11

Torturing Innocents, Resisting Torture

Each blow stupefied me a little more, but at the same time confirmed me in my decision not to give way to these brutes … I remained silent. “He’s playing games with us! Put his head under again!”

—HENRI ALLEG AND HIS TORTURER, La Question, 1957

In the two (sets of) equilibria discussed in this chapter, no Detainee type provides information to the Interrogator. In one set, the Interrogator tortures in response; in the other she does not. Each occurs under both types of questioning. In this chapter we examine both sets, organizing the discussion by the actions underlying them, rather than by questioning type.

NO INFORMATION, TORTURE

In the no information, torture equilibrium, Cooperative and Innocent Detainees do not reveal information, either because they believe they will be tortured anyway (see Appendix A for the full set of beliefs) or because the Innocent Detainee does not have the move to reveal information under objective questioning. Either way, the Interrogator believes that the Detainee is hiding information rather than Innocent () and so tortures.

Equilibrium Features

The relevant thresholds in this equilibrium are thus the information revelation threshold(s) and, in the case of leading questioning, for the Innocent Detainee, the Innocent Detainee recognition threshold , and the information hiding threshold . The way in which they bound the equilibrium is depicted in gray in Figure 11.1, with all the thresholds placed in their usual locations based on their theoretical or empirical properties discussed in Chapter 7. The white space behind the gray-shaded area (everything greater than or equal to ) is the next (and final) equilibrium we discuss later in the chapter: the no information, no torture equilibrium.

Figure 11.1 No Information, Torture Equilibrium

The Detainee q thresholds provide the upper bound or “roof” of the equilibrium. The Cooperative Detainee believes the Interrogator will torture anyway and so refuses to give up what he knows, whether the questioning is objective or leading. In the latter case, the Innocent Detainee has the option of talking, but also believes he will be tortured anyway and so maintains his innocence. What about the area to the north of ?

In this region, the Cooperative Detainee does believe that Interrogator is Pragmatic, but also believes that the Interrogator will torture anyway because she (the Interrogator) will think he has more information to provide or is willing to confirm (). Finally, the Interrogator’s Innocent Detainee recognition threshold encloses the “back end” of the equilibrium. The Interrogator has observed “no information,” and her belief that this is due to an Innocent Detainee is below her threshold . She’s too worried it’s a Resistant Detainee trying to get away with hiding the information and so she tortures.

Although we have said (Observation 7.2) that is likely to approach one-half from the zero side, the greater the degree of confidence the Detainee has that the Interrogator will believe he has revealed all his information (i.e., as drops), the more the upper part of the region shrinks to the left. (This also assumes that the Detainee believes that the Interrogator is Pragmatic [i.e., ].) The less confident she is, the more it moves toward one-half, expanding the maximum.¹

The situation is similar for the Innocent Detainee recognition threshold . It also is less than, but close to, one-half. Lowering it so that the region is reduced requires the Interrogator to lower his threshold, to be more willing to believe that the Detainee is Innocent after observing “no information.” The less willing she is to believe this, as seems empirically likely, the more the threshold pushes toward one-half, expanding the region.

Now consider how the movement of Detainee information revelation thresholds and affect this equilibrium. The more valuable the information v to the Cooperative Detainee and the more important it is to the Innocent Detainee not to falsely confess l, the higher this threshold and so the larger the region. The region can shrink so that torture without information takes up less total space in two ways.

First, the information value v (or desire not to falsely confess l) might be lower. The less valuable the information, the more willing the Cooperative Detainee is to reveal it—that is, the lower the threshold, pushing the region down. The same is true for the Innocent Detainee’s desire not to falsely implicate himself. Second, the Interrogator can drive down the threshold by increasing the severity of the torture, k. In other words, you can get less torture overall with more severe torture or you can get less severe torture but at the expense of expanding the region. This is the same conclusion we came to in Implication 10.1, since the “no information” region is just the complement of the “valuable information” region in that equilibrium.

