4
No Sanctuary

O Lord, thou has searched me and known me! thou discernest my thoughts from afar. Thou searchest out my path and my lying down, and are acquainted with all my ways. Even before a word is on my tongue, lo, O Lord, thou knowest it altogether. Thou dost beset me behind and before, and layest thy hand upon me. Search me, oh God, and know my heart! Try me and know my thoughts! And see if there be any wicked way in me, and lead me in the way everlasting!
Psalm 139

Of the two main theses this book aims to establish—that tests frequently create the personal characteristics they purport to measure and that testing is a means of exercising power—lie detection, a technique for surveillance and control, is obviously most relevant to the second. Using the largely descriptive account of chapter 3 as a foundation, here we investigate how lie detector tests exert power and domination over those who take them. The

analysis leads to the conclusion that lie detection represents an important advance in what Foucault has called the "disciplinary technology of power."^[1]

Communing/Communicating with the Polygraph

When administered with care by experienced and proficient interrogator-examiners, polygraph tests are potent instruments for bringing information to light which people wish to keep hidden. Their power to elicit confessions and revelations may be analyzed in terms of the framework of communication established by the test. A polygraph test replaces the normal pattern of dialogue between two persons with a matrix of communication involving four parties: the polygraph machine, the examiner, and the subject, who is bisected, Cartesian-like, into physical and mental substance, body and mind. The synapse of information transfer occurs between the two physical objects or machines: on the one side, the polygraph, and on the other, the subject as body, a bundle of physiological functions. Standing somewhat to the side are the two mental objects or persons: the polygraph examiner and the subject again, but this time as a volitional and moral agent.

The two machines commune together as the polygraph reaches out to embrace the subject's body with bands, tubes, and clips. The body responds loverlike to the touch, whispering secrets to the polygraph in tiny squeezes, twinges, thrills, and nudges. But both machines are treacherous. The body, seduced by the polygraph's embrace, thoughtlessly prattles the confidences it shares with the subject's mind. The polygraph, a false and uncaring confidant, publishes the secrets it has learned on a chart for all to read. The subject as mind, powerless to chaperone the affair, watches helplessly as the carnal entwining of machines produces its undoing.

Even as he sat in the polygraph examining room chair and watched Dee Wheeler secure the instrument's attachments to his body, he joked and laughed, confident he would leave the room a free man. But his smile faded when the instrument's slender recording pens

spelled out the killer's escape route and pinpointed the spot where he buried the murder weapon.^[2]

But when his brain and autonomic nervous system put deception criteria on the polygrams, Abrose broke down and confessed.^[3]

C. B. Hanscom bore down. His question formulations became exacting and precise. Aaron's autonomic nervous system, in a manner of speaking, blew its control fuse. Hanscom backed off, then began a delicate and deliberate interrogation. It wasn't long before Aaron had confessed not only to the burglary in question but also to the murder.^[4]

Advocates of the psychological stress evaluator (a machine that produces a sound wave chart from a tape recording of the subject's voice) point to the absence of physical connections as an advantage, but in fact the reverse is true. The physical grip of the polygraph contributes to the subject's sense of impotence before the machine, and this is one reason why the polygraph is more effective than the PSE and other lie detection devices in breaking through subject resistance to provoke confessions.

We will return to the subject as mind in a moment, but first we need to acknowledge the fourth figure in the communication matrix: the polygraph examiner. If this person is an incompetent amateur or an unscrupulous chart roller, the test does not approach its potential. The full power of the polygraph is realized only when manipulated by a determined and skilled examiner. Profiles of such individuals are uncommon; it is conceivable that Ferguson, who on occasion refers to them as "champions of justice," goes a little overboard with his:

Who and what may be termed a "competent polygraph examiner"? We might simply describe him as a rare breed, dedicated to the welfare of his country, confident of his ability, well trained and educated; he is an experienced interrogator with a profound knowledge of the psychology of human behavior, thoroughly versed in law and physiology, compassionate and understanding, fair and impartial, ethically oriented. He has been around. He has lived.^[5]

The effective examiner plays two distinct roles in a polygraph test. First he^[6] is the expert technician. He affixes the polygraph's sensors to various parts of the subject's body, calibrates the machine before the test begins in earnest, carefully instructs the subject to relax and to sit completely still while the machine is running, asks the questions in an even monotone, and interprets the charts. In this role, he umpires the tryst between the two machines; he is detached and objective, the consummate scientist. "The primary function of the polygraphist," explains examiner and polygraph theoretician Sylvestro F. Reali, "is to allow the subject to physiologically show the polygraphist the truth from within the subject himself. Allowing for no input to that end from the examiner."^[7]

The second role reveals the human side of the examiner. He seems genuinely to like the subject, to be pulling for the subject to "pass" the test. If deception is indicated on some of the questions, the examiner spares no effort to identify extraneous issues that might be responsible for it, and his greatest satisfaction is to prove the subject not guilty. He seems to understand the feeling of helplessness experienced by the subject in the face of a lie detector test and to sympathize. He wants to assist in any way he can. But the examiner is frustrated in these benevolent intentions unless the subject allows him to help. It is necessary that the subject place complete confidence in the examiner and cooperate fully so that, together, they might bring the test to a happy conclusion.

This, of course, is a role that examiners sometimes adopt treacherously, with the intention of cajoling the subject into a confession. But many examiners are sincere in this posture, and their most satisfying cases are those where they managed, after hard and creative effort, to clear an innocent individual suspected of a crime. One examiner told me of a young man he was called to test, who was accused of stealing stereo equipment from the store where he worked. On the first polygraph chart, the subject's responses to the relevant questions about the theft indicated deception. The polygraph examiner, however, was not convinced of the suspect's guilt. Following the first chart, he interviewed the subject at great length to ascertain whether some extraneous issue

might be causing the anxiety that appeared on the chart. Finally, taking the examiner into his confidence, the young man confessed that some ten years before he had been involved in a homosexual encounter. He was afraid that if his employer ever found out, it would cost him his job. The examiner assured him that this matter was irrelevant to the issue of the stolen equipment and that the secret was safe with him. This reassured the subject, and the examiner then turned on the polygraph and ran a second chart. No deception was indicated on the relevant questions about the theft, and the young man was cleared.

