The CIA's 1963 Torture Manual In Its Entirety, Part IV

Josh Clark

2. The Reconnaissance

If the interrogatee is cooperative at the outset or if rapport is established during the opening phase and the source becomes cooperative, the reconnaissance stage is needless; the interrogator proceeds directly to detailed questioning. But if the interrogatee is withholding, a period of exploration is necessary. Assumptions have normally been made already as to what he is withholding: that he is a fabricator, or an RIS agent, or something else he deems it important to conceal. Or the assumption may be that he had knowledge of such activities carried out by someone else. At any rate, the purpose of the reconnaissance is to provide a quick testing of the assumption and, more importantly, to probe the causes, extent, and intensity of resistance.

During the opening phase the interrogator will have charted the probable areas of resistance by noting those topics which caused emotional or physical reactions, speech blocks, or other indicators. He now begins to probe these areas. Every experienced interrogator has noted that if an interrogatee is withholding, his anxiety increases as the questioning nears the mark. The safer the topic, the more voluble the source. But as the questions make him increasingly uncomfortable, the interrogatee becomes less communicative or perhaps even hostile. During the opening phase the interrogator has gone along with this protective mechanism. Now, however, he keeps coming back to each area of sensitivity until he has determined the location of each and the intensity of the defenses. If resistance is slight, mere persistence may overcome it; and detailed questioning may follow immediately. But if resistance is strong, a new topic should be introduced, and detailed questioning reserved for the third stage.

Two dangers are especially likely to appear during the reconnaissance. Up to this point the interrogator has not continued a line of questioning when resistance was encountered. Now, however, he does so, and rapport may be strained. Some interrogatees will take this change personally and tend to personalize the conflict. The interrogator should resist this tendency. If he succumbs to it, and becomes engaged in a battle of wits, he may not be able to accomplish the task at hand. The second temptation to avoid is the natural inclination to resort prematurely to ruses or coercive techniques in order to settle the matter then and there. The basic purpose of the reconnaissance is to determine the kind and degree of pressure that will be needed in the third stage. The interrogator should reserve his fire-power until he knows what he is up against. 3. The Detailed Questioning

a. If rapport is established and if the interrogatee has nothing significant to hide, detailed questioning presents only routine problems. The major routine considerations are the following:

The interrogator must know exactly what he wants to know. He should have on paper or firmly in mind all the questions to which he seeks answers. It usually happens that the source has a relatively large body of information that has little or no intelligence value and only a small collection of nuggets. He will naturally tend to talk about what he knows best. The interrogator should not show quick impatience, but neither should he allow the results to get out of focus. The determinant remains what we need, not what the interrogatee can most readily provide.

At the same time it is necessary to make every effort to keep the subject from learning through the interrogation process precisely where our informational gaps lie. This principle is especially important if the interrogatee is following his normal life, going home each evening and appearing only once or twice a week for questioning, or if his bona fides remains in doubt. Under almost all circumstances, however, a clear revelation of our interests and knowledge should be avoided. It is usually a poor practice to hand to even the most cooperative interrogatee an orderly list of questions and ask him to write the answers. (This stricture does not apply to the writing of autobiographies or on informational matters not a subject of controversy with the source.) Some time is normally spent on matters of little or no intelligence interest for purposes of concealment. The interrogator can abet the process by making occasional notes -- or pretending to do so -- on items that seem important to the interrogatee but are not of intelligence value. From this point of view an interrogation can be deemed successful if a source who is actually a hostile agent can report to the opposition only the general fields of our interest but cannot pinpoint specifics without including misleading information.

It is sound practice to write up each interrogation report on the day of questioning or, at least, before the next session, so that defects can be promptly remedied and gaps or contradictions noted in time.

It is also a good expedient to have the interrogatee make notes of topics that should be covered, which occur to him while discussing the immediate matters at issue. The act of recording the stray item or thought on paper fixes it in the interrogatee's mind. Usually topics popping up in the course of an interrogation are forgotten if not noted; they tend to disrupt the interrogation plan if covered by way of digression on the spot.

Debriefing questions should usually be couched to provoke a positive answer and should be specific. The questioner should not accept a blanket negative without probing. For example, the question "Do you know anything about Plant X?" is likelier to draw a negative answer then "Do you have any friends who work at Plant X?" or "Can you describe its exterior?"

It is important to determine whether the subject's knowledge of any topic was acquired at first hand, learned indirectly, or represents merely an assumption. If the information was obtained indirectly, the identities of sub-sources and related information about the channel are needed. If statements rest on assumptions, the facts upon which the conclusions are based are necessary to evaluation.

