 about the effects of the techniques used in combination. The OLC requested more information from the CIA, and the latter responded with a December 2004 memo. This memo, along with CIA cables and emails, eyewitness testimony, and other governmental and non-governmental investigations, provides a comprehensive, detailed, and chilling picture of what rendition and torture “lite” looked like.
In fact, it is possible to narrate what the memo called a “prototypical” rendition and interrogation based on this memo, described on the cover as a “generic description of the process,” as well as other declassified CIA documents, eyewitness testimony, journalist reports, and government and non-governmental investigations (Central Intelligence Agency 2003c; Central Intelligence Agency 2004a; Central Intelligence Agency 2004b; Marty 2006; International Committee of the Red Cross 2007; United States Senate 2014a; Carle 2011). In the following story, I have made up the names and some small details for narrative purposes, but the description hews closely to CIA guidelines and descriptions.
“A PROTOTYPICAL INTERROGATION”
Khalid al Zushiri was just exiting the jetway into the airport terminal when the two police officers standing by a door asked him his name. When he confirmed it, they opened the door and politely asked him to step in. Nervous and unsure what was going on, he complied and was greeted by four large men dressed from hooded heads to booted toes in black. All he could see were their eyes through little holes in their masks. He noticed one pair was blue just before a blindfold was tied around his head from behind. The ninjas surround him immediately, one patting him down, one handcuffing his wrists, and another shackling his feet together. The cuffs and foot manacles were then chained together.
Scarcely before he knew what was happening and could struggle or even utter a word of protest, he could feel his clothes being cut from his body and hear the pieces being searched and tossed into a bag. By the time his shirt had been cut clear off his body and put into a bag, he recovered enough from his initial shock to protest and resist but he was told, in a menacing and low tone, to stay quiet. Many pairs of hands held him fast. The ninjas pulled his shoulders back so that he stood up straighter and then let him go, stepping away from his body. For a moment no one seemed to move. Then, even through his blindfold, he detected a flash of light. Suddenly the blindfold was removed and, right before the flash went off again, he saw one of the ninjas pointing a camera at him. He was quickly turned in the opposite direction, there was another flash, and the blindfold covered his eyes once again.
If he was surprised by how quickly they had removed his clothes, leaving him standing stark naked and shackled in a roomful of strangers, he was even more shocked by what came next. Latex-covered fingers began to explore his body, starting at the top and working their way down. They rustled through his hair, then over his ears, and his lips, opening and probing in his mouth. The hands then hoisted him onto a table, at first face down, and then turned on his side. They pulled his knees up to his chest. When he felt the latex gloves start to spread his buttocks, he struggled and squirmed, but he couldn’t move an inch. He gasped and contracted involuntarily as the latex-covered finger probed his anus. Its removal came as a relief, but only for a moment, for the finger returned, this time inserting something in his rectum and leaving it there.
The hands stood him back up on the floor and removed the handcuffs and foot manacles before pulling something over his legs. It felt small like underwear, rather than pants, but of a strange material. As it was secured around his waist, he realized it was an adult diaper. They then began to dress him; and by the time they were putting his arms in the sleeves, he realized it was a one-piece jumpsuit. Any hope this would all now stop once he was dressed ended when he was put back in the cuffs and manacles. None of the men had said a word, not to him and not to each other. Someone then placed earphones or earmuffs on his head, adjusting them so that they stayed snug over his ears. Now it was so silent that his own breathing sounded loud. A long and loose hood dropped over his head, resting on his shoulders.