Polygraph examiners frequently report that "spillover" from anxiety about extraneous issues, as in this case, is a major cause of "false positives" (charts that indicate deception although the subject is actually innocent) in lie detector tests. Efforts are made to minimize this possibility by trying to put the subject at ease during a pretest interview, a process that normally includes reviewing all the questions that will be asked on the test. Often an "outside issue" question—one that asks if the subject is afraid that the examiner might probe into some issue that has not been discussed previously—is included on the test. The rationale is that a subject who is convinced that such outside issues will not be raised will be more relaxed during the test. Alternatively, as happened in the case of the homosexual encounter described above, the catharsis of confessing to an outside issue may relieve tension that could produce a false positive result. In either event, so the theory goes, anxiety stemming from extraneous matters is reduced and is therefore unlikely to produce charts that falsely indicate deception.

One method that helpful examiners use to avoid false positive results is through further inquiry into the control question (to which any subject is expected to respond deceptively and which is used as a point for comparison with responses to the relevant questions). If the initial chart indicates deception on a relevant question, Inbau and Reid recommend that the examiner pursue the source of deception by telling the subject that the charts indicate lies but that probably the deception is spilling over from the response to the control question (which in their system is, "Did you ever steal anything in your life?").^[8] The examiner goes on to

say that to clear the charts, it is necessary to be entirely honest on the control question. This is presented as a minor inconvenience to satisfy the peculiarities of the polygraph but of no real consequence because past thefts are not relevant to the present investigation. The subject is then invited to confess all past thefts, after which another chart is run, this time with the control question rephrased as, "Other than what you have already told me, have you stolen anything in your life?" This process may continue through several more charts, the examiner explaining that deception is still indicated and urging the subject to reveal any additional past wrongdoings. Ultimately, the subject is utterly purged of secrets about the past, the charts are clear of any indication of deception, and the polygraph examiner has fulfilled the professional obligation to preserve the innocent from false accusations.

A well-conducted polygraph test has an uncanny capacity to convince the subject that one's most private thoughts cannot be defended against intrusion. One is maneuvered into a position where the only option available seems to be confession. A young woman explained it this way:

You come in feeling like an honest, decent person and then he starts telling you that you're showing a reaction to some question about drugs or stealing and you try to think of a reason, things you've done, maybe years ago. You get to feeling as if the all-important thing is to get the machine to say that you're all right. You feel as if the thing can almost read your mind but not really accurately, it exaggerates, gets things mixed up. And you feel that if you tell the man everything you can think of, everything you've ever done that you might feel guilty about, why then finally it might come out with a clean record.^[9]

An examiner told me that one of his subjects marveled, "You know more about me than my husband does," and polygraph specialist John Reid claimed, "We get better results than a priest does."^[10] The lengths people go to bare their souls before the polygraph would be funny if so much pathos were not involved. Among employees polygraphed in a fast-food restaurant in Florida, one neatly wrote on the consent form, "I owe 99 cents for soft drinks.

I also owe 30 cents or more for small sundae," and appended her initials. Another confessed, "I have taken coffee and the inside of onion rings without paying," and a third divulged, that "Since I have been employed, I have given ice away with water."

Confessions, of course, are not always innocuous. It may be that the subject really is guilty of the malfeasance under investigation—a circumstance that the thorough polygraph test is, of course, designed to ascertain. And occasionally the extraneous information that is revealed is of much weightier import than the matter being investigated. One examiner told me that he has even gotten confessions of murder in routine preemployment screenings.

The impotence of the subject as mind in the lie detector situation sometimes impels people to lose confidence in their own convictions and to become abjectly dependent on the examiner. Then they may offer to confess to the delict under investigation even if they are innocent. Examiners I have interviewed report that it has sometimes been necessary to tell subjects, very carefully, not to confess unless they actually committed the crime. Working on the basis of a strong police presumption that a woman was guilty of a certain crime, an examiner succeeded in extracting a confession after about forty-five minutes of work with the polygraph. He then discovered, however, that she was unable to supply certain factual details about the case that the perpetrator would surely know. He concluded that she was in fact innocent, but it took him hours to convince her of that and to persuade her to retract the confession.

The experience of confessing before the lie detector may remain an uncomfortable memory for years. During a preemployment polygraph screening, a meat cutter I interviewed was asked if he had done anything for which he could be sent to jail. Although the examiner assured him he was only interested in the previous six or seven years, the urge to confess stimulated by the test resulted in an admission of something he had done some fifteen years before. He got the job, but his experience with the polygraph has rankled ever since. "They've got something that's mine," he told me. He worries that this information, which he considers to have been stolen from him, appears on his employment record.

Nor is such a concern necessarily misplaced. I asked several polygraph and PSE examiners what they do with confessions that are unrelated to the matter they have been charged to investigate. One said he never passes it on, but others answered that they write the information down and include it in their reports. Therefore, it is not always in the subject's interest to extend the trust that the accomplished polygraph examiner courts so ardently. It has been explained how police might encourage a suspect to submit to a polygraph test when their evidence is not strong enough for an indictment. The suspect is told that the test represents the best chance to clear oneself, but in reality, it is a last-ditch effort to obtain a confession, without which the case will be dropped. The police officer who explained this is convinced that when the department uses lie detector tests on its own personnel in internal investigations, the same game is being played. "They tell you they want to clear you of this," he said, "but really they're trying to hang it on you."