As detailed questioning proceeds, addition biographic data will be revealed. Such items should be entered into the record, but it is normally preferable not to diverge from an impersonal topic in order to follow a biographic lead. Such leads can be taken up later unless they raise new doubts about bona fides .

As detailed interrogation continues, and especially at the half-way mark, the interrogator's desire to complete the task may cause him to be increasingly business-like or even brusque. He may tend to curtail or drop the usual inquiries about the subject's well-being with which he opened earlier sessions. He may feel like dealing more and more abruptly with reminiscences or digressions. His interest has shifted from the interrogatee himself, who jut a while ago was an interesting person, to the atsk of getting at what he knows. But if rapport has been established, the interrogatee will be quick to sense and resent this change of attitude. This point is particularly important if the interrogatee is a defector faced with bewildering changes and in a highly emotional state. Any interrogatee has his ups and downs, times when he is tired or half-ill, times when his personal problems have left his nerves frayed. The peculiar intimacy of the interrogation situation and the very fact that the interrogator has deliberately fostered rapport will often lead the subject to talk about his doubts, fears, and other personal reactions. The interrogator should neither cut off this flow abruptly nor show impatience unless it takes up an inordinate amount of time or unless it seems likely that all the talking about personal matters is being used deliberately as a smoke screen to keep the interrogator from doing his job. If the interrogatee is believed cooperative, then from the beginning to the end of the process he should feel that the interrogator's interest in him has remained constant. Unless the interrogation is soon over, the interrogatee's attitude toward his questioner is not likely to remain constant. He will feel more and more drawn to the questioner or increasingly antagonistic. As a rule, the best way for the interrogator to keep the relationship on an even keel is to maintain the same quiet, relaxed, and open-minded attitude from start to finish.

Detailed interrogation ends only when (1) all useful counterintelligence information has been obtained; (2) diminishing returns and more pressing commitments compel a cessation; or (3) the base, station, [one or two words deleted] admits full or partial defeat. Termination for any reason other than the first is only temporary. It is a profound mistake to write off a successfully resistant interrogatee or one whose questioning was ended before his potential was exhausted. KUBARK must keep track of such persons, because people and circumstances change. Until the source dies or tells us everything that he knows that is pertinent to our purposes, his interrogation may be interrupted, perhaps for years -- but it has not been completed. 4. The Conclusion

The end of an interrogation is not the end of the interrogator's responsibilities. From the beginning of planning to the end of questioning it has been necessary to understand and guard against the various troubles that a vengeful ex-source can cause. As was pointed out earlier, KUBARK's lack of executive authority abroad and its operational need for facelessness make it peculiarly vulnerable to attack in the courts or the press. The best defense against such attacks is prevention, through enlistment or enforcement of compliance. However real cooperation is achieved, its existence seems to act as a deterrent to later hostility. The initially resistant subject may become cooperative because of a partial identification with the interrogator and his interests, or the source may make such an identification because of his cooperation. In either event, he is unlikely to cause serious trouble in the future. Real difficulties are more frequently created by interrogatees who have succeeded in withholding.

The following steps are normally a routine part of the conclusion:

a. [approx. 10 lines deleted]

d. [approx. 7 lines deleted]

e. [approx. 7 lines deleted]

f. [approx. 4 lines deleted] C. Techniques of Non-Coercive Interrogation of Resistant Sources

If source resistance is encountered during screening or during the opening or reconnaissance phases of the interrogation, non-coercive methods of sapping opposition and strengthening the tendency to yield and to cooperate may be applied. Although these methods appear here in an approximate order of increasing pressure, it should not be inferred that each is to be tried until the key fits the lock. On the contrary, a large part of the skill and the success of the experienced interrogator lies in his ability to match method to source. The use of unsuccessful techniques will of itself increase the interrogatee's will and ability to resist.

This principle also affects the decision to employ coercive techniques and governs the choice of these methods. If in the opinion of the interrogator a totally resistant source has the skill and determination to withstand any con-coercive method or combination of methods, it is better to avoid them completely.

The effectiveness of most of the non-coercive techniques depends upon their unsettling effect. The interrogation situation is in itself disturbing to most people encountering it for the first time. The aim is to enhance this effect, to disrupt radically the familiar emotional and psychological associations of the subject. When this aim is achieved, resistance is seriously impaired. There is an interval -- which may be extremely brief -- of suspended animation, a kind of psychological shock or paralysis. It is caused by a traumatic or sub-traumatic experience which explodes, as it were, the world that is familiar to the subject as well as his image of himself within that world. Experienced interrogators recognize this effect when it appears and know that at this moment the source is far more open to suggestion, far likelier to comply, than he was just before he experienced the shock.

