In the first part of this series, I described the torture of Jose Padilla, and started to answer the question: why is this happening? I described the "mosaic theory" of intelligence gathering -- the view that one should try to collect tiny pieces of intelligence and fit them together into a bigger picture.
As I said in that post, the mosaic theory is fine in itself, but it does provide a reason to hold people indefinitely, since one can never know when they might produce some tiny snippet of information whose importance they might not even be aware of. And I quoted an FBI affidavit that suggested that the government regarded the mosaic theory in that way.
Nonetheless, the FBI's use of the mosaic theory has always been limited. The FBI operates within the United States, and both the information it gathers and its treatment of prisoners, have to stand up in court. For this reason it can neither detain prisoners indefinitely, nor violate their rights while they are in custody.
The CIA is in a different situation entirely. It normally operates outside the US, and its goal is normally not criminal prosecution but intelligence gathering. For this reason it is not subject to the same strictures as the FBI or domestic police forces. It does not need to justify its detention of people to US courts, or to interrogate them in ways that would stand up in a court of law. It would be surprising if its interrogation procedures did not reflect these differences.
However, there are practical limits on what the CIA can do. One of the most important, for our purposes, is that it is hard for it to detain prisoners indefinitely. It would have to do so outside the US in order to avoid the requirements of the US legal system, and this means that its detentions either exist at the sufferance of other governments, whose continued acquiescence cannot be counted on, or rely heavily on secrecy, which cannot be counted on either. In either case, it's hard to see how the CIA could count on being able to detain people -- especially large numbers of people -- indefinitely. Possibly the CIA has detained some people for long stretches of time, but these would have to be a small number of exceptional cases.
This means that the CIA has also had to operate within certain practical limits. I imagine that there are many fewer limits on what the CIA can do in a particular case than there are for the FBI -- to my knowledge, the President cannot order the FBI to assassinate someone, for instance. But the CIA cannot detain any significant number of people for long stretches of time without risking discovery and the compromise of its operations. And that means that, for very different reasons, neither our domestic law enforcement agencies nor our intelligence services could seriously consider the possibility of detaining people in perpetuity.
After 9/11, however, the administration decided that this had changed, and it came up with several ways of detaining people indefinitely: prisons in Afghanistan, Guantanamo, and elsewhere, and the idea of designating people as 'unlawful enemy combatants' to avoid both US law and the Geneva Conventions. In so doing, it created enough of a legal rationale that agencies that had always seen themselves as bound by the law had some reason for thinking that they could go along with things like the indefinite detention of a US citizen against whom no charges had been filed, and who had no access to counsel.
This created a possibility that neither domestic law enforcement agencies nor the intelligence community had needed to take seriously before: the possibility of detaining large numbers of people for as long as necessary, without having to answer to the US courts, foreign governments, or anyone else. I suspect that the fact that they had not needed to take this possibility seriously before meant that they had not thought it through seriously: earlier, the question how one would handle the indefinite detention of large numbers of people would have been just idle speculation.
Moreover, the point of this detention, after 9/11, was partly to get dangerous people into custody and off the street, but also, and importantly, to gather intelligence. And this fact, I think, meant that the interrogations used on detainees would have been based on practices used in the intelligence community, not domestic law enforcement. This is the kind of interrogation that the intelligence community has always done, and it stands to reason that its procedure would have been adopted. (Anyone remember all those news stories about the FBI scrambling to adjust to the post-9/11 world of intelligence gathering as opposed to law enforcement?)
So: what were those CIA procedures?
The best source I've been able to find (from Joseph Margulies again) is the KUBARK manual, available online here and here. It's a CIA interrogation manual dating from 1963, and released as the result of a FOIA request in 1996. (According to Wikipedia, KUBARK was a code name the CIA used for itself.) You might ask: why rely on a manual from 1963? Well: it was still being used as the basis for training manuals as of 1983. It's also very good, as manuals for interrogation go: the author is plainly very smart and thoughtful, if a bit chilling, and it's very perceptive. If I ran an agency that had a continuing need for such a manual, which God forbid, I would have updated it (the manual spends a lot of time discussing relevant psychological research, which has progressed a lot since 1963), but I would not be in a hurry to replace it altogether.
The main reason for thinking that it can illuminate present interrogation practices, though, is simply that it describes them so well. For instance, here's the advice on detention (sec. IX.D):
"If, through the cooperation of a liaison service or by unilateral means, arrangements have been made for the confinement of a resistant source, the circumstances of detention are arranged to enhance within the subject his feelings of being cut off from the known and the reassuring, and of being plunged into the strange. Usually his own clothes are immediately taken away, because familiar clothing reinforces identity and thus the capacity for resistance. (Prisons give close hair cuts and issue prison garb for the same reason.) If the interrogatee is especially proud or neat, it may be useful to give him an outfit that is one or two sizes too large and to fail to provide a belt, so that he must hold his pants up.
The point is that man's sense of identity depends upon a continuity in his surroundings, habits, appearance, actions, relations with others, etc. Detention permits the interrogator to cut through these links and throw the interrogatee back upon his own unaided internal resources."