Asymmetry and the Power of Lie Detection

Our definition of a test includes the notion that it is an information-obtaining device applied by an agency to an individual. It is obvious that the individual to whom a lie detector test is applied enters the test situation alone and essentially defenseless. The test giver, in contrast, is an agent who comes to the test armed with technological apparatus and the official authority of the sponsoring organization. It is an utterly asymmetrical situation that pits the solitary subject against the full weight and resources of the organization. Even when the purpose of the test has to do with some possible benefit for the subject, such as screening for a job, the asymmetry of the situation is in the forefront. The test giver asks the questions, and the subject answers them, with full recognition by everyone concerned that the organization will evaluate the answers and use the resulting information to decide what to do about the subject. These observations apply to written integrity tests as well as polygraph tests. In fact, they apply to essentially all types of tests, and they highlight an important

aspect of the power that characterizes testing in general. Information still flows essentially one way from subject to test givers, that information is extracted and processed according to an agenda and technological apparatus controlled by the test givers, and it is still used by the test-giving organization to make decisions about the individual test taker. The sense of impotence may be exaggerated in the case of a polygraph test, when the subject is literally in the clutches of the machine's sensing apparatus, but takers of other kinds of tests are also held in the grasp—now metaphorical but no less meaningful—of test-giving organizations. As with preindustrial authenticity tests such as trial by ordeal, the asymmetry that places a scrutinizing organization on one side and a scrutinized individual on the other is a relation of power. If we are to understand the social consequences of lie detection, it is essential to know how that power works.

On occasion, it works so well that it literally creates what it purports to measure rather than identifying something that was previously there. I am referring to cases where people, disoriented by the circumstances of a polygraph test, come to distrust their own memories and confess to crimes they did not commit. Confessions can be created in the crucible of polygraph tests with much less expenditure of violence and energy than was visited on those unfortunates of four or five centuries ago who, utterly broken by torture, would come to believe that they were, indeed, witches. A particularly effective technique is for the polygraph examiner to run a chart and pin it to the wall. He then points to the response monitored for the relevant or "hot" question and asks why the subject "hit" on it. (In fact, the response may indicate deception only very slightly, if at all. But subjects normally have no knowledge of how to read polygraph charts and so have no way to interpret what they are shown apart from what the examiner says.) The examiner then leaves the room, allowing the subject to study the chart and stew about the response. The examiner returns and runs another chart, asking the same questions in the same order. The subject knows full well when the hot question is coming and is worried that the chart might again indicate deception on it, a worry that subsides after the question has been asked and answered. The new chart is pinned on the wall, and,

sure enough, the subject did "hit" on the relevant question again, even more this time, and a distinctive "peak of tension" pattern (responses building up over a few questions prior to the hot one, climaxing on it, and then relaxing on subsequent questions) starts to appear. The subject is again left alone to ponder the chart, after which the examiner returns and runs yet another chart—same questions, same order. The subject is increasingly nervous about the course that events are taking, and when this chart is pinned on the wall, the peak of tension pattern is unmistakable, even to untrained eyes. Confused and bewildered by a pattern of physiological responses that objective, scientific measures make obvious, ultimately the subject concludes that his or her own memory must be blotted out or playing tricks and confesses. What else could one do? After all, the subject is there alone, a single person relying on memory that, as everyone knows, can be fallible. How much can that count against the growing suspicion of a trained examiner and the scientific evidence of a precise machine, recorded so obviously that even the increasingly distraught subject can see it more clearly with each new chart?

The preceding scenario is likelier to occur in polygraph than in written integrity tests, primarily because the polygraph is used for meticulous interrogations in connection with particular crimes while written tests are used for brief and general checks of the honesty of job applicants who are not under suspicion for any specific misdeed. The following observations concerning the power of lie detection pertain, however, with at least equal force to written integrity tests as to the polygraph.

A Matter of Trust

One ingenious technique to identify a guilty party, reported from India, is to bring all those suspected of a crime together before a tent. They are told that inside is a magic donkey that will bray if a guilty individual pulls its tail but will make no sound if the one administering the tug is innocent. The suspects are then sent into the pitch dark tent one at a time and told to pull the donkey's tail. They are not told that the tail (which is actually attached to no

donkey at all) had been liberally smeared with lampblack. After the last one emerges from the tent, they are reassembled and ordered to show their hands. The one with clean hands is judged to be the guilty party.^[11]

Setting aside the obvious point that this technique could not be used very often because the word would quickly get around, consideration of it reveals one of the prime characteristics of the asymmetrically skewed logic of lie detection. This can be brought out by the question: why should it be expected that only one of the suspects would emerge with clean hands? Any reasonably prudent individual, innocent or guilty, would probably refrain from pulling the tail. After all, why should one trust an ass, no matter how reputedly magic or sage, not to make a mistake? If any suspect should in fact emerge with dirty hands, that person might be judged innocent by the investigators but would also have shown oneself to be a fool.^[12]

This vignette reveals a core assumption in the asymmetrical logic of lie detection: innocent or honest test takers are expected to behave completely ingenuously in the test, but the street is one way because the test itself is disingenuous. It is, in this case, a trick based on the lie that the object to be pulled is a genuine donkey's tail. Equivalents of a fake donkey's tail are found everywhere in modern lie detection. One is the interviewer who cajoles an applicant into destructive admissions by framing questions in the spirit of a friendly conspiracy, such as, "You know, Jack, when I have something better to do than go to work, I just call in sick. How about you?" Another is the police who tell a suspect that the case looks very bad and the only way to clear oneself is to take a polygraph test, when in fact the case is so flimsy that the polygraph is used in a final effort to get a confession before dropping the charges. Stanley Abrams, a lie detector advocate, proudly notes that polygraph examiner training schools require students to take a polygraph test to assure that they are of "good moral caliber"^[13] and, in the same book, urges these morally straight examiners to deceive subjects (if often for their own good) about the import of test results to make the test sufficiently "meaningful" to them that their responses will not be muted by nonchalance.^[14]

Asymmetrical expectations concerning truth are matched or exceeded by asymmetrical expectations about trust. While I have stressed that specific information in these tests flows in one direction only, from test taker to test giver, the very fact that a test is given conveys one unmistakable message in the opposite direction—that the organizations giving the tests do not trust the individuals who must take them. Takers of polygraph or integrity tests, in contrast, are expected to respond to test givers with complete trust and openness. Theoreticians of lie detector tests depict honest persons as peoplewho hold nothing back from public inspection; who, having done nothing wrong, have nothing to hide.^[15] They are trusting souls, fully convinced both of the efficacy of modern science (as represented by the polygraph or integrity tests) and of the essential goodwill of others, including (perhaps especially) the test givers.

