The Lie Detectors (33 page)

In 1998 this contradictory landscape came to the attention of the U.S. Supreme Court in
Scheffer v. United States,
a case involving a military officer who had been denied the right to introduce polygraph evidence that seemed to exonerate him for failing a drug test. But in the end,
Scheffer
was decided on narrow grounds. Writing for an eight-to-one majority, Justice Clarence Thomas—perhaps recalling the imbroglio with Anita Hill—upheld the prerogative of the appropriate rule-maker (here the president in his capacity as commander in chief) to formally ban the polygraph so long as the rule-maker provided some minimal rationalization. Yet only three justices (forming a conservative minority) agreed with the remainder of Thomas’s opinion: that the polygraph threatened to usurp the jury’s role in making assessments ("the
jury
is the lie detector," quoted Thomas) or his fear that the machine’s "aura of infallibility" would overwhelm their deliberations. The four partially concurring justices and the lone dissenter (forming a liberal majority) were troubled by America’s double standard for the polygraph and defended the jury’s ability to assess all forms of evidence, including scientific evidence. Indeed, in their eagerness to assure the court that jurors would not be flummoxed by lie detector evidence, advocates of the polygraph had submitted briefs arguing that the results of polygraph tests now had a minimal influence on jurors’ judgments. It would be ironic, indeed, if the polygraph were finally admitted into court because most people no longer believed in it.

The muddled ruling left lower courts free to admit the polygraph, as some have cautiously begun to consider doing. In the courts, as elsewhere, the debate over the polygraph turns not so much on its reliability as on other questions. For conservatives, the debate turns on whether the machine is used in conjunction with institutional authority. For liberals, it turns on the right of the individual citizen to mount a full defense, especially when the accused seems likely to lose a swearing contest with the police. As always, with the polygraph what counts is less its reliability than the circumstances that surround it.

To those circumstances we must now add the resurgent national security regime and its "war on terror"—and the domestic posturing these campaigns enable. The CIA has long required its agents to take polygraph tests even as it trained its operatives on how to beat them. From prison, Aldridge Ames, the CIA agent who betrayed his country for twenty years—despite being repeatedly tested on the polygraph—now mocks the machine. "Like most junk science that just won’t die (graphology, astrology and homeopathy come to mind), because of the usefulness or profit their practitioners enjoy, the polygraph stays with us." After the FBI agent Robert Hanssen was also discovered to have betrayed his country for twenty years, the Bureau publicly announced that it would polygraph those agents who have access to intelligence information. Thus, to great fanfare, the polygraph lock was ceremonially affixed to the barn door.

In 1999, forty-six years after the AEC shut down the Oak Ridge screening program as unscientific and counterproductive, its successor agency, the Department of Energy (DOE), reinstated the ritual of political purification at the nuclear weapons labs. The impetus was again a politically motivated leak alleging that nuclear secrets had been leaked. That spring, with the Republican Congress excoriating the Clinton administration for supposedly kowtowing to China, the
New York Times
published a story based on anonymous sources suggesting that the Chinese had miniaturized nuclear warheads using a design possibly stolen from America. Behind the scenes, a panicked search for a culprit had already led security officers to fasten their suspicions on Wen Ho Lee, a Taiwanese-born American scientist who had worked at Los Alamos for twenty years and had been suspected of meeting furtively with Chinese scientists. One hitch was that a polygraph operator subcontracted by the DOE had judged Lee "not deceptive," and it was three weeks before the FBI reinterpreted Lee’s polygraph record as inconclusive and demanded a redo. This time the FBI’s examiner implied that Lee had done poorly, but there was no judgment until the article in the
Times
asserted that the government’s chief suspect was an unnamed Chinese-American weapons scientist who had taken a polygraph exam and was "found to be deceptive."

In a follow-up interrogation the next day, the FBI’s special agent Carol Covert accused Lee twenty-six times of having failed his polygraph. She also threatened him with the loss of his job, and with the same fate that befell the Rosenbergs. Two days later, the
Times
revealed that the DOE’s principal suspect was named Wen Ho Lee and that he had "stonewalled" investigators. When Lee was also alleged to have transferred to an open computer a library of software codes for simulating nuclear tests, he was placed in solitary confinement and shackled. Within a year, however, the government’s case against Lee had unraveled, and almost none of it has ever been substantiated. Among many other misrepresentations, the FBI agents had lied when they accused Lee of failing his polygraph exam. On the first exam he had received a high score for honesty. The second had been arranged solely to accuse him of failing. Many Americans were shocked to learn that such ruses were "not uncommon." In the end, the Justice Department had to withdraw the bulk of the case against Lee and release him.

Yet the DOE’s solution—its public-relations solution—was to make polygraph tests obligatory for lab personnel. This time around, the weapons scientists noisily expressed their disgust. At public hearings in Livermore, Albuquerque, Los Alamos, and Washington, D.C., they condemned the tests as unfair, demoralizing, lacking empirical validation, and as yet untested by the open debate on which science thrived. For purely political reasons, they said, they were being subjected to tests that "would never pass muster in a high school science fair." In 2003 a survey by the National Academy of Sciences of decades of lie detection studies criticized dragnet screening as useless, warning that it did no good in itself and undermined national security by producing a false sense of security. Later that year the DOE seems to have misread the report to justify an increase in the number of exams and forced out a physician-physicist who protested vehemently. For more than five years, some 20,000 employees have been subject to mandatory exams, a return to the situation of the early cold war—although in late 2006 DOE seemed to rescind polygraph screening and this flawed approach to national security.

