Authors: Ken Alder
Was Wilkens guilty? Could the lie detector be beaten? The old-style cops were furious that science had bungled the case. Larson heard rumors that the San Francisco police had vowed never to use the instrument again. The city’s captain of detectives stood on the podium at the next meeting of the International Association of Chiefs of Police and declared that lie detectors "would not be countenanced."
For his part, Larson blamed the press. He had been duped by the papers, he decided, and not for the first time. Already he had encountered the sort of unscrupulous editor who begged to sit in on an interrogation in return for a promise to keep the story under wraps—only to run the scoop in the morning. The newspapers fed all manner of misconceptions, stirring up wild fantasies of instantaneous justice. The Sunday supplements had started running stories about "How the Electric Detective Catches Criminals," invariably illustrated with "a pretty girl attached to an imposing array of apparatus." One crank had begun writing to Larson obsessively, contending that he too could read people’s minds. Larson made a vow: he would refuse to talk to the press about ongoing investigations, except when cases were already sensationalized. That way at least he would do no harm.
But there was still one question Larson couldn’t stop turning over in his mind: Was Wilkens guilty or not? It’s the question everyone always wants to answer: Whodunit? Larson also had his scientific results to clear up, not to mention his reputation. In his frustration, he tried an unusual approach; he befriended Wilkens. He congratulated him on his acquittal, helped him find a job in Oakland under an assumed name, and drove him around town while he went house-hunting. Then Larson persuaded Wilkens to sit for another round on the lie detector. After all, Wilkens could not be tried twice for the same crime. But the second results mirrored the first in their ambiguity, and again Wilkens insisted on his innocence.
Unsatisfied, Larson tried yet another tack. The lie detector was not the only new solution to the hoary problem of extracting the truth from recalcitrant witnesses. Since the publicity given to Larson’s exploits, Dr. R. E. House of Ferris, Texas, had announced his own method: "truth serum." This was the drug scopolamine, used by obstetricians to induce a "twilight sleep" during childbirth and ease the memory of pain. House now wished to promote it as a memory aide, coaxing drugged subjects to recall events honestly. One year after Wilkens’s acquittal, when House came to San Francisco to present his work to the American Medical Association and test his drug on prisoners at San Quentin, Larson asked him to try his serum on Wilkens.
Wilkens was game. "I am going to take every step to establish my honesty in the community." For several hours on the evening of June 25, 1923, a semi-somnolent Wilkens reiterated his previous claims of innocence while Larson, Vollmer, House, and several psychiatrists posed questions. But Larson was disappointed. Wilkens, he noted, had managed to tell lies under the influence of the drug, falsely denying, for instance, that he had had sex with his housekeeper. Worse, Larson decided, truth serum was unethical because suspects’ words were not under their conscious control, so they could not be held responsible even for words that were true. At a minimum, this form of unwilled self-incrimination could never be accepted in a courtroom.
By then, Larson was preparing to resume his training in criminal psychiatry. At the end of 1923, with Vollmer’s blessing, he was to leave Berkeley for Chicago to attend Rush Medical School and work part-time at the Institute for Juvenile Research, the nation’s leading center for behavioral approaches to crime prevention. There Larson would seek to transform his technique into a method for curing crime as well as detecting it.
But in the meantime, the Wilkens case had raised questions: Was it appropriate to introduce lie detector evidence in court? Should it be permitted if the suspect had agreed to take the test voluntarily? What sort of truth did the lie detector coax from the body? Larson addressed these questions when the American Bar Association met in San Francisco in 1922. He entertained the hope that its graphical results might one day be shown to juries, as they "are so striking that they could be easily recognized." However, these presentations would have to be made by qualified experts and then only after "careful standardization…, much cooperation and experimental work."
But there were others, more impatient. The publicity surrounding the Wilkens case had attracted the notice of William Moulton Marston, the Harvard-trained psychologist whose work had inspired Larson’s work in the first place. Marston still hoped to validate his own brand of lie detection, though he understood (as Larson did not) that the lie detector was not antithetical to publicity, but fed on it, even magnified it. Marston was in Washington, D.C., having used his method to exonerate an accused murderer, and he hoped to take the case all the way to the Supreme Court. He wrote to Larson in the hope that they might pool their efforts.
