The Lie Detectors (26 page)

Keeler found that 36 percent of the subjects harbored Nazi sympathies or, more to the point, considered it stressful to be questioned on this topic. Among the 126 recommended for the program, nearly 40 percent denied any affiliation with a political party and exhibited no reaction to questions about current loyalty; about the same percentage admitted membership in the Nazi Party but seemed to have no current loyalty to it; two admitted having been in the SS; and one reacted so as to indicate sympathy for communistic ideas. Of the 96 candidates rejected, half were admitted Nazis with signs of ongoing sympathy; a quarter denied any Nazi affiliation but reacted strongly about current sympathy; 15 percent reacted as loyal to communistic ideas; 10 percent were rejected as mentally unstable; and a few had bad hearts.

Those conducting the test judged it a success, not just for flushing out potential traitors, but for improving morale by enabling "those with a sincere desire to cooperate…to weed out the Nazis among their group." In other words, the lie detector fostered mutual trust by encouraging future cops to rat on their comrades. That this procedure did not entirely accord with the program’s ethos was noted by some rejected applicants. "We often were told to feel free and we were taught upon the rights of a free man too. I think therefore, that the school should give all students the opportunity of a defense if anybody states something against another student, instead of creating an atmosphere of mistrust." The school’s commander dismissed this particular prisoner as suffering from kleptomania, "a habit not desirable in a policeman."

Soon after the testing was complete, the army publicly acknowledged the reeducation program, and Keeler extolled his role to the press. But its achievements proved ephemeral. While only 6 percent of departing policemen still thought of Germans as the master race, 57 percent still blamed the war on the Jews. In the end, the police graduates were given no special role in Germany, in part because the occupation authority forbade any association among former veterans. On the other hand, this testing inspired some columnists to call for lie tests for all returning German prisoners.

For a self-styled pragmatic nation, America took an intensely ideological approach to sorting Germans. The Russians sorted their prisoners not by politics but by rank; they then dosed the soldiers with socialistic propaganda and "turned" the officers by playing to their opportunism. The British separated "whites" from "blacks" on the basis of conformism. Both the Russians and the British focused on group dynamics, whereas the Americans focused on individuals so as to identify and then transform their ideological allegiance. In a sense the lie detector had always been a loyalty monitor at one remove, probing the subjects’ conscience for their own belief that they had betrayed the law, their employer, or a spouse. The political struggles of the twentieth century simply reconnected the lie detector to its core competence as a touchstone, not of the truth, but of belief.

Keeler’s work screening German POWs landed him the biggest job of his career: a contract to test the workers in America’s premier nuclear city. On the surface, this may seem an unlikely alliance: Keeler’s humble placebo-ina-box protecting history’s most destructive device. But the lie detector and the bomb actually performed analogous tasks. After all, the bomb’s potency derived as much from its symbolic force as its destructive capability. Deterrence, the name for this new doctrine, was the very stuff of which the lie detector was made.

Safeguarding the bomb, after all, meant securing not just its physical matériel, but the idea too. To protect the fissile fuel, its defenders erected guard posts and radiation monitors. To protect the nuclear know-how, they deployed mental guard posts and psychical monitors. For the latter, they brought in Leonarde Keeler, ostensibly to prevent traitors from taking fissile uranium out, but actually to remind those with nuclear know-how that the knowledge they possessed was not theirs to give away.

Whereas Los Alamos was dedicated to the assembly of heterogeneous knowledge, Oak Ridge was a city of purification. Its gigantic industrial plants distilled the tiny percentage of fissile uranium 235 out of the many tons of inert uranium 238. To carry out this task required a heterogeneous array of actors. The federal government set aside thousands of acres of rural Tennessee; major chemical corporations invested vast sums; thousands of scientists, engineers, and technical workers were recruited from far-flung urban areas; and tens of thousands of semiskilled laborers were lured from the rural South. This agglomeration was strictly managed. The races were segregated; management and labor were sharply divided; unnecessary shoptalk was prohibited. During the war, the town was denied political representation, a free press, and labor unions. When the physicist Richard Feynman visited Oak Ridge from Los Alamos, he violated security protocol when he told production engineers what they were doing so that they could do it right.