Substantive Interpretation

This equilibrium corresponds to a very important outcome in the real world: A detainee fails to provide valuable information and is tortured. It remains forever unclear whether the torture was “justified” because the detainee actually possessed the valuable information or was unjustified even according to the pragmatic view because the detainee was innocent or cooperative but innocent (i.e., ignorant) of that particular piece of information. Even in the former case, it is forever unclear whether more (severe) torture would have compelled the hidden information because the detainee was cooperative or whether no amount of torture would have wrung the information out of him (because he was Resistant).

For a case in the real world to match the “no information, torture” outcome, it must meet the following conditions:

1. The detainee was any of the three types.

2. The detainee was subjected to objective or leading questioning.

3. The detainee gave up no or very little valuable information.

4. The detainee was tortured afterward.

We have already seen a real-world case with Pacha Wazir after he was taken out of Carle’s hands and rendered to the Salt Pit in Afghanistan. In the view of CIA headquarters, he had not provided the information they sought and so was tortured at the Salt Pit. The other cases to explore are the Resistant and Innocent types, particularly since, as we have seen, both are tortured in other equilibria for playing “no information.” We have cases clearly illustrating each type: Khaled El-Masri, an innocent unemployed German car salesman kidnapped by the CIA and tortured in black sites in Morocco and Afghanistan, and Henri Alleg, a French newspaper editor and journalist tortured by French paratroopers during the Battle of Algiers in 1957.

Torturing Innocents: Khaled El-Masri

On the last day of 2003, El-Masri left Ulm, Germany, on a long bus ride to Skopje, Macedonia, where he planned to take a short vacation.² At 3:00 that afternoon the bus crossed from Serbia into Macedonia. Border guards there considered El-Masri’s new German passport suspicious and took him off the bus to look through his baggage and question him further about connections to Islamic groups and terrorism. He denied any connection. After seven hours, a group of armed men in plainclothes drove him in a convoy of three cars to the Skopski Merak Hotel in the Macedonian capital, across the street from the zoo.

A nine-member Macedonian security team held him in a room on the top floor for the next 23 days, with three of them present at all times. He was interrogated constantly. They repeatedly denied his requests to speak with someone from the German Embassy. He began a hunger strike on day 13 and didn’t eat another bite for the next ten days. At one point, he angrily tried to walk out of the hotel room, but the Macedonian officers drew their pistols and motioned him away from the door.

On January 22, 2004, a thousand miles away on a Spanish resort island in the Mediterranean, a seven-member CIA rendition team and three contract pilots were also staying at a hotel, the five-star luxury Marriot Club Son Antem. The next morning, one of them enjoyed a massage at the spa and by the afternoon, the entire team, two of whom were women, had checked out of the hotel individually and assembled at the Palma de Majorca airport. At 5:40 they took off in the CIA-contracted Boeing 737 jet, tail number N313P, and landed in Skopje just over two hours later.

Around the time the CIA jet was landing, El-Masri was asked to make a video statement that he had not been harmed and was then blindfolded, handcuffed, put into a waiting jeep, and driven to the airport. He sat alone in a chair, still blindfolded and still handcuffed, ostensibly waiting for a medical exam before being returned to Germany.

Suddenly two people pulled his arms back hard and others began to beat him from all sides. They cut his clothes away from his body and threw him on the floor. With a boot on his naked back holding him to the floor, someone forced a suppository into his anus. This done, they jerked him back to his feet, bound his feet together, and removed his blindfold. There was an immediate flash of a camera. As his vision cleared, he saw seven or eight men around him, clad in black, including black ski masks. They put a diaper and then a tracksuit on him. They shackled his feet and hands and attached both to a belt around his waist. They put pads over his eyes, a blindfold over that, and earmuffs over his ears before hooding him. They marched him (he shuffled with the short shackles) to the waiting Boeing, and once inside he was thrown face down onto the floor. His cuffs and foot shackles were detached from his belt just long enough to pull and manacle him into a spread eagle position. He could detect no seats, and it felt like what he imagined a cargo plane was like. They gave him at least two injections and he was unconscious for most of the flight to Kabul via Baghdad.