This belief that the innocent have in the accuracy of the lie-detector, and that they will be exonerated, is usually shown by their attitude. This attitude is one of genuine confidence in both the machine and the examiner. Because of this confidence they regard the examination as an experience they will want to relate to their family and friends. . . . [They are] at ease, light-hearted, and talkative. However, they are very sincere and their straight-forwardness is displayed when they discuss the case during the interview.^[16]

Comments on a questionnaire administered to 220 individuals immediately after taking a routine polygraph test indicated that a number of them left the test with increased respect and confidence in the polygraph, and that they "found the test new, interesting, and enjoyable."^[17] The asymmetry is again painfully obvious: these depictions do not explain why honest applicants and employees should extend such confidence to the very persons and organizations that, by requiring a polygraph test, obviously have no confidence in them.

In the minds of test givers, people who do not manifest the total openness in the foregoing profile of the honest individual are suspect from the start. One polygraph examiner told me, quite vehemently, that the only people who complain about lie detector

tests are those who have something to hide and that the notion that polygraph tests represent an invasion of privacy is "a bunch of bull brought up by people who can't pass the polygraph." Interviewers are instructed to beware of certain "weasel" words that indicate dishonesty, such as an applicant evasively explaining the departure from a previous job in terms of "personal reasons," "personality conflict," or "I was laid off."

Experienced polygraph examiners often express confidence in their ability to identify guilty subjects quite apart from the test itself. Usually the telltale signs of guilt pertain to deviance from the above depiction of fully frank individuals who, with nothing to hide, compliantly welcome any and all intrusions into their private lives.

Muscular flexing, or a shifting or lifting of the cuff arm, during the early part of the test questioning, which is ordinarily accompanied by the subject's complaint that the cuff is hurting his arm, is usually indicative of deception. The innocent subject will almost always reserve any such movement or complaint until the latter part or end of the test.^[18]

Guilty subjects tend to be late for their test appointments, they appear quite nervous, they will not look the examiner in the eye, they frequently try to convince the examiner that they are deeply religious, they complain that the test is taking too long, and they want to get away as soon as possible, claiming to have another appointment; "when leaving they often quickly shake the examiner's hand and hurry out of the laboratory."^[19]

Warren Holmes, one of the most celebrated interrogators and polygraph examiners in the United States, explains some of the intricacies involved in distinguishing between the innocent and the guilty as follows:

The guilty tell their story in general terms without any specific details. . . . The innocent offer infinite details, and tell their story in such an animated way that you say to yourself, "Damn, nobody could make that up." When you listen to an innocent person, you can sense that he is reliving an actual experience. The innocent sound and look like they're telling the truth.^[20]

One wonders if it is quite that simple, however, especially on learning that the general/specific distinction is flatly contradicted by another polygraph expert. For Travis Patterson, it is the liar who refers to specifics, while "the innocent will be general in his remarks."^[21]

The notion that honest people gladly take lie detector tests is not easy to establish, because it runs counter to the widespread opinion that people—innocent as well as guilty—are fearful of such tests and do not wish to submit to them. Therefore, to convince both others and themselves, proponents of lie detection devote a good deal of imagination to advancing the proposition that the tests are voluntary. One simple expedient is to tell applicants that a polygraph or written integrity test is a prerequisite for the job and then ask if they are willing to take it. The overwhelming agreement by job applicants to do so is then presented as evidence for the voluntary nature of the test. A protest that applicants agree to take the test only because otherwise they will be dropped from consideration for the job is usually dismissed with the response that the individual is under no compulsion to apply for the job in the first place.

Ratiocinations regarding current employees are similar. My research revealed that if an officer in a metropolitan police department refuses to take a lie detector test in the course of an internal investigation, the chief of police issues a direct order to submit to the test. At that point, the officer who persists in refusing is terminated. It is still maintained that the polygraph test is voluntary, however, because the reason for termination is cited as insubordination to the chief's order, not refusal to take the test.

The zenith of the theory that honest people happily bare themselves to lie detection is represented by a few polygraph examiners and employers who have managed to convince themselves that employees, far from viewing lie detection as a weapon deployed against them, welcome it as promoting their own security and morale. One polygraph examiner explained it to me as follows. Workers in relatively large establishments often do not know each other personally. If the polygraph is used in preemployment screenings, that proves to people that their fellow employees are "decent people." A manager from Zale Corporation, a

large jewelry chain that used the polygraph extensively, said, "We explain very carefully to the employee that the best way we know how to make sure the persons working around him are honest, and not stealing, is through polygraphy. . . . Ten percent of our profits, before taxes, goes into an employee profit-sharing plan. It isn't hard to convince the employee that dishonesty costs him personally. Our employees don't want dishonest co-workers around."^[22] More than that, the polygraph is transformed into a bulwark of protection for the innocent employee. Each employee of Zale Corporation is told "that if he ever should be suspected or accused of any wrong doing, the polygraph is instantly available as a means of clearing himself."^[23]