The hands then half-carried, half-walked him across the room to a van parked right next to the door. He could tell it was a van because they laid him down roughly on the floor and the ninjas all got in with him. The van drove for only a minute or two before stopping. The door opened and the hands dragged him out. They again half-carried, half-walked him to some steps. He noticed that he had started to feel a little woozy and had trouble moving his feet properly for the steps. The hands didn’t seem to notice, though, and he was propelled upward for a couple of seconds. One of the hands pushed his head down and he stepped onto a new surface. As they pulled and pushed him along, he banged into the sides and realized they were plush armrests. He was on an airplane, a luxury one. Panic again started to rise in his throat, but it competed with sleepiness now and didn’t get very far. He was set down on the floor, and his cuffs and manacles were chained to the seat legs on the floor of the plane. Someone pulled his hood back for a moment and then they let him go. There was movement back and forth around him for a minute or two before the floor started rumbling and vibrating as the plane start to taxi. The last thing he felt before he drifted into unconsciousness was the floor pressing on him as the plane ascended steeply into the air.
Barely 20 minutes had passed since he stepped into the room.
He had vague memories of take-offs and landings, people bustling around him, and someone lifting his hood every once and awhile, but he didn’t fully come to until he found himself bouncing around in the back of small van. He asked where they were taking him, why they had kidnapped him. When no one answered, he raised his voice, but was told to keep quiet. The door to the vehicle opened and he felt very cold air before being hustled inside a building. They dragged him along, turning once or twice, before standing him up against a wall. He heard a door close behind him.
Someone removed the hood but he remained blindfolded. Again he tried to protest and again he was told to keep quiet. As several large, strong men held him tight, including his head, another began cutting the hair off his head, adjusting the blindfold when necessary. When they had cut everything off close to his scalp, they switched to an electric razor and shaved him bald before shaving off his sparse beard. Then they removed his manacles, blindfold, diaper, and jumpsuit, and took several photographs of him front and back.
After he had been put back in the diaper and manacled sitting down in a chair, a new masked man came into the room. He told Khalid he was going to give him a physical examination and started asking Khalid questions about his health while he probed and prodded. Did he have any allergies? Any history of heart problems? Khalid lurched back and forth between trying to answer these questions and begging the “doctor” to find out why he had been kidnapped as well as asking, What is going on? The doctor said only that Khalid “should start cooperating” and that the sooner he cooperated, the better off he’d be, while listening to his heartbeat with a stethoscope.
Following the medical questions, the doctor began asking him other questions, including whether anyone in his family had a mental illness. Despite his terror, Khalid was indignant enough to retort that the only crazy people he knew were standing around in masks. Immediately upon uttering these words, one of the men holding him slapped him hard across the face. Khalid was quiet except for answering the doctor after that. When the doctor was finished, they put the blindfold and hood back on him, stood him up, and led him along a short hallway to another room. The manacles made it difficult to keep up with the guards. Inside the room they adjusted his chains in order to shackle him to a chair. The room was cold and he requested clothes. They ignored him. When they removed the hood and blindfold, he blinked and squinted, for the room was all white and very brightly lit. The only other object in the room beside the chair was a bucket and a horizontal bar over his head near the ceiling. One of the masked ninjas told him they’d be back soon to question him and that Khalid would be given one chance to cooperate with them. Suddenly, as he shut the door, screaming-type rock music began blasting from a speaker near the ceiling.
True to their word, they came back sometime later; exactly how long he couldn’t say. They removed his diaper and walked him naked, hooded, and manacled, back down the hallway to the first room, where he was again chained to the chair. When they removed the hood, there was an American sitting on a chair across from him, the two chairs centered in the small room. The American showed him some pictures of young men and asked Khalid to identify them. When Khalid said that he couldn’t, that he didn’t know any of them, the American put the pictures down and, saying this was Khalid’s “last chance,” asked for the email address of a Walid bin al-Shibawi. Sweating heavily, Khalid claimed he’d never heard of al-Shibawi and had no idea what his email address was. He tried to continue, protesting his innocence, protesting that they had the wrong man, but his words were cut short by a slap to his stomach from one of the masked guards. The American told Khalid that they would do what it took to get information from him, no one knew where he was, and that things were about to get a lot harder for him. Then he rose from his chair and left the room without another word. The ninjas unchained him from the chair, refastened his manacles, put his hood on, and returned him to his cell.