On landing he was put in a room long enough that he began to think from the newspapers there that he was in Afghanistan. Then they put him in the trunk of a car and drove for ten minutes. He awoke to find he was in a small, filthy, concrete cell furnished with a threadbare military-issue blanket and rags of clothes bundled into a pillow. A dirty plastic bottle contained yellow, foul-smelling water for drinking. There was one small window, high on the wall. Previous occupants had scratched in Arabic and Farsi on the walls. El-Masri had arrived at the Salt Pit, the brick factory converted into a secret CIA prison near the Bagram military airfield.

Later that same night, masked men took him for a medical exam, including taking photos of him naked and a blood sample. The doctor was also masked and dismissed his complaints about the water. The masked men came for him later the same night and took him from his cell to an interrogation room. All his interrogators were also masked. After four interrogations over four days, they deemed his refusal to acknowledge any connection with, or having information on, 9/11 hijacker Mohammed Atta or other terrorists as uncooperative. He was left alone in his cell for three weeks, let out only three times a day to use the toilet. The food gave him diarrhea.

Finally, with some of the other prisoners, he decided to try a hunger strike in March. Although others eventually gave in, El-Masri refused all food for 37 days, drinking only the stagnant water. After he refused their requests to end the strike, on April 10, 2004, his captors ended it for him. Four masked men entered his cell and shackled him in handcuffs and manacles and then carried him to the interrogation room. After binding him to a chair, one of the men pulled his head back and a doctor pushed a tube up his nostril and down his throat, force feeding him from a funnel at the other end. They threatened to keep doing it if El-Masri didn’t eat. Eventually, they agreed to give him better food and water and books to read and he agreed to end the strike.

On May 16, while still recovering from the hunger strike, “Sam,” a tall, blond Westerner who spoke German with a northern accent, interrogated El-Masri and later promised that he would be going back to Germany. When nothing happened for several days, El-Masri started another hunger strike. Then, on May 28, 2004, he was blindfolded, handcuffed, and driven away from the prison. He heard a plane land while he was locked in a cargo container. The door opened and he was ordered to change into his old clothes. When he had done so, he was again blindfolded, the earmuffs were put on, and he was taken to the waiting plane. They chained him to his seat.

After landing some hours later, his manacles were exchanged for a rope binding his hands but the blindfold remained. He was put into a car with several men. Over the course of the next seven or eight hours, the car drove over rough and bumpy roads, up and down mountains, changing drivers and passengers several times. There was almost no talking. Then the car stopped and he was taken out.

When they removed his blindfold, he could see that it was dark and they had pulled off a deserted stretch of a dirt road. They returned his passport and suitcase to him and untied the rope around his wrists. They pointed to a path and told him to start walking and not to look back.

His spine tingled as he walked, and he wondered whether they were going to shoot him in the back.

But he did as they instructed and when he rounded a bend, three Albanian border guards were waiting for him with a packed lunch. They drove him to the airport in Tirana, Albania’s capital. After paying for his own ticket back to Germany, they shepherded him through customs and immigration with no inspections and put him on a plane to Frankfurt. When he arrived in Frankfurt—unkempt, long hair, shaggy beard, and sixty pounds lighter—the German officer at passport control demanded to see other documents because he looked so different from the passport photo taken just eight months earlier.