By now we have disappeared completely into the magic wood, where fake donkey tails hang from every tree. The very instrument of the employer's abiding mistrust of employees is promoted as something not just to be tolerated by those against whom it is directed but actually to be welcomed by them as a boost to their morale. The machine—a literal objectification of mistrust—is held up to employees as a certification of their trust in each other. More, the polygraph is redefined as an instrument of solace for those unlucky enough to fall under the shadow of false suspicion, whether in employment or criminal circumstances. Lie detection becomes a helping profession, a fortress for the protection of the innocent. This ideal is enshrined in the Statement of Purpose of the American Polygraph Association:

Fortunate is he who, being accused or suspected of misconduct, is able to produce credible witnesses to attest to his innocence. Now therefore, and be it known henceforth, it shall be the primary responsibility of the American Polygraph Association to foster and to perpetuate an accurate, reliable and scientific means for the protection of the innocent. To verify the truth—fairly, impartially and objectively—shall be our purpose.^[24]

As a step toward putting this principle into practice, in 1960, the APA offered a polygraph test to anyone who claimed wrongful conviction of a capital crime.^[25]

Polygraph supporters recognize that these general claims about the positive attitude that honest and innocent subjects

adopt toward the test need to be bolstered with empirical evidence. To that end, a number of surveys have been conducted to ask people who have taken polygraph tests if they found them to invade their privacy or to be humiliating or offensive in any other way. Some of these demonstrate a certain concern for objectivity, but others belong to that part of the spectrum far beyond the point where credibility has been strained to the breaking point and snapped. A marvelous example of the latter is the survey conducted in 1978 by the Washoe County Sheriff's Department in Reno, Nevada. Deputy Sheriff and polygraph examiner Richard Putnam resolved to get at the truth of the common accusation that lie detector tests violate people's privacy by conducting a survey among 85 applicants for positions in the Sheriff's Department. The survey document is remarkable and deserves to be reproduced in full.^[26]

SURVEY

I, _____________, submitted myself to polygraph examination on _______________ at the Washoe County Sheriff's Department, after being advised of my constitutional rights and the fact that the examination was voluntary.

No promise or reward was made to me for answering the following questions:

1. Were you in any manner embarrassed, humiliated, or degraded by any part of the polygraph examination process?

Yes or No ___

2. In your opinion, was there any objectionable or unwarranted invasion of your privacy during the conduct of the polygraph examination?

Yes or No ___

3. Should you be hired, do you believe you will be more secure and comfortable in your work environment knowing that polygraph is used to assist in personnel evaluation?

Yes or No ___

I have answered these questions of my own free will and hereby authorize the release of my answers to these questions to any person or parties having an interest in them.

Signature of Person Examined: ______________________

Examiner's Signature: __________________________

Putnam reports with satisfaction that all 85 applicants completed the affidavit even though they were assured that participation in the survey was voluntary. The results were equally gratifying: subjects unanimously affirmed, on question 2, that their privacy had not been violated by the test; 83 of them answered yes to the third question, I answered no, and 1 had no opinion. A hint of trouble appeared when 6 applicants responded yes to the first question. But, Putnam hastens to assure us, "five of the six indicated their answer was based solely upon embarrassment caused by information they provided to the examiner which had not been directly solicited by the questions asked."^[27] This absolves the examiner from any suspicion of impropriety (while incidentally providing further evidence of the polygraph's capacity to elicit confessions) and cleans up the only place in the survey where the rave reviews of the polygraph drop below 97 percent.

The way in which the questionnaire was administered severely compromises Putnam's survey results. Because they want to be hired, job applicants are highly interested parties in the circumstances of this research. Their primary consideration in completing a questionnaire is not to say what they really think but to say what they think the employer wants to hear and, above all, to avoid saying anything that might dispose the employer against them. Therefore, if research based on a questionnaire is to be credible in these circumstances, it is essential that subjects be convinced that it is impossible to identify who completed what questionnaires. In this case, the form's official-looking stipulations, the required signatures of both subject and examiner, and the authorization of release of the answers to any interested party reduce the whole "study" to a travesty of research. The only surprise is that results were not 100 percent on all three questions.

A survey that initially appears to have been more sensitive to methodological considerations was conducted in 1980 by Toronto polygraph examiner Ben A. Silverberg. He collected questionnaires from 102 job applicants and 118 current employees immediately after they had taken routine polygraph tests. To protect anonymity, subjects were left alone to complete their questionnaires, which they then deposited in a sealed container. The polygraph again passed with flying colors. Responses to ques-

tions about fairness of the test, invasion of privacy, willingness to take it on a regular basis, and so on, ranged from 89 to 100 percent favorable to the test.^[28]

Silverberg's is one of several surveys, all of which show similar results. As Frank Horvath generalized the results of seven surveys of attitudes of those who have taken preemployment polygraph tests,

between 85 and 95 percent of the persons who have experienced polygraph examinations said that their examinations were not offensive, objectionable, or an invasion of their privacy. In addition, the overwhelming majority of these people expressed a willingness to take polygraph tests in the future should the situation require it.^[29]

A methodological difficulty with these studies, however, is that subjects were asked to complete questionnaires immediately after taking the polygraph test, while still in the examiner's office, and, therefore, before the result of their job application was known. It is possible that they felt their anonymity might not be respected and that their responses might jeopardize their chances of being hired.^[30] Further consideration of the Silverberg study suggests that its strategy of leaving subjects alone to complete their questionnaires and drop them in a sealed container was not sufficient to assure anonymity. For one thing, respondents might have realized that their questionnaires could be removed from the container immediately after they left and thus easily linked to them. Further, "each questionnaire instructed the subject to complete a short biographical sketch."^[31] It would not be difficult to connect the biographical sketches with information on application forms or employment records to identify respondents to particular questionnaires. Subjects who were concerned about these possibilities had good reason to be, because Silverberg did in fact subvert the ostensible anonymity of the study. One of his tables breaks down the responses to the eight questions according to the 185 subjects who were recommended for hiring or retention, the 14 who were given a qualified recommendation, and the 21 who were not recommended.^[32] How could he have done that unless he knew who filled out what questionnaires?