This time his handcuffs were chained to the bar near the ceiling, with the chain adjusted so his arms were fully extended up toward the ceiling. If he slumped, his wrists pulled painfully against the cuffs. His foot manacles were chained to the floor. The screaming music now alternated with a loud hissing sound, like steam escaping from a radiator. Sometimes it sounded more like crackling. He was still nude, save for the diaper, and the room was cold. To his surprise, when they removed the hood and blindfold, it was still pitch dark. They had turned the lights off. At least, he thought, he had some measure of privacy now and might even be able eventually to sleep, despite the painful pressure on his arms. A splash of cold water in his face disabused him of that illusion. Every time he would start to fall asleep, one of the guards would slap him or spray water in his face.
They came for him about a day later, though he couldn’t tell whether it was day or night. After he was led into another room, he felt a towel go around his neck and panicked for a moment, thinking they meant to strangle him. But the towel was held tight only against the back of his neck, not his throat, and the hood was removed.
A new American stood in front of him now, holding the ends of the towel. He explained to Khalid that he was alone with them and no one knew where he was. His future depended completely on whether and how he cooperated with them. If he cooperated, his situation would improve immediately. If it didn’t, then it would get worse—a lot worse. Khalid again tried to protest that they had the wrong man, but was slapped hard by the American before he could finish his sentence.
Begging and protesting only earned him more slaps, in the face and on the stomach. Then suddenly the man holding the towel spun him around once and slammed Khalid into the wall. The towel prevented his head from bearing the full force of the blow. Khalid was surprised at the sound it made, and he realized after several more throws against the wall that it was made of plywood and covered the masonry wall behind it. He still didn’t tell them what they wanted to hear, and the slaps and throws against the wall continued. Other times they would make him stand leaning against a wall, holding himself up by his fingertips or using only his forehead, with his hands cuffed behind his back and his legs spread wide apart. When he dropped his arms or tried to reposition himself, the guards slapped him and put him back against the wall until he finally collapsed.
This cycle of hanging from a chain in his cell, sometimes in the dark, sometimes under bright lights, followed by the wall slamming, slapping, and standing against the wall, went on for days, perhaps weeks. He lost track of time. They fed him only a nutrition drink and he was always naked, cold, hungry, and filthy, for he hadn’t bathed in weeks. He was forced to clean himself when the diaper was removed using only a bucket of water and his hand.
One day, when the hood was pulled off in the interrogation room, he noticed what looked like a simple coffin on the floor. As his eyes widened, the interrogator told him the only way he’d be leaving the prison was in that box. Following another round of fruitless questioning, the guards forced him into the box and closed the lid. Khalid now noticed the holes in the box for breathing, but he couldn’t look out of them. The guards had draped a cloth over them and it quickly became very hot and stuffy. He screamed and pushed against the box but couldn’t get out. Buffeted by the constant loud music, he lost track of time, eventually soiling himself.
At some point, the box opened and he was roughly dragged out and held firmly by the guards in front of the interrogator. More questions. More responses deemed unacceptable. More slaps and walling. Then they turned him around and he saw a much smaller box, less than three feet on a side. His struggles went nowhere as they dragged him to the little box. The ninjas stuffed Khalid into the small box, his knees jammed tightly against his chest in a fetal position, and locked the lid. It was dark, hot, and stuffy and once more the harsh music blared. This lasted for several hours before he was pulled out and questioned again.
He could no longer keep track of the sequence; but various combinations of standing for hours, hanging from the handcuffs, being splashed with water, slapping, wall-slamming, and being stuffed in the boxes went on for days. A few days after his first experience with the boxes, they once more brought him to the interrogation room. When they removed his hood, he saw something new: a portable bed like you see in a hospital, but with leather straps on the sides. On a table next to the gurney was a small towel and a dozen bottles of spring water. He wasn’t sure what it all meant and at first he didn’t even struggle as they led him to the gurney. As they began to tie him to it, though, he tried to protest and wiggle out, to no avail. The bed was tilted with his head on the down side. One of the interrogators gripped Khalid’s head so that he couldn’t twist it to the side. He looked up and saw the other interrogator’s face as a cloth was placed over his mouth and nose. He screamed as the interrogator began soaking the cloth with water.