The bombing of Japan and the formal disclosure of the existence of Oak Ridge intensified this obsession with atomic secrecy in some quarters, although it convinced others that nuclear know-how ought to be made transparent. The U.S. Army wanted to control atomic know-how, and its congressional allies demanded that the government protect America’s nuclear secret at all costs. The scientific elite and their allies in the Truman administration wanted to assert civilian control over atomic power; they warned that there were no atomic "secrets" as such: in five years, or fifteen at most, the Russians would master the atom. Given this inevitability, the United States should share its nuclear knowledge through the United Nations and control the proliferation of weaponry. Nativist Congressmen denounced these internationalists as naive and disloyal. Then rumors of leaks began. In January 1946 the press learned that the Russians had acquired nuclear information from Canadian spies. The House Un-American Activities Committee hinted that a comparable spy ring was operating at Oak Ridge.

To assuage these fears, the Army Corps of Engineers brought Keeler and his team to Oak Ridge in February 1946 to determine "insofar as possible, the loyalty, integrity, reliability, mental stability, and suitability" of its nuclear workers. Keeler unearthed no spy ring, although of the 690 individuals he tested—at $13 apiece—nine admitted having "stolen product material," and seven knew someone who had done so. In each case, however, the infraction proved minor: a tiny chip of uranium removed as a souvenir or a practical joke. More common (3 percent) were unreported radioactive spills. Far more common were the shenanigans that always surround job sites: 10 percent had lied on their job application, 12 percent had stolen tools, 3 percent had used an alias at some point in their life, and so on.

Soon after, the army signed a contract with Russell Chatham, one of Keeler’s team, to test workers at uranium separation plants run by the Carbide and Carbon Chemicals Corporation. Keeler did not bid on the contract; he would not have wanted to give up his high-flying image and move to Tennessee. In 1947, when the civilian Atomic Energy Commission (AEC) wrested control over nuclear power from the military, it expanded on Chatham’s testing program so as to assure the nation of its diligence. The chairman of the House Un-American Activities Committee had again generated huge headlines by announcing that Oak Ridge was "heavily infested" with communists, that nuclear secrets had been stolen, and that scientists had allied themselves with the workers’ union to protect traitors and keep the military out. Fighting public leaks of secret information with leaked announcements about security arrangements has since become standard procedure in American security politics, leaks being the pressure valves in the continual conflict between the public’s right to know and the demands of the security regime.

By the early 1950s, Chatham and his team of polygraphers were periodically testing 5,000 scientists, engineers, mid-level managers, and laborers at Oak Ridge—including one another. Employees were tested at hiring, at regular intervals thereafter, and on departure. Employees who objected were sent to a superintendent who threatened to revoke their security clearance, making the tests voluntary in name only. Only two employees dared refuse, and they were transferred to other jobs. Because of the sheer volume of tests—1,700 in June 1951—most exams lasted twenty minutes, with two sets of questions for each subject separated by an interval during which subjects could explain ambiguities and make confessions. This mass screening was designed not so much to uncover deceit as to enforce a new form of loyalty.

Only a small number of subjects disclosed "derogatory" information: 2 percent in 1952, of which one-third were tagged as having "friends or relatives associated with organizations considered un-American." Yet on closer examination, those who "sympathized with the Communist movement" were citizens who supported federally subsidized housing and the Tennessee Valley Authority. It was such "liberal thoughts" as these that caused one subject to wonder if "deep down he could be sympathetic toward Communism." In his case the polygraph operator indicated that the subject had probably confused his "isms." The majority of those affiliated with un-American organizations were patriotic members of the Ku Klux Klan. The managers of Carbide and Carbon Chemicals Corporation were far more pleased by the reduction in stolen tools and other items, especially Kleenex, that were in short supply during the postwar period. More generally, the tests helped keep the labor force quiescent during the CIO’s union drives, and damped down protests against hazardous releases of radiation and toxic mercury.