To avoid some of these problems, Frank Horvath and Richard Phannenstill conducted a survey in which questionnaires were mailed to respondents well after hiring decisions had been made. Questionnaires were mailed to 596 individuals who had taken preemployment polygraph tests in Milwaukee between 1983 and 1985, and 218 responses were received. Responses were generally positive toward polygraph tests, although less enthusiastically than in the other studies: 72 percent found the test they had taken to be fair, 82 percent said the questions were not objectionable, and 73 percent would take a polygraph test in connection with another job application.^[33] A methodological problem may also be detected in this study, however, because "although no identifying information was requested from respondents, each questionnaire was conspicuously identified with a code number."^[34] This enabled the researchers to correlate questionnaire responses about attitudes toward the test with information such as admissions of wrongdoing that the individual might have made during the test and whether or not test results warranted recommendation for employment. Research procedures were established which prevented identification of the respondents. Nevertheless, the conspicuous code number on each questionnaire may well have aroused their suspicion that the researchers intended to connect returned questionnaires with the names of individuals to whom they had been sent, and that could have had an effect on the answers.

On the other side of the issue, Lawrence White's psychological experiments and interviews suggest that applicants respond to preemployment polygraph screenings with lowered self-esteem, diminished satisfaction with the job, and increased inclination to steal from the employer.^[35] This indicates that lie detection might actually exacerbate the very conditions it aims to curtail. I conducted extended interviews with eight individuals who have taken polygraph tests as preemployment, periodic, or specific investigatory tests. Although my sample is very small, the results correspond much more closely to White's findings than to the studies that report positive attitudes toward polygraph tests on the part of those who take them. Nearly all of my subjects expressed resentment that the sheer circumstances of the test

placed them under suspicion of wrongdoing. They judged the test to be embarrassing or uncomfortable and an invasion of their privacy. The test was a particularly unpleasant experience for three of my subjects. One of them, who worked for a janitorial service, was tested under suspicion of theft by what seems to have been an unscrupulous chart roller, who even demanded that she pay for the test herself. She lost her job as a result of the test. Another subject was tested in connection with missing money at the retail store where she worked. She was so humiliated by being suspected and required to take the test that she intended to quit her job as soon as the actual culprit was identified. (She planned to remain until then because she did not want to give the impression that she was quitting because she was guilty.) The third individual was subjected to questions pertaining to her sexual activity and induced to admit to lesbian encounters, although this issue was irrelevant to the ostensible purpose of the test. Nevertheless, this woman took the position that, for all their problems, the tests are still warranted because they contribute to controlling serious social problems such as employee theft and drug use.

Lie Detection and the Disciplinary Technology of Power

A common reaction to lie detector tests is to despise them as unwarranted and humiliating invasions of privacy that, particularly in light of their questionable accuracy, may work a great deal of harm on helpless and innocent people. Certainly, the discussion of the asymmetry of power and trust entailed by lie detection is conducive to a conclusion of this sort. But to reach that conclusion and leave the matter there does not generate a satisfying analysis of lie detection. It does not explain, for example, why some people (certainly fewer than the figures suggested by polygraph proponents but probably a significant number nonetheless) quite willingly submit themselves to lie detector tests. Nor does it encourage exploration of what lie detection can reveal about the nature of contemporary society. A more instruc-

tive path is one that delays judgment of lie detection and investigates further the asymmetry that lie detection establishes between potent test givers and impotent test takers, asking what kind of social system it is that succeeds in placing people in such a vulnerable position vis-à-vis the organizations that demand to test them.

The genius of the polygraph test is its promised road to exoneration, which, however, frequently requires the subject to reveal much extraneous information along the way. The examiner may tell the subject that deception is indicated on a relevant question but then suggests that the reason is probably knowledge of the wrongdoing of someone else or spillover from anxiety about some unrelated issue. To clear oneself of suspicion, it then becomes necessary to locate the cause of the problem. Captured by the imperative to get the charts to look right, one may end up pouring out every private, embarrassing, or incriminating fact that memory can locate about oneself or others, until the subject is utterly drained and the chart, finally, runs clear, purged of all indications of deception. And all of this happens because it is time for a periodic test, or the subject happens to work in a department where a crime may have been committed, or is, in all innocence, simply applying for a job.

The necessity to divulge extraneous information is, polygraph examiners insist, just a by-product of the peculiar technology of polygraph tests. It is instructive, however, to imagine that it is not an epiphenomenon at all but the point of the whole process. Then investigations of wrongdoing, periodic checks on employees, and the screening of job applicants emerge as pretexts for getting people into a situation where they are forced to reveal a wide range of potentially damaging information about themselves. From this perspective, lie detection expands immensely in significance to become a technique for maintaining surveillance over people's departures from social norms for some quite general purpose. But what might that purpose be?

A theoretical context in terms of which this expanded view of lie detection makes sense has been developed by social historian and philosopher Michel Foucault. His book, Discipline and Punish ,^[36] is an analysis of the evolution of power in Western

civilization. Until about the eighteenth century, power was applied by means of a veritable orgy of violence, examples of which we have encountered in the ordeal and trial by battle. Foucault describes how the bodies of those found guilty of defying the king's justice and power were burned, broken, crushed, cut, or torn asunder in public spectacles of torture and execution designed to terrorize all who might contemplate challenging the established order. In the end, it was, however, an inefficient strategy. Power was applied in sporadic explosions of excess rather than in regular and measured doses. The agony of the victim might attract the crowd's sympathy rather than focusing their attention on the intended lesson that if they did not mind their own behavior, they could well come to a similarly sticky end. If the executioners bungled their job, the ostensibly omnipotent and terrible justice of the king would look foolish and the people might be tempted to confront the king's violence with rebellious violence of their own.^[37]