This is torture “lite.”
It would be pedantic in the extreme to document here how these methods have their origins in traditional tortures and have traditionally and legally been considered torture. Others have already established this very well (Rejali 2007, McCoy 2006, Otterman 2007, Conroy 2000, Wallach 2006, Department of Justice 2009, pp. 251–254, 260–261; Ohlin 2010, Parry 2010, Scharf 2010, Haas 2009, Schwarz and Huq 2013, pp. 65–82)). Even our own military considered them torture less than a decade before 9/11. A U.S. Army Field Manual classified several of these techniques (“infliction of pain through … bondage,” “forcing an individual to stand, sit, or kneel in abnormal positions for prolonged periods of time,” “any form of beating”) as “physical torture” and others (“food” and “sleep deprivation”) as “mental torture” and expressly forbid their use, arguing that “the use of force is a poor technique, as it yields unreliable results, may damage subsequent collection efforts, and can induce the source to say whatever he thinks the interrogator wants to hear” (Department of the Army 1992, pp. 1–8)).
Americans are certainly skeptical of claims that the techniques do not constitute torture. Following the release of the Senate Intelligence Committee’s report on the CIA interrogation program in December 2014, Huffington Post/YouGov asked Americans whether it is acceptable or unacceptable for the United States to use some of the CIA’s interrogation methods. Americans responded as shown in Table 2.1.7 Rectal feeding, by which pureed food is forced up a detainee’s anus, received the least support, followed by the confinement box, threatening a detainee or his family, waterboarding, nudity, and walling. Only sleep deprivation was deemed acceptable by an absolute majority, with slapping and punching considered acceptable by a plurality.
Table 2.1 DECEMBER 2014 HUFFINGTON POST/YOUGOV SURVEY ON TORTURE TECHNIQUES
A CBS poll at the same time asked whether respondents consider four techniques to be torture or not.8 Seven in ten viewed threatening family members (73%), sleep deprivation (70%), and waterboarding (69%) to be torture. Another 57% considered ice water baths to be torture. In short, like waterboarding, the other EITs pass the interocular trauma test.
“Enhanced interrogation techniques” is a euphemism for interrogational torture.
The “tough” statements of Bush administration and CIA officials about going to the “dark side” and the “gloves coming off” suggest that what they have really objected to was the possibility of legal proceedings, not the fact of torture itself. Given their political and personal responsibility to prevent another attack after 9/11, an attack they thought imminent, one can sympathize with their dilemma. (Try to imagine for a moment that it was your job to keep America safe while the smoke still rose from lower Manhattan.) Like many of their predecessors (including many Democrats by the way), they believed (and no doubt still believe) that interrogational torture works. If everyone—including Bush administration officials—know that the EIT program was torture, it is all the more important that the program was effective in gathering otherwise unobtainable information that saved innocent lives.
Still, it does not follow from the fact that interrogational torture’s effectiveness is in the political interest of its proponents that interrogational torture was not effective. We take up the question of effectiveness next.
A FRAGILE AND DANGEROUS THING
Gasping for breath, Ammar has just been waterboarded but has yet to give up any information on the Saudi group to Daniel, a CIA interrogator. Daniel tells him, “It’s cool that you’re strong. I respect it, I do. But in the end everybody breaks, bro. It’s biology” (Boal 2011, p. 6). After some humiliation, time in a confinement box, deception, and the temptation of good solid food, Ammar does indeed “break” and give up some names which will eventually help lead to the location of bin Laden. This may be a scene from the Hollywood movie Zero Dark Thirty, but it captures the widely held assumption that everyone breaks under torture. Just contemplating a routine trip to the dentist makes the claim that torture works seem reasonable.