That the test uncovered so few "traitors" suggests that Chatham had set a relaxed or "friendly" threshold of suspicion. Polygraph operators have many ways of adjusting the trade-offs between false positives (wrongly accusing truth-tellers of lying) and false negatives (wrongly exonerating liars). So long as the percentage of spies was assumed to be small, setting a severe threshold for false negatives would hardly increase the odds of catching one, and would badly damage morale. This is an intrinsic feature of any probabilistic search for needles in a haystack. Suppose there are ten spies among 10,000 workers (surely a sufficiently paranoid guess). And suppose the polygraph has an accuracy of 90 percent (higher than any field study has found). Even if operators set out to trap half of the ten spies as deceivers, the test would net 345 "guilty" individuals, among whom the five spies would still have to be fingered, leaving 340 innocent workers wrongly accused and the five other spies entirely undisturbed. If interrogators decided to get even tougher and trap eight of the ten spies, they would need to accuse 1,606 individuals of lying, among whom interrogators would still have to pinpoint the eight true spies, while still letting two spies go scot-free.

An opinion poll conducted for Chatham confirms that the tests for the AEC aimed more at deterrence. To be sure, the fifty-nine responses are hardly to be taken at face value (the subjects’ names were listed). But they do suggest what security-minded employees thought security officers wanted to hear. Several subjects praised the lie detector as "a major deterrent." They said the test kept employees "on the ball," and made a person "think twice" before doing or saying anything to jeopardize security. One officer said that the tests cut "loose talk" by "70 percent," a number Chatham blithely quoted as a demonstrable fact. The test also gave employees confidence in one another. Said one engineer: "Passing the test makes me feel better personally and feel that the organization must be an honorable one." Several wanted the test to be required of all Oak Ridge employees and to be given at other nuclear sites as well. Then the same individuals praised the machine for clearing them of guilt in their own minds too. They thanked the test for giving them "the confidence that they have not done wrong." One physicist phrased his satisfaction in charmingly physicist-like terms: "I have a personal satisfaction in passing each test, to have a recalibration so to speak." In a world permeated by suspicion, with shifting presumptions about guilt, the lie detector provided "mental relief from worry" about whether one could count oneself among the honest.

By 1952 some 50,000 tests had been administered to more than 18,000 individuals, making Oak Ridge a sort of Tennessee Eden, purifying the world’s most potent product, and staffed by America’s most self-assured, trusting, honest technicians.

Why not extend this solace to a nation anxious about nuclear annihilation? The first Americans to whom this assurance was offered were those expected to be first to experience this destruction personally: American soldiers. Because General Douglas MacArthur was eager to use tactical nuclear weapons on the Korean peninsula, the army needed to convince Americans (and America’s foes) that U.S. troops had the stomach for nuclear war. At the tests code-named Desert Rock in the fall of 1951 American troops were put through a set of showy pseudo-secret maneuvers at a nuclear test site in Yucca Flats, Nevada. Their goal, according to army psychologists, was to ease the irrational fear of radiation that is "almost universal among the uninitiated." To assess (and ease) this fear, the paratroopers who carried out this first advance into a zone of nuclear devastation were subjected to before-and-after polygraph exams by Keeler’s associate Paul Trovillo. Trovillo accompanied the men through their education sessions on radiation and as they marched to within 500 yards of ground zero, where they found charred earth, blasted equipment, half-blind sheep, and not much else. His results compared the soldiers’ psychophysiological reactions to this experience (their "private" views) with their verbal reactions (their "public" views) and their observed behavior as recorded on film.