For a brief period at the end of the eighteenth century and the beginning of the nineteenth, the principles of the Enlightenment were applied to the exercise of power and punishment when efforts were made to base conformity to social rules on people's rationality. Criminals had broken the social contract, placing themselves beyond the bounds of society. Punishments should dramatize their status as social exiles and vary appropriately to the specific crimes by which they had defied the social order. Prisons would be open to the public, for the edification of schoolchildren and others, who would observe and learn the inexorably rational connection between breaking the law and paying for it.^[38] The appeal to reason seems not to have been an effective means of social control, however, and it was not long before the application of power took yet another form. It shifted toward the development of subjects who were disciplined to embody—automatically and unthinkingly—all the proper habits for following authority. This is the disciplinary technology of power, which Foucault describes in terms of the military dream of society:

Historians of ideas usually attribute the dream of a perfect society to the philosophers and jurists of the eighteenth century; but there

was also a military dream of society; its fundamental reference was not to the state of nature, but to the meticulously subordinated cogs of a machine, not to the primal social contract, but to permanent coercions, not to fundamental rights, but to indefinitely progressive forms of training, not to the general will but to automatic docility.^[39]

Such a system is especially well suited to the special requirements of industrial and postindustrial society, where individuals must be trained to perform highly specialized tasks and where they must perform them regularly and reliably so as to maintain the proper coordination of the mutually dependent components of the economy and other parts of society.

The disciplinary technology of power requires that people be kept under regular surveillance, in order that deviations from proper behavior might be rapidly detected and immediately redressed. This allows corrective power to be applied sparingly, for small doses are sufficient to curtail small strayings. One instrument designed to facilitate surveillance was Jeremy Bentham's panopticon.^[40] Here was an arrangement whereby inmates could be isolated from each other in individual cells and conveniently observed from a central tower. A peculiar advantage of the panopticon in terms of efficient utilization of resources was that as the inmates could not see into the observation tower, an observer did not actually have to be present at all times. Never knowing when they were being observed, inmates would have to behave as if the surveillance were constant. Although Bentham's architectural device has been used primarily in the construction of prisons, the full title of his book makes it clear that he by no means intended it to be limited to penal institutions: "Panopticon; or, the Inspection-House, Containing the Idea of a New Principle of Construction Applicable to any Sort of Establishment, in Which Persons of any Description are to be Kept Under Inspection: and in Particular to Penitentiary-Houses, Prisons, Houses of Industry, Work-Houses, Poor-Houses, Manufactories, Mad-Houses, Lazarettos, Hospitals, and Schools . . ."

A different technique of surveillance was tried for a time by the Ford Motor Company when Henry Ford, concerned by the effect

of high turnover rates on productivity, resolved to attract and keep good workers by making them an offer they could not refuse. In 1914, when the going wage for unskilled labor was $2.34 per day, Ford began paying his laborers the unheard-of sum of $5 per day. Ford's primary conditions were that workers justify this munificent wage by "proper living" and that their new affluence neither increase their cupidity nor make them slothful. A "Sociological Department" was formed at Ford to maintain surveillance over employee behavior off the job as well as on it. Its investigators (at the peak of the program, there were 100 of them) would visit employees' homes unannounced to certify that they did not engage in excessive drinking or improper sexual behavior, that no boarders were taken in, and that their houses were kept neat and clean. Employees whose comportment fell short of company standards dropped back to the $2.34 per day wage for a probationary period, but if they mended their ways, they could be reinstated to the $5 day.

Ford's scheme, one of the most dramatic efforts ever adopted by an American company to intervene in the overall lives of its employees, was short-lived. The Sociological Department's surveillance was unpopular, and Ford himself came to hold the opinion that it was better to reward employees for productivity on the job and length of service than for proper living at home. The cost of the program may have contributed to Ford's second thoughts about it. Salaries for the one hundred investigators and other expenses connected with the Sociological Department must have amounted to a considerable sum. And to maintain the premium wage at double the standard wage would require raising the former at twice the rate of the latter, a practice that could become very expensive indeed. As it happened, the daily $5 was not adjusted to keep pace with inflation, so in the years following 1914, it looked less and less extraordinary. Despite its spectacular beginning, the experiment of luring people to hard work and proper living with the inducement of uncommonly high wages faded away after only six years.^[41]

If we look at lie detection from the perspective of Foucault's analysis, it may be readily understood as yet another device for surveillance, one that represents a major advance in convenience

and economy over both the panopticon and Ford's $5 day. The leap forward achieved by lie detection by polygraph is that it has succeeded in enlisting people as their own watchmen. The central power need simply collect occasional reports of self-surveillance by means of lie detector tests. As the personnel director of the Zale Corporation remarked, "We feel the best place to go to find out the background of a person is to the person himself, and the polygraph is the best way to do this."^[42]

The information gathered in this manner should be unrivaled in detail and completeness. Far beyond what can be observed from a central tower or gleaned in occasional visits by company investigators, lie detection achieves the ultimate in surveillance because it taps the constant and undivided attention that constitutes self-awareness—an awareness not only of overt deeds but also of intentions, desires, impulses, and other "inner" phenomena that, but for the test's probing, may never become public. Moreover, lie detector tests operate to control behavior as well as to report on it. The strategic assumption, at any rate, is that people will be deterred from disapproved activities (such as stealing from one's employer) by fear that it will be found out in the next test.