Does it? Does torture work? The Roman jurist Ulpian said that torture “is a weak and dangerous thing that may fail the truth.”9
This commonplace idea that torture eventually works on everyone, the gut feeling that there is only so much any human can withstand, is an instance of what we might call the dangerous seduction of intuition. It is seductive because it seems so self-evidently obvious. It is dangerous because it can nevertheless be wrong. After all, Senator John McCain gave up no useful information to his North Vietnamese captors; he just provided the names of the Green Bay Packers offensive line as fellow members of his squadron and the names of cities that had already been bombed as supposed future targets (Salter and McCain 1999, p. 194).
Not that it is surprising that we are seduced by it. There are plenty of other examples of such successful seductions. It seems pretty reasonable (and did to humans for thousands and thousands of years before Galileo and Copernicus) that the sun revolves around the earth. It does not. It seems reasonable that the earth is flat, but it is a sphere. It seems obvious that the various features of humans and other animals must have been designed, yet logical reasoning and overwhelming empirical evidence demonstrates that evolution by natural selection and random mutation accounts for what we see. It seems reasonable that measuring something cannot change the nature of the thing you are measuring, yet quantum mechanics tells us that this is not true at the subatomic level. These examples point to the fact that intuition can sometimes lead us astray. Interrogational torture is another one of those instances, as the Gonzàlez, Salas, Bourdil, and other cases underscore.
These cases (and there are others) challenge our intuition that torture always works. Still, one case (or even a few cases) cannot tell us very much about how it works in general. This is an empirical question and really requires an empirical answer—that is, an answer based on actual evidence or data. Empirical evidence comes in two forms: observational data we would get from the real world of interrogations and experimental data we would get from the artificial world of running an experiment.
In the latter, we would start with the “null” or baseline hypothesis that torture is not effective in eliciting information. We would then see whether we could, through lots of experiments, amass evidence showing that torture did generate information. If we have enough, that would give us enough confidence to reject the “ineffective” hypothesis. Those experiments would involve randomly assigning people to groups with different information levels and types and degrees of torture.
Now, if you felt like you were reading a page from Nazi doctor Joseph Mengele’s diary when you read this description of an experiment, then you understand the obvious ethical problem ruling out an experimental approach to the problem. So experimental data are out. What about observational data? It comes in two forms: the Full Monty or dribs and drabs. Each has its problems.
In the Full Monty version, an analyst would sit down in front of a CIA computer and open up a database of all CIA interrogations. This file would include everything the CIA knows about the detainees, including their background, the information they provided, whether or not they were tortured and, if so, which techniques, and so on. Each row represents the interrogation of a detainee; and all the other information, such as the amount of information they provided or the severity of the torture employed, is given in the columns.
Ethical considerations forbid this type of research as well, as the CIA itself recognized in its response to the Senate Intelligence Committee report. Explaining why no assessment of effectiveness was ever completed, it noted, among other factors, that “Federal policy on the protection of human subjects” would have “encumbered” a “systematic study over time of the effectiveness of the techniques” (Central Intelligence Agency 2014, p. 48). In other words, it would be unethical for the CIA (or anyone else) to figure out whether what the CIA did was effective or not!
But let’s press on. The nice thing about the Full Monty version is that everything that can be known is right there. If you’ve ever seen a spreadsheet, it’s easy to imagine what that data file might look like.
The first column would be filled with the names of the detainees, with the succeeding columns containing the data on the variables for each detainee. So the CIA might have some scale for each of the variables in the columns after the detainee’s name, let’s say 0 to 100, with 0 representing no information or information of no value and 100 representing lots of information or extremely high-quality information. Similarly, the CIA might have some scale for the severity of the torture employed, from no torture, 0, to the maximum, perhaps 100.10 There might be a variable for other interrogation methods used by the FBI and military interrogators.