The reason for his release? His name had been confused with a Khalid al-Masri with suspected connections to terrorists because of a transliteration error. Even so, many in the CIA said the agency should wait until the Germans checked out his passport before rendering him. The Al Qaeda unit head at the CIA—the hard-charging red-headed woman behind the “Maya” character in the film Zero Dark Thirty—overruled them, however. She had a “hunch” El-Masri was involved in terrorism (Grey 2006, p. 94).³

Resisting Torture: Henri Alleg

A little before four o’clock in the afternoon on June 12, 1957, Henri Alleg knocked on the door of a friend’s third-floor apartment in Algiers.⁴ It was opened not by his friend Maurice Audin, but by a detective. Alleg ran back down the stairs trying to escape. He had nearly made it to the street when the detective caught up with him. With the revolver at his back, he walked back up to Audin’s apartment. The detective put in a quick call to the “paras”—paratroopers—and by four o’clock, they and the police had arrived. The lieutenant from the 10th Paratrooper Division considered Alleg an “excellent catch,” for Alleg was the former editor-in-chief of the banned Alger Républicain, a left-wing daily supportive of Algerian independence.

In truth, Alleg was not surprised the moment had come. He had gone underground eight months before, not much more than a year after the Alger Républicain had been banned by the French authorities and after his co-workers were starting to be rounded up by the police and paras. But he had continued to write on behalf of Algerian independence in the communist daily l’Humanité from his various places in hiding.

The paras wanted to know who was doing that hiding and where. The lieutenant asked him exactly that before they had even cuffed him. Alleg’s response was “That I won’t tell you!” (Alleg Calder and Sartre 2006, p. 38).

He was led handcuffed down the stairs to a passenger car waiting across the street. The barrel of the guard’s submachine gun poked into his ribs during the ride into the upper part of Algiers. They stopped at a construction site near the El-Biar Square. Alleg was taken out of the car and marched across the courtyard in front of a large apartment building under construction, then into the building itself and up the stairs.

There were no handrails on the steps. Rebar poked out from the bare ends of unfinished concrete walls. Electrical wires hung from the ceilings. Guards shoved disheveled prisoners up and down stairs and in and out of rooms. Several stories up, Alleg’s guards took him into what might one day be someone’s living room. They took off his handcuffs and a different lieutenant waiting there handed him a pencil and paper. He politely told Alleg to write down who protected him, what he had been doing, with whom he had met, and so on. Alleg responded,

“I have nothing else to say to you. I shall write nothing and don’t count on me to betray those who have had the courage to hide me” (Alleg Calder and Sartre 2006, p. 40).

A young sunburned para in a blue beret then fetched him down one floor to the kitchen. There, next to the sink and oven, he was ordered to undress. At first he refused, but when they said they would do it by force if necessary, he complied. Naked, he was ordered to lie down on a vomit-covered wooden plank and his ankles and wrists were fastened to the plank by leather straps. The short and sunburned soldier stood over him, confidently telling him that “Everybody talks. You’ll have to tell us everything—and not only a little bit of the truth, but everything!” (Alleg Calder and Sartre 2006, p. 41). He started to shiver and the blue berets thought he was afraid. No, he said, he was just cold.

A captain then joined the two lieutenants and the four enlisted men in the kitchen. Upon Alleg’s refusal to say anything more, they lifted him into the adjoining room where the light was better. The soldiers then brought in boxes for the officers to sit on. The lieutenant who had arrested him showed Alleg the gégène, the field telephone. Alleg had written about others who had suffered from it and was all too aware of what would happen next. All he said, however, was a demand that the paras follow proper procedures in charging him and to address him appropriately—using the formal “vous” rather than the informal “tu” they had used up to that point.

A sergeant sat on his chest and showed Alleg the alligator clips leading from the gégène. Then he attached one clip to Alleg’s earlobe and the other to his finger, both on his right side. The first shock felt like lightning exploding next to his head, and he jumped against his bonds and screamed. The lieutenant timed the cadence of his question—“Where have you been hiding?”—with the shocks of the telephone.

Alleg’s answer, however, did not bear on the question, instead telling the lieutenant that he would regret torturing.

This did not make the lieutenant happy at all, and he turned the dial to its maximum. Despite this, Alleg still did not talk, and it wasn’t because they had stuffed his shirt into his mouth to muffle his screams. The sergeant moved one of the clips to Alleg’s penis to see if that might compel him to reveal his hiding place. Its only effect, however, was to cause such convulsions that they had to retie his straps.