Polygraph tests also represent an advance in the efficiency with which power is applied. Conducted in private, testing avoids the massive expenditure of energy and resources required for public spectacles of torture and execution in the monarchial system of power. Polygraph tests also achieve surveillance over more people more cheaply and conveniently than is possible by means of primitive instruments of a disciplinary technology of power such as the panopticon or Ford's $5 day. Scarcely an hour in duration and requiring only a portable machine and a quiet room, the tests obviate the architectural, construction, and maintenance costs connected with the panopticon, as well as the expense and inconvenience of assembling and confining people for extended periods within its precincts. Ford tried to control employees' behavior with doubled wages and surprise visits by members of the Sociological Department, strategies that required a major financial outlay by the company. Lie detection controls behavior by threatening people with losing the wages they have, and that costs

the company very little. Employers seldom go to the trouble and expense of prosecuting those identified by lie detector tests as deceptive or dishonest. Applicants who test deceptive in preemployment tests are not hired; current employees in that unhappy situation may be terminated. This necessitates no legal apparatus to investigate and try cases, nor prisons to hold the offenders. Undesirables are simply excluded from employment. If the measure of efficiency in the application of power is maximum impact at minimum cost, the ratio achieved by lie detection is overwhelming. A procedure requiring small outlay by test givers places test takers at the large risk of not getting, or losing, a job. To the extent that numerous employers apply lie detection, and results of different tests hold constant, some individuals will suffer the cumulative and devastating consequences of losing their ability to earn a livelihood.

The Employee Polygraph Protection Act of 1988 may be thought to have stymied the perfection of this aspect of the disciplinary technology of power when it outlawed mechanical lie detector tests for most purposes in the private sector. Perhaps so, but the antipolygraph law is stimulating phenomenal growth of written integrity tests, and these may prove to be as effective tools of the disciplinary technology of power as the polygraph ever was. To be sure, written tests cannot equal the polygraph's capacity to elicit confessions of past misdeeds of any and all descriptions, and thus they cannot achieve the same meticulous degree of individual surveillance. But since they do not require one-on-one administration, they are considerably cheaper than polgyraph tests, and they may conveniently be given to a much larger proportion of the population. Therefore, it may be that the disciplinary technology of power has not in fact retreated before the antipolygraph law but has made a lateral shift preparatory to further growth. What individual surveillance may lose in depth of fine-grained detail it will gain in breadth of application.

Foucault also suggests that the disciplinary technology of power reordered the concept of the individual. Earlier periods of history were marked by what might be termed "ascending individualization": the persons who were known as distinctive individuals were those at the center of power, and/or those who were

exalted as heroes or vilified as execrable villains. Information about these exemplary figures flowed to the anonymous beings who composed the masses. There the stories about famous central figures inspired awe (thus helping to maintain the rulers in power) and provided role models for desirable unacceptable behavior. Foucault continues:

In a disciplinary regime, on the other hand, individualization is "descending": as power becomes more anonymous and more functional, those on whom it is exercised tend to be more strongly individualized; it is exercised by surveillance rather than ceremonies, by observation rather than commemorative accounts.^[43]

That is to say, information no longer passes from focal figures to anonymous masses but in the reverse direction: the people who are controlled are known, while those who hold power are anonymous. The asymmetrical, one-way flow of information from those who take lie detector tests to those who give them is an outstanding example of Foucault's descending individualization. But, of course, the examiners who give the tests are not themselves in power. They are only agents of the real power centers. And those centers are so depersonalized that they are likelier to be organizations than persons—corporations, the government, and, ultimately, the total sociocultural system.

The efficient communication of specific information about named and identifiable individuals to anonymous centers of power is accomplished not just by lie detection but by testing of all sorts. An outstanding and very recent addition to the rapidly accumulating set of examples in the contemporary United States is DNA testing or "genetic fingerprinting" of prison inmates. Given a recidivism rate as high as 80 to 90 percent for murder and sex-crime parolees, current inmates are tested and their genetic profiles stored so that they may be readily identified from any body tissue, hair, blood, semen, or saliva that they might leave at the scene of some future crime. By the end of 1991, fifteen states had laws mandating DNA testing of inmates convicted for certain kinds of offenses. Leaders in toughness are Virginia, which tests all felons and which had deposited some 30,000 samples in its

data bank in a period of about eighteen months, and South Dakota, where just being arrested leads to DNA testing.^[44] With the accelerating development of new technologies and new tests derived from them, Foucault's insights about testing gain in relevance and importance with the passage of time: "The examination is the technique by which power, instead of emitting the signs of its potency, instead of imposing its mark on its subjects, holds them in a mechanism of objectification. . . . We are entering the age of the infinite examination and of compulsory objectification."^[45]

There is more. In addition to displaying and fixing the individual as an object of knowledge for the purposes of subjugation, Foucault suggests that the role of testing in bringing about a shift from ascending to descending individualization has contributed to a fundamental redefinition of what the individual is:

Thanks to the whole apparatus of writing [and now we would add, most emphatically, computer data banks] that accompanied it, the examination opened up . . . the constitution of the individual as a describable, analysable object, not in order to reduce him to "specific" features [those characteristic of the species generally], as did the naturalists in relation to living beings, but in order to maintain him in his individual features, in his particular evolution, in his own aptitudes or abilities, under the gaze of a permanent corpus of knowledge.^[46]

Far from being the primal, irreducible, and invariable atom of human society, then, this argument holds that the individual as currently known and conceptualized is a relatively recent invention, an invention generated in large measure by the procedures of testing and serving the disciplinary technology of power.^[47] We may not wish to follow Foucault's dark genius to the point of agreeing that the individual was created just in order to be controlled and dominated. Nevertheless, he leads us to the compelling if paradoxical conclusion that the concept of the individual, on which is erected our civilization's particular construction of human freedom and dignity, is itself partly built on testing—perhaps the most pervasive and efficient technique for the appli-

cation of power and domination to have evolved so far. Lie detection has served as a lens that reveals certain features of testing with great clarity by magnifying or distorting them to the point of caricature. In the following chapters, many of those same characteristics will surface again in other sorts of testing. Their further analysis, and that of numerous other features of testing yet to be analyzed, will afford a better view of the new kind of social order that has evolved over the last two centuries or so and the new kind of individual that has developed with it.