This is echoed in another passage, describing an interrogation technique that the manual calls "Alice in Wonderland" (VIII.C.9):
"The aim of the Alice in Wonderland or confusion technique is to confound the expectations and conditioned reactions of the interrogatee. He is accustomed to a world that makes some sense, at least to him: a world of continuity and logic, a predictable world. He clings to this world to reinforce his identity and powers of resistance.
The confusion technique is designed not only to obliterate the familiar but to replace it with the weird. Although this method can be employed by a single interrogator, it is better adapted to use by two or three. When the subject enters the room, the first interrogator asks a doubletalk question -- one which seems straightforward but is essentially nonsensical. Whether the interrogatee tries to answer or not, the second interrogator follows up (interrupting any attempted response) with a wholly unrelated and equally illogical query. Sometimes two or more questions are asked simultaneously. Pitch, tone, and volume of the interrogators' voices are unrelated to the import of the questions. No pattern of questions and answers is permitted to develop, nor do the questions themselves relate logically to each other. In this strange atmosphere the subject finds that the pattern of speech and thought which he has learned to consider normal have been replaced by an eerie meaninglessness. The interrogatee may start laughing or refuse to take the situation seriously. But as the process continues, day after day if necessary, the subject begins to try to make sense of the situation, which becomes mentally intolerable. Now he is likely to make significant admissions, or even to pour out his story, just to stop the flow of babble which assails him. This technique may be especially effective with the orderly, obstinate type."
Likewise, an interrogation technique called "Regression":
"There are a number of non-coercive techniques for inducing regression. All depend upon the interrogator's control of the environment and, as always, a proper matching of method to source. Some interrogatees can be regressed by persistent manipulation of time, by retarding and advancing clocks and serving meals at odd times -- ten minutes or ten hours after the last food was given. Day and night are jumbled. Interrogation sessions are similarly unpatterned the subject may be brought back for more questioning just a few minutes after being dismissed for the night. Half-hearted efforts to cooperate can be ignored, and conversely he can be rewarded for non-cooperation. (For example, a successfully resisting source may become distraught if given some reward for the "valuable contribution" that he has made.) The Alice in Wonderland technique can reinforce the effect. Two or more interrogators, questioning as a team and in relays (and thoroughly jumbling the timing of both methods) can ask questions which make it impossible for the interrogatee to give sensible, significant answers. A subject who is cut off from the world he knows seeks to recreate it, in some measure, in the new and strange environment. He may try to keep track of time, to live in the familiar past, to cling to old concepts of loyalty, to establish -- with one or more interrogators -- interpersonal relations resembling those that he has had earlier with other people, and to build other bridges back to the known. Thwarting his attempts to do so is likely to drive him deeper and deeper into himself, until he is no longer able to control his responses in adult fashion."
The attempt to disrupt everything familiar and to replace it with a world of surreal cruelty is a persistent theme in all the detainee stories. Things that make no sense are done to and around detainees, apparently for no reason at all. It's hard to read any of the detainees' stories without thinking: their captors seem to have been making a deliberate and well-thought-out attempt to create a world in which nothing makes any sense whatsoever, and there is no hope, and the world outside has simply dropped away.
On social isolation (Sec. VIII.A):
"If the interrogatee is under detention, the interrogator can also manipulate his environment. Merely by cutting off all other human contacts, "the interrogator monopolizes the social environment of the source."(3) He exercises the powers of an all-powerful parent, determining when the source will be sent to bed, when and what he will eat, whether he will be rewarded for good behavior or punished for being bad. The interrogator can and does make the subject's world not only unlike the world to which he had been accustomed but also strange in itself - a world in which familiar patterns of time, space, and sensory perception are overthrown."
On sensory deprivation (Sec. IX.E, after a survey of relevant research):
"These findings suggest - but by no means prove - the following theories about solitary confinement and isolation:
1. The more completely the place of confinement eliminates sensory stimuli, the more rapidly and deeply will the interrogatee be affected. Results produced only after weeks or months of imprisonment in an ordinary cell can be duplicated in hours or days in a cell which has no light (or weak artificial light which never varies), which is sound-proofed, in which odors are eliminated, etc. An environment still more subject to control, such as water-tank or iron lung, is even more effective.
2. An early effect of such an environment is anxiety. How soon it appears and how strong it is depends upon the psychological characteristics of the individual.
3. The interrogator can benefit from the subject's anxiety. As the interrogator becomes linked in the subject's mind with the reward of lessened anxiety, human contact, and meaningful activity, and thus with providing relief for growing discomfort, the questioner assumes a benevolent role. (7)
4. The deprivation of stimuli induces regression by depriving the subject's mind of contact with an outer world and thus forcing it in upon itself. At the same time, the calculated provision of stimuli during interrogation tends to make the regressed subject view the interrogator as a father-figure. The result, normally, is a strengthening of the subject's tendencies toward compliance."