Another effect frequently produced by non-coercive (as well as coercive) methods is the evocation within the interrogatee of feelings of guilt. Most persons have areas of guilt in their emotional topographies, and an interrogator can often chart these areas just by noting refusals to follow certain lines of questioning. Whether the sense of guilt has real or imaginary causes does not affect the result of intensification of guilt feelings. Making a person feel more and more guilty normally increases both his anxiety and his urge to cooperate as a means of escape.

In brief, the techniques that follow should match the personality of the individual interrogatee, and their effectiveness is intensified by good timing and rapid exploitation of the moment of shock. (A few of the following items are drawn from Sheehan.) (32)

1. Going Next Door

Occasionally the information needed from a recalcitrant interrogatee is obtainable from a willing source. The interrogator should decide whether a confession is essential to his purpose or whether information which may be held by others as well as the unwilling source is really his goal. The labor of extracting the truth from unwilling interrogatees should be undertaken only if the same information is not more easily obtainable elsewhere or if operational considerations require self-incrimination.

2. Nobody Loves You

An interrogatee who is withholding items of no grave consequence to himself may sometimes be persuaded to talk by the simple tactic of pointing out that to date all of the information about his case has come from persons other than himself. The interrogator wants to be fair. He recognizes that some of the denouncers may have been biased or malicious. In any case, there is bound to be some slanting of the facts unless the interrogatee redresses the balance. The source owes it to himself to be sure that the interrogator hears both sides of the story.

3. The All-Seeing Eye (or Confession is Good for the Soul)

The interrogator who already knows part of the story explains to the source that the purpose of the questioning is not to gain information; the interrogator knows everything already. His real purpose is to test the sincerity (reliability, honor, etc.) of the source. The interrogator then asks a few questions to which he knows the answers. If the subject lies, he is informed firmly and dispassionately that he has lied. By skilled manipulation of the known, the questioner can convince a naive subject that all his secrets are out and that further resistance would be not only pointless but dangerous. If this technique does not work very quickly, it must be dropped before the interrogatee learns the true limits of the questioner's knowledge.

4. The Informer

Detention makes a number of tricks possible. One of these, planting an informant as the source's cellmate, is so well-known, especially in Communist countries, that its usefulness is impaired if not destroyed. Less well known is the trick of planting two informants in the cell. One of them, A, tries now and then to pry a little information from the source; B remains quiet. At the proper time, and during A's absence, B warns the source not to tell A anything because B suspects him of being an informant planted by the authorities.

Suspicion against a single cellmate may sometimes be broken down if he shows the source a hidden microphone that he has "found" and suggests that they talk only in whispers at the other end of the room.

5. News from Home

Allowing an interrogatee to receive carefully selected letters from home can contribute to effects desired by the interrogator. Allowing the source to write letters, especially if he can be led to believe that they will be smuggled out without the knowledge of the authorities, may produce information which is difficult to extract by direct questioning.

6. The Witness

If others have accused the interrogatee of spying for a hostile service or of other activity which he denies, there is a temptation to confront the recalcitrant source with his accuser or accusers. But a quick confrontation has two weaknesses: it is likely to intensify the stubbornness of denials, and it spoils the chance to use more subtle methods.

One of these is to place the interrogatee in an outer office and escort past him, and into the inner office, an accuser whom he knows personally or, in fact, any person -- even one who is friendly to the source and uncooperative with the interrogators -- who is believed to know something about whatever the interrogatee is concealing. It is also essential that the interrogatee know or suspect that the witness may be in possession of the incriminating information. The witness is whisked past the interrogatee; the two are not allowed to speak to each other. A guard and a stenographer remain in the outer office with the interrogatee. After about an hour the interrogator who has been questioning the interrogatee in past sessions opens the door and asks the stenographer to come in, with steno pad and pencils. After a time she re-emerges and types material from her pad, making several carbons. She pauses, points at the interrogatee, and asks the guard how his name is spelled. She may also ask the interrogatee directly for the proper spelling of a street, a prison, the name of a Communist intelligence officer, or any other factor closely linked to the activity of which he is accused. She takes her completed work into the inner office, comes back out, and telephones a request that someone come up to act as legal witness. Another man appears and enters the inner office. The person cast in the informer's role may have been let out a back door at the beginning of these proceedings; or if cooperative, he may continue his role. In either event, a couple of interrogators, with or without the "informer", now emerge from the inner office. In contrast to their earlier demeanor, they are now relaxed and smiling. The interrogator in charge says to the guard, "O.K., Tom, take him back. We don't need him any more." Even if the interrogatee now insists on telling his side of the story, he is told to relax, because the interrogator will get around to him tomorrow or the next day.