So they tried other tricks. They stripped the wire and taped it across the width of his chest. They doused him with cold water for direct discomfort as well as to increase the effects of the electricity. The spectators drank beer. Alleg didn’t talk.

They untied him from the plank and began to beat him while some were sent to fetch his friend Audin from another building. They showed him a battered Audin, who told Alleg that “It’s hard, Henri,” before he was taken away (Alleg Calder and Sartre 2006, p. 41). Alleg’s continued refusal to talk began to frustrate the paras. The other lieutenant continued to beat him, moving from fists and knees to a piece of lumber.

He was taken back to the room with the gégène and tied to the plank once again. This time, however, they used a larger machine and Alleg could feel the difference, as his entire body spasmed. Still he did not talk.

Now he was asked if he knew “how to swim” and was carried on the plank to the kitchen. The end with his head rested over the sink and a couple of paras held up the end with his feet. The paras jammed a wooden wedge between his teeth and wrapped a rag around his mouth and nose. The para told him that when he was ready to talk, he should move his fingers.

Water flowed from the tap through a rubber hose onto the rag, soaking it.

At first he was able to snatch a breath or two and keep some of the water out of his throat. But it didn’t last long.

I had the impression of drowning, and a terrible agony, that of death itself, took possession of me. … In spite of myself, the fingers of my two hands shook uncontrollably (Alleg Calder and Sartre 2006, p. 49).

The paras misinterpreted this involuntary movement as willingness to talk and removed the rag in triumph. Alleg gulped in air as the captain pounded his stomach to expel the water Alleg had swallowed. When they realized he was staying silent, they put him back under.

Three times he experienced “the terrible moment where [he] felt [himself] losing consciousness, while at the same time fighting with all [his] powers not to die.” The first two times they brought him back from the brink, and he managed to both gasp for air and throw up water at the same time. The third time he blacked out.

He never again moved his hands.

When he regained consciousness, he found himself on the floor of the adjoining room, naked but removed from the plank. The paras pulled him up off the floor and began to beat him like a punching bag, tossing him back and forth between them. Then they tied his feet and suspended him by his feet from a shelf in the kitchen. One of the paras then lit a piece of paper like a torch and burned Alleg’s penis with it. They moved to other parts of Alleg’s body, his legs, his nipples. They beat him. They stepped on his fingers trailing on the hard floor.

After a while of this, he was half-dragged, half-carried to a cell. He crawled onto the mattress, but was stabbed when he put his weight on it. A guard outside the door laughed that he had put barbed wire in it. But his colleague recognized what Alleg had accomplished. “All the same, he has gained a night for his friends to get away” (Alleg Calder and Sartre 2006, p. 52).

The next day they gave him several more sessions with the gégène. Frustrated with Alleg’s continued refusal to talk, they stuck the wire into the back of his mouth. Once the shocks began, the paras didn’t even need to hold the wire; the current caused Alleg’s jaw to clench shut and his teeth held the wire in place, no matter how hard he tried to open his mouth. He felt like his eyeballs were being pushed out of their sockets from the inside. Smashing his head against the concrete brought some relief, but the shocks continued for a while longer.

He didn’t talk.

Sometime later they began to burn him again, scorching his nipples and the bottom of his swollen feet. He said nothing. Furious, they beat him. He said nothing. They tempted him with water, knowing the electricity had made him desperately thirsty. He said nothing. They punched him in the face. He said nothing. More shocks. He said nothing.

Later a high-ranking officer came to see him and offered to take Alleg back to France if he told what he knew; if Alleg refused he would “disappear” (Alleg Calder and Sartre 2006, p. 65). Alleg’s response?

“Too bad” (Alleg Calder and Sartre 2006, p. 65).

The officer told him his only option now was suicide and walked out.

Back in his cell, Alleg heard the screams of a woman being tortured nearby. For a while he thought he recognized his wife’s voice and that the paras had followed through on their threat to torture her to get him to talk. He stayed silent in his cell.