POLONIUS:
See you now,
Your bait of falsehood takes this carp of truth,
And thus do we of wisdom and of reach,
With windlasses and with assays of bias,
By indirections find directions out.
—WILLIAM SHAKESPEARE,
HAMLET
IN JULY 1922—ONE MONTH AFTER LARSON FIRST TESTED
Wilkens—a young African-American named James Alphonso Frye was visited in his jail cell in Washington, D.C. by William Moulton Marston, the Harvard-trained psychologist and lawyer who had inspired Larson’s research and who still had ambitions for his own lie detector. Frye, a twenty-five-year-old veteran of the Great War, had been arrested on August 27, 1921, for a car heist in collusion with four other black men. A few days later the police had induced him to confess to the unsolved murder of a prominent black physician, which had taken place almost a year earlier. Then a few days after that—to the consternation of his attorney—Frye had retracted his confession, having made it, he said, in exchange for a policeman’s promise to drop the robbery charge and for a half share of the $1,000 reward. Frye now claimed that he had agreed to this implausible deal because he knew he would never be convicted, thanks to an airtight alibi. But when the witnesses to his alibi refused to come forward, Frye’s new lawyers hunted for a way to counteract the confession. Coincidentally, Dr. William Marston, teaching at nearby American University, had run extensive lie detector tests on Negro suspects during the war.
Marston’s goal in testing Frye was to create a legal precedent, thereby realizing the ambition of his mentor, Hugo Münsterberg, to usher in a new era of scientific justice. In this he would succeed, though not as he intended. Frye’s case would effectively end the chance for the lie detector to have a career in the criminal courtroom, even as it gave the device free rein outside the formal procedures of the law.
The justice system has long wrestled with methods for gauging honesty. Criminal activity, almost by definition, cloaks itself in just the sort of falsehood society wishes to uncover. There have been three major phases in the development of such tests in the European west, each with its preferred methods of extracting the truth from recalcitrant human beings—and each increasingly in tune with prevailing assumptions about how to extract the truth from inanimate nature. Proponents of lie detection hoped to introduce a fourth and final phase in the administration of justice in which the methods for humans and inanimate nature would finally converge.
In the first phase, the medieval trial by ordeal, the community obliged the accused to confront a physical challenge so that God might judge the outcome. Sometimes this ordeal involved mortal combat. At other times, the faith of the accused was tested by an act of endurance. For instance, presumed liars were asked to lick a burning hot poker; if God wanted to commend their honesty, their tongues would not be scorched.
The second phase, which lasted from the twelfth to the eighteenth century, came in two distinct forms, one on the European continent, the other in England. Though the continental system of inquisitorial justice would be later denounced as cruel and unusual (see the Fifth and Eighth Amendments to the U.S. Constitution), it actually operated according to its own impeccable logic. The goal of inquisitorial justice was certainty of judgment—that, and the reintegration of the guilty into the kingdom of the saved. In an age when almost all crimes were punished by death, no Christian magistrate dared substitute earthly judgment for God’s say-so, at least not without proof "as clear as the sun at noon." Hence, the judge calculated guilt according to an arithmetic rule. The sworn testimony of one reliable eyewitness constituted a half proof; that of two independent eyewitnesses constituted a full proof. Unfortunately, as few crimes were witnessed by even one person, let alone two people, magistrates were authorized to torture the accused—confession being considered the "queen of proof."
In theory, judicial torture was justified to save the soul of the accused, lest the person die having violated a sacred oath of innocence and suffer the far greater torments of hell. It was also thought to be conducive to truth-telling. Early modern jurists believed that the truth was less a production of the will than a spontaneous utterance, and hence that bodily pain, by crushing the will, would release the truth, much as the pain of childbirth would induce an unmarried woman to honestly name her infant’s father. Yet the jurists also recognized what their Roman predecessors had long known: that many people resisted torture to the end, and that others, to end their agony, simply told interrogators whatever the authorities wanted to hear. For just these reasons the Roman jurist Ulpian was often quoted as saying that torture was "weak and dangerous, and inimical to the truth."