A session with the witness may be recorded. If the witness denounces the interrogatee there is no problem. If he does not, the interrogator makes an effort to draw him out about a hostile agent recently convicted in court or otherwise known to the witness. During the next interrogation session with the source, a part of the taped denunciation can be played back to him if necessary. Or the witnesses' remarks about the known spy, edited as necessary, can be so played back that the interrogatee is persuaded that he is the subject of the remarks.

Cooperative witnesses may be coached to exaggerate so that if a recording is played for the interrogatee or a confrontation is arranged, the source -- for example, a suspected courier -- finds the witness overstating his importance. The witness claims that the interrogatee is only incidentally a courier, that actually he is the head of an RIS kidnapping gang. The interrogator pretends amazement and says into the recorder, "I thought he was only a courier; and if he had told us the truth, I planned to let him go. But this is much more serious. On the basis of charges like these I'll have to hand him over to the local police for trial." On hearing these remarks, the interrogatee may confess the truth about the lesser guilt in order to avoid heavier punishment. If he continues to withhold, the interrogator may take his side by stating, "You know, I'm not at all convinced that so-and-so told a straight story. I feel, personally, that he was exaggerating a great deal. Wasn't he? What's the true story?"

7. Joint Suspects

If two or more interrogation sources are suspected of joint complicity in acts directed against U.S. security, they should be separated immediately. If time permits, it may be a good idea (depending upon the psychological assessment of both) to postpone interrogation for about a week. Any anxious inquiries from either can be met by a knowing grin and some such reply as, "We'll get to you in due time. There's no hurry now ." If documents, witnesses, or other sources yield information about interrogatee A, such remarks as "B says it was in Smolensk that you denounced so-and-so to the secret police. Is that right? Was it in 1937?" help to establish in A's mind the impression that B is talking.

If the interrogator is quite certain of the facts in the case but cannot secure an admission from either A or B, a written confession may be prepared and A's signature may be reproduced on it. (It is helpful if B can recognize A's signature, but not essential.) The confession contains the salient facts, but they are distorted; the confession shows that A is attempting to throw the entire responsibility upon B. Edited tape recordings which sound as though A had denounced B may also be used for the purpose, separately or in conjunction with the written "confession." If A is feeling a little ill or dispirited, he can also be led past a window or otherwise shown to B without creating a chance for conversation; B is likely to interpret A's hang-dog look as evidence of confession and denunciation. (It is important that in all such gambits, A be the weaker of the two, emotionally and psychologically.) B then reads (or hears) A's "confession." If B persists in withholding, the interrogator should dismiss him promptly, saying that A's signed confession is sufficient for the purpose and that it does not matter whether B corroborates it or not. At the following session with B, the interrogator selects some minor matter, not substantively damaging to B but nevertheless exaggerated, and says, "I'm not sure A was really fair to you here. Would you care to tell me your side of the story?" If B rises to this bait, the interrogator moves on to areas of greater significance.

The outer-and-inner office routine may also be employed. A, the weaker, is brought into the inner office, and the door is left slightly ajar or the transom open. B is later brought into the outer office by a guard and placed where he can hear, though not too clearly. The interrogator begins routine questioning of A, speaking rather softly and inducing A to follow suit. Another person in the inner office, acting by prearrangement, then quietly leads A out through another door. Any noises of departure are covered by the interrogator, who rattles the ash tray or moves a table or large chair. As soon as the second door is closed again and A is out of earshot, the interrogator resumes his questioning. His voice grows louder and angrier. He tells A to speak up, that he can hardly hear him. He grows abusive, reaches a climax, and then says, "Well, that's better. Why didn't you say so in the first place?" The rest of the monologue is designed to give B the impression that A has now started to tell the truth. Suddenly the interrogator pops his head through the doorway and is angry on seeing B and the guard. "You jerk!" he says to the guard, "What are you doing here?" He rides down the guard's mumbled attempt to explain the mistake, shouting, "Get him out of here! I'll take care of you later!"

When, in the judgment of the interrogator, B is fairly well convinced that A has broken down and told his story, the interrogator may elect to say to B, "Now that A has come clean with us, I'd like to let him go. But I hate to release one of you before the other; you ought to get out at the same time. A seems to be pretty angry with you -- feels that you got him into this jam. He might even go back to your Soviet case officer and say that you haven't returned because you agreed to stay here and work for us. Wouldn't it be better for you if I set you both free together? Wouldn't it be better to tell me your side of the story?"