September 11, 2001, which was supposed to change everything, has reprised every bugaboo concerning lie detection. The FBI has been using the lie detector on suspected terrorists with mixed results. When Abdallah Higazy, an Egyptian citizen who had been staying at a hotel across from the World Trade Center on 9/11, returned several months later to collect his possessions, he was held as a material witness. A security guard said he had found a pilot’s radio stashed in the safe of Higazy’s hotel room. In the course of a four-hour interrogation on the polygraph, the FBI’s operator screamed at Higazy, repeatedly told him he had failed the exam, and warned him that Egyptian security forces would investigate his family in Egypt (this being tantamount to threatening the family with torture). Sobbing with fear, Higazy admitted that the radio was his. He was placed in solitary confinement until the radio’s owner turned up to retrieve it a few days later. The security guard was later convicted of lying to the FBI, even though he too had previously passed a polygraph exam.

As usual, it is the innocent who are tripped up by polygraphs; the bad guys aren’t fazed. Documents seized by U.S. forces in Afghanistan show that jihadists have been trained in countermeasures against the lie detector, which they consider a perversion of the methods used by the medieval Islamic philosopher-physician Avicenna, who monitored the pulse to determine his patients’ true feelings. Iraqi jihadists have posted this information on the Web.

In the meantime, American interrogators have hastily reassembled the coercive techniques of the cold war. The road to Guantánamo and Abu Ghraib may have been laid in the month after September 11, 2001, when President George W. Bush issued secret orders exempting detainees in the war on terror from the Geneva Conventions and from the protection ofU.S. laws and international treaties against torture. But in charting this course the president was reviving long-honed techniques of psychological torture used by the CIA in Vietnam in the 1960s and its Latin American proxies in the 1970s and 1980s. As the historian Alfred McCoy has shown, nothing better dramatizes that legacy than the iconic photograph of the hooded Iraqi prisoner standing on a box, his arms outstretched, with fake electrical leads trailing from his fingers. Widely read as evidence that the interrogations at Abu Ghraib were the handiwork of a few perverted guards, the image actually documents the return of the CIA’s techniques of psychological torture, including sensory deprivation (the hood), self-inflicted pain (the outstretched arms), and the dire threat of physical pain (the fake electric leads).

In important respects, of course, polygraph techniques differ radically from these coercive methods. The polygraph can be used only on subjects who are ostensibly cooperative, and resistant subjects can always disrupt a test if so inclined. The physical discomfort experienced during a polygraph exam is nothing like the pain caused by forcing prisoners to assume stress positions. The anxiety of taking a polygraph exam is nothing like the terror of simulated drowning known as "water-boarding." Although the practice of lie detection may follow the
logic
of torture, it is not torture, not even psychological torture. To define torture "downward" until it encompasses such procedures would diminish our abhorrence for coercive techniques that cause acute physical or mental suffering.

Yet the similarities too are revealing. Although polygraph tests are nominally voluntary, prisoners may well have little choice but to take them. And once strapped down, few will refuse to answer all the polygrapher’s questions, even when examiners stray from agreed-upon topics. Even subjects who clam up are still supplying information involuntarily through their bodily responses. Both psychological torture and polygraph interrogation are predicated on total control over the subject’s environment. Keeler carefully designed his interrogation room as a kind of theater-for-one; the Darrow Photopolygraph monitored all the subject’s sensory inputs. Moreover, both sets of techniques deliberately rule out brute force lest it engender resistance or prompt false confessions. Like polygraphers, interrogators who use coercive techniques are told to remain cool, detached, and scientific. In both cases, the goal is to exert psychological pressure until a confession is forthcoming. It is telling that these new coercive methods were first developed in the 1950s under the cover of polygraph exams.

Finally, even though many polygraph operators and professional military interrogators have publicly denounced America’s use of torture, some polygraph subjects conflate their experience with psychological torture. And no wonder. For instance, the Pentagon has currently deployed large numbers of polygraph operators to Iraq to examine prisoners and informants. Exam conditions there are far from the scientific ideal, sometimes conducted with the aid of balky interpreters or in the bathrooms of bombed-out buildings. As one examiner acknowledged, "Some examinees were familiar with the American ‘truth machine’ while others were scared to death of having electrical wires attached to them."

Until recently, most Americans, when they gave any thought at all to the distasteful topic of torture, imagined that it involved the infliction of brutal, if metered, violence—as typified by the iron maidens of medieval dungeons. And most Americans—to judge by the popularity of television dramas like
24
and
NYPD Blue
—think even good cops sometimes have to break the rules and hurt bad guys because that is the only way to get the necessary information. Unfortunately, these false impressions have been reinforced by the ambiguous behavior of the U.S. government.

In the first place, according to the UN Convention of 1984, the definition of torture encompasses severe mental, as well as physical, suffering when "intentionally inflicted on a person for such purpose as obtaining from him or a third person information or a confession." Yet the United States, when it ratified the Convention in the 1990s, appended a set of little-noticed reservations that largely preserved the CIA’s clandestine psychological techniques. And since the attacks of September 11, 2001, the government has repeatedly denied that it has authorized torture, even while authorizing techniques such as sensory deprivation, stress positions, disorientation, and water-boarding that clearly constitute torture as defined by U.S. laws and treaties.

In the second place, there is reason to doubt that psychological torture generates information any more reliably than physical torture. As recently as September 2006, U.S. officials asserted that extraordinary techniques were the only way CIA interrogators had been able to pry crucial information from captured al-Qaeda figures like senior aide Abu Zubaydah, who confirmed that Khalid Shaikh Mohammed was the mastermind of the September 11 attacks. Yet these official claims were contradicted by reports which suggested that all the reliable information from Abu Zubaydah had been obtained in the initial, conventional interrogation by the FBI and that the crucial intelligence that led to the capture of Khalid Shaikh Mohammed was in fact obtained from a tipster who collected a $25 million reward. Other accounts have suggested that Abu Zubaydah is no longer psychologically stable.