At one point he splashed water on his face and a guard mockingly asked him whether he was feeling better. Alleg replied, “Yes, you’ll soon be able to start on me again” (Alleg Calder and Sartre 2006, p. 71).

They didn’t though, not really. They tried a “truth serum,” but that didn’t work either. Henri Alleg never talked. He was transferred to another prison and eventually escaped his torturers, spending time in Czechoslovakia before making his way back to France.

Back to the Model

That El-Masri was innocent and had nothing to do with Al Qaeda or terrorism is now beyond dispute, though the United States has never apologized or recognized his claims and, in fact, used its diplomatic muscle to deter the nominally independent German judiciary from pursuing criminal action against the United States.⁵ During his captivity, El-Masri was apparently asked leading questions about his connections with the 9/11 hijacker Mohammed Atta and other terrorists. His resistance was remarkably powerful, suggesting his aversion against falsely implicating himself l was high, something also supported by his multiple hunger strikes. If he was asked objective questions, he could not have provided satisfactory answers to them and so was tortured.

Henri Alleg, in contrast, was no innocent swept up in the paratrooper’s broad net. His refusal to provide the information sought by the paras was met with multiple tortures, some very similar to the CIA and others well beyond what we know the CIA practiced, including electrical shocks and burning. And yet he never talked.

It is worthwhile noting that because of the peculiarity of these cases, we happen to have the background information necessary to establish their type (Innocent and Resistant, respectively). To an Interrogator at the time, however, their behavior was observationally equivalent. Alleg = El-Masri.

NO INFORMATION, NO TORTURE

In this equilibrium, no Detainee provides information but the Interrogator does not torture because she believes the Detainee is Innocent and not Cooperative or Resistant (i.e. ). The ideal outcome here is if the Detainee is in fact Innocent, for he is not tortured for telling the truth. If, however, he does have information, then he got away with hiding it. From the proponents’ point of view, this is particularly problematic if the Detainee was Cooperative, because he likely would have given up at least some of his information had the Interrogator tortured.

Equilibrium Features

There is only one threshold to consider in this equilibrium and that is the Interrogator’s innocent detainee recognition threshold . The reason for this is that the Detainee does not need to form beliefs about whether the Interrogator is Pragmatic nor what his information hiding threshold might be because the Interrogator does not torture after “no information.” No type of detainee can do any better than this, and so none of them provides information. Since , this equilibrium must take up more than half of the entire equilibrium space, as illustrated in Figure 11.2.

Figure 11.2 No Information, No Torture Equilibrium

This means that most of the parameter space is taken up by detainees who were questioned and threatened with torture but gave no information, and were never tortured. This makes sense theoretically, given that we built in disincentives to torture in order to give the pragmatic account, to give proponents, the best possible case they can make. But does it make sense empirically? How does this correspond to the real world?

For a case in the real world to match the “no information, no torture” outcome, it must meet the following conditions:

1. The Detainee was any of the three types.

2. The Detainee was subjected to objective or leading questioning.

3. The Detainee gave up no valuable information.

4. The Detainee was not tortured afterward.

What do we have that corresponds to this case?

Not much, unfortunately.

The problem is this. Once torture becomes an interrogation tool, it tends to be used. By the time a detainee ended up at a CIA black site, he had already been tortured and would continue to be tortured by the conditions of confinement and perhaps “standard” techniques—whether or not he was ever subjected to EITs. We just don’t have good accounts of someone who was brought before an interrogator in a torture program, questioned, threatened with torture, and provided no or misleading information and who was not tortured afterward because he was deemed innocent. Some must surely exist, but they are apparently few and far between.