8. Ivan Is a Dope

It may be useful to point out to a hostile agent that the cover story was ill-contrived, that the other service botched the job, that it is typical of the other service to ignore the welfare of its agents. The interrogator may personalize this pitch by explaining that he has been impressed by the agent's courage and intelligence. He sells the agent the idea that the interrogator, not his old service, represents a true friend, who understands him and will look after his welfare.

9. Joint Interrogators

The commonest of the joint interrogator techniques is the Mutt-and-Jeff routine: the brutal, angry, domineering type contrasted with the friendly, quiet type. This routine works best with women, teenagers, and timid men. If the interrogator who has done the bulk of the questioning up to this point has established a measure of rapport, he should play the friendly role. If rapport is absent, and especially if antagonism has developed, the principal interrogator may take the other part. The angry interrogator speaks loudly from the beginning; and unless the interrogatee clearly indicates that he is now ready to tell his story, the angry interrogator shouts down his answers and cuts him off. He thumps the table. The quiet interrogator should not watch the show unmoved but give subtle indications that he too is somewhat afraid of his colleague. The angry interrogator accuses the subject of other offenses, any offenses, especially those that are heinous or demeaning. He makes it plain that he personally considers the interrogatee the vilest person on earth. During the harangue the friendly, quiet interrogator breaks in to say, "Wait a minute, Jim. Take it easy." The angry interrogator shouts back, "Shut up! I'm handling this. I've broken crumb-bums before, and I'll break this one, wide open." He expresses his disgust by spitting on the floor or holding his nose or any gross gesture. Finally, red-faced and furious, he says, "I'm going to take a break, have a couple of stiff drinks. But I'll be back at two -- and you, you bum, you better be ready to talk." When the door slams behind him, the second interrogator tells the subject how sorry he is, how he hates to work with a man like that but has no choice, how if maybe brutes like that would keep quiet and give a man a fair chance to tell his side of the story, etc., etc.

An interrogator working alone can also use the Mutt-and-Jeff technique. After a number of tense and hostile sessions the interrogatee is ushered into a different or refurnished room with comfortable furniture, cigarettes, etc. The interrogator invites him to sit down and explains his regret that the source's former stubbornness forced the interrogator to use such tactics. Now everything will be different. The interrogator talks man-to-man. An American POW, debriefed on his interrogation by a hostile service that used this approach, has described the result: "Well, I went in and there was a man, an officer he was... -- he asked me to sit down and was very friendly.... It was very terrific. I, well, I almost felt like I had a friend sitting there. I had to stop every now and then and realize that this man wasn't a friend of mine.... I also felt as though I couldn't be rude to him.... It was much more difficult for me to -- well, I almost felt I had as much responsibility to talk to him and reason and justification as I have to talk to you right now."(18)

Another joint technique casts both interrogators in friendly roles. But whereas the interrogator in charge is sincere, the second interrogator's manner and voice convey the impression that he is merely pretending sympathy in order to trap the interrogated. He slips in a few trick questions of the "When-did-you-stop-beating-your-wife?" category. The interrogator in charge warns his colleague to desist. When he repeats the tactics, the interrogator in charge says, with a slight show of anger, "We're not here to trap people but to get at the truth. I suggest that you leave now. I'll handle this."

It is usually unproductive to cast both interrogators in hostile roles.


If the recalcitrant subject speaks more than one language, it is better to question him in the tongue with which he is least familiar as long as the purpose of interrogation is to obtain a confession. After the interrogatee admits hostile intent or activity, a switch to the better-known language will facilitate follow-up.

An abrupt switch of languages may trick a resistant source. If an interrogatee has withstood a barrage of questions in German or Korean, for example, a sudden shift to "Who is your case officer?" in Russian may trigger the answer before the source can stop himself.

An interrogator quite at home in the language being used may nevertheless elect to use an interpreter if the interrogatee does not know the language to be used between the interrogator and interpreter and also does not know that the interrogator knows his own tongue. The principal advantage here is that hearing everything twice helps the interrogator to note voice, expression, gestures, and other indicators more attentively. This gambit is obviously unsuitable for any form of rapid-fire questioning, and in any case it has the disadvantage of allowing the subject to pull himself together after each query. It should be used only with an interpreter who has been trained in the technique.

It is of basic importance that the interrogator not using an interpreter be adept in the language selected for use. If he is not, if slips of grammar or a strong accent mar his speech, the resistant source will usually feel fortified. Almost all people have been conditioned to relate verbal skill to intelligence, education, social status, etc. Errors or mispronunciations also permit the interrogatee to misunderstand or feign misunderstanding and thus gain time. He may also resort to polysyllabic obfuscations upon realizing the limitations of the interrogator's vocabulary.

<--- Previous page | Next page --->