One might think that the analyst could then just perform some pretty simple statistical tests to compare using torture against other methods and to look for the effects of torture on information. Do the numbers in the information columns tend to go up as the numbers in the torture columns go up (supporting the pro-torture position)? Or do they go down (supporting the anti-torture position)? Or is there no real tendency at all (again supporting the anti-torture position)?11 If there is a tendency to go up, how strong does this tendency have to be to say torture works? We would have to decide this (ahead of time!), but this would seem to be the right way to go about answering the question.
Not so fast. The not-so-nice thing about even the Full Monty version is that everything that can be known is not everything that needs to be known. Leaving secrecy considerations aside, any such database would suffer from methodological, empirical, and epistemological problems.
Methodologically, the relationship between torture and information is not as straightforward as it might seem. The type and severity of torture employed by the interrogator was not independent of the information acquired and believed to be possessed by the detainee. In fact, the more likely a detainee was to have information, the more likely he was to get tortured. In other words, torture is endogenous to information, biasing any attempts to infer the effectiveness of torture.12
The empirical problem concerns the ratio of the complexity of the cases to the number of cases. The individual interrogation cases (one detainee, one set of interrogations) are likely to be very different from each other, if not unique. They will differ along many dimensions: each detainee’s position in the enemy organization, what they know, how fresh that information is, their level of training, their cultural background, their idiosyncratic strength of will and pain thresholds, the circumstances of their arrest, the techniques employed and how they were employed (timing, sequence, severity, number of repetitions), the quantity of information divulged, if any (however you would measure that), and the quality of any information divulged (ditto). Each one of these factors would be an additional column in the spreadsheet.
All this, by the way, assumes no cases of mistaken identity, something we know does happen (more on this later). Against this complexity is the number of interrogations in which some coercive methods were used. Although there have been thousands of interrogations in Iraq, Afghanistan, Guantanamo Bay, and various black sites, it would appear that the number of interrogations employing torture is small (in absolute as well as relative terms) even if we include the renditioning and conditioning techniques (as we should). This presents a real problem for teasing out whether torture works: Lots of variables and few cases will create the problem of indeterminacy. You just cannot draw any conclusions from the data.
Here’s a more prosaic example of this problem: home brewing. When you brew beer at home, as I do, a thousand things can screw up your beer or at least make it come out differently than you thought (and hoped): different combinations of malt extract and hops; bad extract or stale hops; bad yeast pack; sluggish initial yeast production; contamination of the wort during cooling; sanitization problems with the fermenters, spoons, bottles, kegs, etc.; poor temperature control; and lots more. So if you end up with skunky beer, it can be pretty hard to figure out what happened and make the appropriate changes. This is especially true if you only brew a couple times a year, as I do.
The same is true with interrogations. There just are not enough cases of coercive interrogations (one would hope, since the claim was to use them in only extreme cases) to draw firm conclusions about whether they worked or not. In truth, the only way to avoid the indeterminacy problem is to torture more. Since the complicating factors are not going to change, the only way to counteract their effect is to increase the number of cases in which torture is used, so that there are more cases (row entries) with each particular combination of factors (values in the columns). This takes us back into Mengele territory. Moreover, it bears emphasizing that this would only make doing the analysis possible; whether or not that analysis would actually end up showing that interrogational torture works would still be up for grabs.
The problem is, however, even worse than this. Even if this problem were somehow circumvented, there would still be a fundamental epistemological problem—and always will be. Those cool fMRI pictures of the brain from notwithstanding, we can never know what is in another person’s head. This is just as true for the CIA. Indeed, this fact is what make interrogation necessary in the first place. But it results in two problems, one if we observe information and one if we do not.