We might think we could include some cases of military interrogations in Iraq and Afghanistan. In some of those cases, of course, detainees were actually threatened with torture and were actually tortured. In many others, though, military interrogators followed the old norms and never threatened torture (Alexander 2011, Alexander and Bruning 2008). What if at least some of those detainees feared they would be tortured if they did not satisfy their captors, even though their interrogators never had any intention whatsoever to do so, even if they refused to provide information? No doubt many of these detainees were not tortured despite having provided no information. This isn’t really close enough to qualify as cases for our model, though. We are modeling what happens when interrogators (their bosses!) decide to use torture to extract information. This element is missing from the non-torturous military interrogation cases and we are left wanting for examples. We return to this problem in Chapter 13.

The argument thus far has shown that:

1. EITs are torture and the effectiveness of interrogational torture is an open question. (Chapter 2)

2. The Bush program approximates closely the ideal model of interrogational torture and includes limits on torture; the Bush and ideal models provide benchmarks for comparison with the game theory models to come. (Chapter 3)

3. The Bush model generates strange, quixotic outcomes. (Chapter 4)

4. The Bush interrogational torture program is more realistically modeled as objective and leading question variants of an incomplete information game, with three types of detainees, two types of interrogators, and uncertainty about the amount and value of information provided. (Chapter 5)

5. By positing a set of Detainee strategies, calculating the Interrogator’s expected utility using Bayes’ Rule to identify her best response, and checking for incentives to deviate by any of the Detainee types, it is possible to derive a perfect Bayesian equilibrium in which a Detainee is tortured after providing information. (Chapter 6)

6. The RIT model generates nine perfect Bayesian equilibria, the formal and empirical characteristics of which generate important observations, propositions, and implications, including:

(a) The Interrogator’s thresholds for believing that a Detainee is Innocent after “no information” are less than one-half, with close to one-half and closer to zero.

(b) The Interrogator’s information hiding threshold under objective questioning is greater than or equal to one-half, whereas her information hiding threshold under leading questioning as well as the Detainee’s version of are a little less than one-half.

(c) Objective questioning (potentially) provides better information, but is necessarily accompanied by more torture, than leading questioning, which, however, provides less valuable information.

(d) All things begin equal, Interrogators are more likely to get less valuable information than highly valuable information. (Chapter 7)

7. Surprise torture of a Cooperative Detainee—even if he has provided all his information—is not only likely, but perversely more likely

(a) the more willing the Detainee is to divulge information and

(b) the more important the information is in terms of “unknown unknowns.” (Chapter 8)

8. The perversity under objective questioning persists under leading questioning, namely:

(a) The more the Innocent Detainee believes Interrogator promises of no torture in exchange for confirmation, the more likely is ambiguous and false information.

(b) The more brutal the torture, the more likely is ambiguous and false information.

(c) The more important the confirmation is to the Interrogator and the more difficult it is for the Detainee to understand what is being asked, the greater the likelihood of torture.

(d) Leading questioning will not eliminate torture; the higher the Interrogator’s demand for confirmation, the more torture. (Chapter 9)

9. The valuable information, selective torture equilibrium reveals four further trade-offs and a paradox:

(a) Everything else being equal, eliciting information

i. requires either more frequent or more brutal torture,

ii. is more likely when the information is less valuable or the torture is more severe, but not both,

iii. is more likely when the standard of cooperation is lowered, even though it increases the likelihood of information hiding,

iv. is necessarily accompanied by innocent torture; decreasing the likelihood of innocent torture requires decreasing the likelihood of information.

(b) The very assumptions justifying interrogational torture paradoxically make it less reliable in terms of the likelihood of getting valuable information. (Chapter 10)

10. From the two equilibria in which no Detainee provides information we learn that:

(a) Everything else being equal, torture becomes less likely (the equilibrium region shrinks to a smaller proportion of the parameter space) after “no information” only if the information becomes less valuable or the torture becomes more severe, not both.

(b) It is difficult to find real-world cases in which detainees subjected to the threat of torture in interrogations were not tortured after failing to provide information. (Chapter 11)

The next step in the argument is to summarize the results of our inquiry and compare them to the benchmarks from Chapter 3.