To mitigate against such deceptions, the jurists generally authorized torture only on the basis of circumstantial evidence of the sort we would consider sufficient to convict. Moreover, jurists introduced safeguards to ensure that torture was not abused. Examiners were forbidden to ask suggestive questions; the confessor had to supply corroborative information that only the guilty party could know; and the confessor had to repeat the confession after the torture had ceased. The torture itself was executed according to strict procedures. Judicial torture was not wanton sadism, but a formalized practice for obtaining the truth, one which matched the most up-to-date understanding of how to gain knowledge by putting nature "on the rack."
In England, by contrast, the common law actually retained many features of the trial by ordeal, except that parties contended in verbal rather than physical combat, with local jurymen posing questions and passing final judgment, while law judges, appointed by the king, served as neutral adjudicators. In the common law, torture was forbidden and no man was bound to accuse himself, although in practice the defendant did have to answer the charges or risk almost certain condemnation. Still, there were those in England who longed for the rigor of the continental system.
These currents clashed in the writing of Francis Bacon, often called the godfather of experimental science. Bacon was also the last Englishman to direct treason trials where torture was ordered by the king’s Privy Council. Bacon criticized the scholastics for merely touching nature "by the fingertips," and urged a more aggressive line of interrogation. As he put it, "the nature of things betrays itself more readily under the vexations of art than in its natural freedom." He urged investigators to bind nature in chains, driving matter "to extremities," causing it to "turn and transform itself into strange shapes" until—coming full circle—"it returned at last to itself." Bacon acknowledged that experimenters who forced nature in this way risked misconstruing its qualities. "When bodies are tormented by fire or other means, many qualities are communicated by the fire itself…whence strange fallacies have arisen." Yet he was confident that God’s benign superintendence would "return" nature to its true properties. Experimentation was less a form of torture than a struggle, from which both nature and experimenter emerged purified. Bacon’s natural philosophy found an uneasy echo in his jurisprudence. He condoned judicial torture, but only to discover corroborative evidence, not to extract confessions as evidence of guilt. As he acknowledged: "By the laws of England, no man is bound to accuse himself." The queen, he said, would not "make windows into men’s souls."
The resolution of this contradiction emerged later in the seventeenth century, when the third and current phase of judicial inquiry took shape. It was then that elite jurists and natural philosophers modestly agreed that justice could depend on judgments short of certainty. Henceforth, judges and adversarial lawyers were to probe the statements of witnesses in cross-examination, with judgment on their veracity—and the guilt of the accused—to be assessed in probabilistic terms. We can still hear the echo of this new approach to proof in such phrases as "beyond a reasonable doubt" and "a preponderance of evidence." And it was under this probabilistic banner, as much as under Enlightenment humanism, that the continental states gradually abandoned torture. Not coincidentally, natural philosophers at this time began to accept the notion that scientific knowledge of nature was provisional and probabilistic.
This grand accommodation, however, did not find favor with those who preferred swift and certain justice. Built into this new probabilistic justice were two contradictory trends, at least in the Anglo-American legal system. In a democratic vein, the courts increasingly granted lay jurors unfettered license to weigh the truthfulness of testimony. In an authoritarian vein, the courts increasingly put their confidence in expert-interpreted circumstantial evidence because it could not be dissembled. The ambition of advocates of the lie detector was to resolve this contradiction—in favor of the experts.