Suppose we observe that a detainee has provided useful information. The analyst is sitting at that computer in CIA headquarters and sees that in case 16, a detainee revealed the location of a safe house, but nothing else after repeated waterboarding. Assume for simplicity that quality is not a problem; it was truly valuable information (the house was still active and previously unknown to us). How would we know whether to put this down as a case of “a little information,” “some information,” or “full information,” let alone a particular number between 0 and 100? The fact is, we cannot know. It may be that this was all the detainee knew, but it may also be the case that it was a small fraction of what he knew.13
Now suppose we observe that a different detainee has failed to provide useful information. The analyst has moved on down to case number 17 in the CIA data file. In this case, the detainee refused to provide anything of value even after waterboarding. What can we infer from this? The detainee might be hiding information, but it might be the case that he truly does not know anything of value; our lack of access to a detainee’s mind means that the detainee can never prove what he does not know. This is just as true of the completely innocent as it is of the detainee who has provided everything of value he knows, but is asked to provide more. Thus, the two problems of observing information and observing no information are actually equivalent. If we expect the detainee to provide some information X, and he provides some other information Y, where Y can be anything from false information to “I don’t know,” then we can never know whether the absence of X is due to deception or ignorance.
The upshot is that these data cannot tell us whether torture works as claimed because we will never know whether the detainee knew more than was revealed, and this will be true of any CIA dataset, past, present, and future. There will never be a column in that CIA file called “Total information known by the detainee.” These fundamental and irresolvable ethical, methodological, empirical, and epistemological problems mean that even the Full Monty would not solve the problem.
So, with the Full Monty not an option either, we are left with what evidence of torture’s effectiveness has emerged in dribs and drabs over the centuries of the practice, from the Ancient Greeks to more recent torture by France, the United States, the United Kingdom, and Israel. It does not add up to anything very clear. A review of cases from the early history of juridical and ecclesiastical interrogational torture (Pennington 2008, Lea and Peters 1973, Peters 1999, Langbein 2006) to the use of “stealthy” torture in the modern world, including democracies (Pfiffner 2014, Rejali 2007), provides plenty of examples of false confessions and fabricated information, obstinate resistance to torture, and far fewer cases of torture producing information. Nevertheless, the “data” emerging in dribs and drabs from the bleak history of interrogational torture are “too fragmentary, dirty, and wrapped in national mythology, accusation, and rumor to allow for precise, validated causal claims” (Rejali 2007, pp. 7, 566). In other words, the information we have available about torture’s effectiveness in generating information falls far short of the normal social scientific standard required to draw firm conclusions one way or another.
Not even the Senate Intelligence Committee’s comprehensive (over 6,700 pages and over 38,000 footnotes relying on CIA cables, emails, interviews, and other documents) report—discussed in the Postscript—of the CIA program has been able to convince defenders of the program (United States Senate 2014a). The Republican minority on the committee released its own report; and former members of the Bush administration, including President Bush himself, came forward to vigorously defend the program (United States Senate 2014b). The CIA too defended itself in a response originally written back in June 2013, but released with the committee’s summary and findings in December 2014 (Central Intelligence Agency 2014).
Public opinion mirrors this disagreement and ambivalence, as captured in four surveys taken shortly after the release of the report.14 The question wording differs slightly across the four polls, asking whether the CIA methods resulted in “mostly reliable” information, “produced valuable information,” “provided intelligence,” and the like. I put the poll results together in Table 2.2 (collapsing “often” and “sometimes” into Reliable and “rarely” and “never” into Unreliable for the CBS News survey). There is as little consensus among the broader public as there is among politicians. A little more than half think that the methods generate reliable information in the first three surveys, with the other half saying that they are unreliable or are not sure. The Huffington Post/YouGov survey finds a plurality not sure, but essentially a three-way tie. Certainly the evidence available has failed to convince torture proponents. Bagaric and Clarke, in their defense of interrogational torture, dismiss the evidence as “a distracting and superficial numbers game” in which torture proponents and opponents hurl competing anecdotes at one another (Bagaric and Clarke 2007, p. 58).
Table 2.2 SUMMARY OF SURVEYS ON INFORMATION FROM TORTURE
The argument thus far has shown that
1. EITs are torture and the effectiveness of interrogational torture is an open question. (Chapter 2)
The next step in the argument is to examine how closely the Bush program approximates the ideal model of interrogational torture and identify the benchmarks defining success of the pragmatic model.