Recall this passage from Padilla's brief:
"His isolation, furthermore, was aggravated by the efforts of his captors to maintain complete sensory deprivation. His tiny cell – nine feet by seven feet – had no view to the outside world. The door to his cell had a window, however, it was covered by a magnetic sticker, depriving Mr. Padilla of even a view into the hallway and adjacent common areas of his unit. He was not given a clock or a watch and for most of the time of his captivity, he was unaware whether it was day or night, or what time of year or day it was."
"For a long time Mr. Padilla had no reading materials, access to any media, radio or television, and the only thing he possessed in his room was a mirror. The mirror was abruptly taken away, leaving Mr. Padilla with even less sensory stimulus."
"Several guards in camouflage and riot gear approached cell No. 103. They unlocked a rectangular panel at the bottom of the door and Mr. Padilla’s bare feet slid through, eerily disembodied. As one guard held down a foot with his black boot, the others shackled Mr. Padilla’s legs. Next, his hands emerged through another hole to be manacled.
Wordlessly, the guards, pushing into the cell, chained Mr. Padilla’s cuffed hands to a metal belt. Briefly, his expressionless eyes met the camera before he lowered his head submissively in expectation of what came next: noise-blocking headphones over his ears and blacked-out goggles over his eyes. Then the guards, whose faces were hidden behind plastic visors, marched their masked, clanking prisoner down the hall to his root canal."
No light, no social contact, the window carefully masked, his meals passed through a slot in the door, even his mirror taken away, having to wear a hood, blackened goggles, and noise-cancelling headphones whenever he left his cell: this is sensory deprivation, and virtually complete social isolation. Padilla lived this way for years.
Stress Positions (from sec. IX.H):
"It has been plausibly suggested that, whereas pain inflicted on a person from outside himself may actually focus or intensify his will to resist, his resistance is likelier to be sapped by pain which he seems to inflict upon himself. "In the simple torture situation the contest is one between the individual and his tormentor (.... and he can frequently endure). When the individual is told to stand at attention for long periods, an intervening factor is introduced. The immediate source of pain is not the interrogator but the victim himself. The motivational strength of the individual is likely to exhaust itself in this internal encounter.... As long as the subject remains standing, he is attributing to his captor the power to do something worse to him, but there is actually no showdown of the ability of the interrogator to do so." (4)"
I take it I don't need to repeat any of the many stories concerning the use of stress positions in interrogations.
What is the point of all of these bizarre techniques? The KUBARK manual is quite clear. From Sec. IX.L:
"A brief summary of the foregoing may help to pull the major concepts of coercive interrogation together:
1. The principal coercive techniques are arrest, detention, the deprivation of sensory stimuli, threats and fear, debility, pain, heightened suggestibility and hypnosis, and drugs.
2. If a coercive technique is to be used, or if two or more are to be employed jointly, they should be chosen for their effect upon the individual and carefully selected to match his personality.
3. The usual effect of coercion is regression. The interrogatee's mature defenses crumbles as he becomes more childlike. During the process of regression the subject may experience feelings of guilt, and it is usually useful to intensify these.
4. When regression has proceeded far enough so that the subject's desire to yield begins to overbalance his resistance, the interrogator should supply a face-saving rationalization. Like the coercive technique, the rationalization must be carefully chosen to fit the subject's personality.
5. The pressures of duress should be slackened or lifted after compliance has been obtained, so that the interrogatee's voluntary cooperation will not be impeded."
The two key steps are inducing regression, and then presenting compliance as a way out, together with a rationalization that enables the subject to choose it. I will say more about regression in the next installment. For now, however, I want to note one important thing:
The KUBARK manual describes an interrogation process with a determinate endpoint: compliance. Moreover, the interrogator can know when that endpoint has been reached: after all, s/he is told to slacken the pressure on the subject once the subject has complied, and that would be pointless advice if the interrogator could never know whether compliance had been achieved, or if the process of achieving it was endless.
By contrast, the mosaic theory, according to which each and every piece of information a subject knows might be important, whether or not the subject realizes it, and according to which the interrogator therefore tries to extract every shred of knowledge a subject possesses, is essentially endless. Neither the interrogator nor the subject is ever in a position to know whether one last seemingly unimportant detail has yet to be revealed. For this reason, absent the constraints of the legal system, or something else that places limits on this sort of interrogation, it will never be completed.
If my experience thinking about this is any guide, it takes a fair amount of reflection on this before you see that there's a huge problem with trying to combine the KUBARK manual with the mosaic theory. And I have both the benefit of hindsight and the luxury of not working right after 9/11, with the lives of my fellow citizens in my hands. So I can understand why, at the time, people might not have seen that there was a serious conceptual problem with trying to combine the two; and a conceptual problem that could have horrific results. Namely:
If you remove the legal constraints on interrogation, and simultaneously try to wed the KUBARK model of interrogation to the mosaic theory, you risk using KUBARK techniques to induce regression and confession, without taking the KUBARK manual's next steps: providing an avenue for compliance, and terminating the interrogation. Instead of going through steps one through five above, you risk getting caught in an endlessly repeating loop of steps three and four, since step five -- the achievement of the goal of the interrogation -- is in principle impossible to reach.
More on the implications of this later.