Only gradually over the past three centuries have Anglo-American judges allowed jurors to hear the defendant and witnesses for the defense testify under oath in criminal cases. Judges thought they had good reasons for this hesitation. Some worried jurors might be reluctant to convict defendants who swore oaths of innocence. Others feared defendants would readily perjure themselves, damning their souls and providing an excuse for further jeopardy. As late as the eighteenth century, certain sorts of people were thought so prone to lie on the stand that they were not allowed to take the oath: children, the spouse of the accused, those with a financial interest in the outcome, slaves, and the defendant. It was only after the American Civil War that U.S. courts finally allowed defendants to be sworn in when they took the stand in their own defense (and then mainly to avoid the accusation of hypocrisy as newly freed slaves were allowed to testify under oath). Only then, with all parties sworn in, did jurors have full freedom to weigh the testimony of all witnesses in an evenhanded manner, assessing their words and demeanor: the way they moved, blinked, blushed, sweated, or spoke. The institution of the jury may be 1,000 years old, but only in the past century have jurors acquired an unfettered license to distinguish truth-tellers from liars.
Yet over this same period, the justice system increasingly sought to corroborate human testimony with circumstantial evidence beyond the power of human beings to dissemble. Because lay juries and magistrates were themselves unable to assess such evidence, the courts increasingly turned to expert witnesses. The courts already drew on the testimony of skilled practitioners of specialized crafts—surveyors, physicians, trades workers—regarding technical matters of which they had direct knowledge. But only during the eighteenth century were adversarial lawyers in England permitted to call experts who extrapolated from special experimentation using general scientific principles to give their opinion regarding the case at hand. With the growing reach and authority of science in the nineteenth century, reformers hoped these experts would underwrite the reliability of a justice system otherwise in the hands of a lay jury.
The problem was that adversarial lawyers proved adept at finding adversarial experts to make diametrically opposed arguments. By the latter half of the nineteenth century, quarreling among experts in the courtroom had become a subject of general mockery. One legal aphorism of the period had it that there were three kinds of liars, "the common liar, the damned liar, and the scientific expert." It didn’t help that a vast array of practitioners of new sciences clamored for admission to the courtroom. How would the law decide whom to admit? The dream of certainty dies hard.
In the early decades of the twentieth century a coterie of reform-minded Americans set out to inaugurate a new phase in the administration of justice: the scientific interrogation of accused persons and other witnesses. The time had come, these reformers proclaimed, for the law’s archaic methods of assessing human honesty to give way to the new science of psychology. Just as the discovery of X-rays now allowed expert radiologists to peer into a patient’s body—and, according to some physicians, perhaps the mind as well—so might these new instruments allow expert psychologists to peer into a witness’s body to infer whether the conscience was disturbed. The human body, suggested these investigators, could serve as a kind of circumstantial evidence for the mind.
This challenge to the law first issued from the laboratory of Marston’s mentor, Hugo Münsterberg of Harvard. In the years before World War I, Münsterberg, a German émigré, founded "brass instrument" psychology. Lured to Harvard by William James, but despised by him in later years, Münsterberg was the first scientist to lay out the rationale for a science of lie detection. The seventeenth-century polymath Gottfried Leibniz had famously asked whether we would understand human consciousness any better if we could visit the machinery of our minds as if taking a factory tour. At the turn of the twentieth century Münsterberg created such a factory over Harvard Yard: a lab filled with students transforming their inner feelings into the kind of public phenomena we call science.
Münsterberg hoped to explore the deepest questions about the human mind. What was consciousness? What was emotion? What were honor, loyalty, and love? To learn the soul’s mute truth, he suggested, one had to listen to the body, precisely because it was not subject to the caprice of the human will or the vicissitudes of language. "We must bring man before a registering apparatus to find out…whether sunshine or cloudiness prevails in his mind." As one student at Radcliffe recalled: "[O]ur pulses strapped to his recording needles and cylinders, we registered irrevocably our susceptibility to patriotism, romance, horror, joy and a dozen other influences of daily life." Converted into graphical tracings—the new universal language of science—the qualities of subjective feelings could at last be accumulated, compared, and controlled.
In this, Münsterberg was pushing William James’s own theory of the emotions to its limits. The tone of our feelings, James proclaimed, derived not from our minds but from our bodies. As he put it, in one of those overly vivid examples that caused him so much grief, when we see a bear, we do not run because we are afraid; rather, it is our running that
makes
us afraid. Subtract bodily sensation from our feelings, James suggested, and no emotion-stuff remained.