The Lie Detectors: The History of an American Obsession (review)

A few years ago, a polygraph examiner in Chicago recounted to me how one subject had beaten the test. After the suspect "passed" (the technical term is NDI, "no deception indicated"), the examiner, who was convinced that he was guilty of the crime, asked him how he had managed to outwit the machine. The subject, apparently more proud of his ruse than concerned to maintain his innocence, opened his shirt and revealed that his torso was wrapped in aluminum foil. That should have no effect on the results at all, the examiner explained, but he surmised that the subject's utter confidence in his tactic enabled him to avoid the tell-tale bodily responses that indicate deception.

This little story demonstrates Ken Adler's point in this informative study that the lie detector works mainly as a placebo does: its effectiveness owes less to what it can do than to what people think it can do. Hence polygraph tests are primarily effective for eliciting confessions, which subjects often offer when they become convinced that the machine has caught them dead to rights. One problem is that some subjects come to credit the machine's powers above their own memories and make false confessions. The same polygraph examiner told me of a case where, during the test, he became convinced of a subject's innocence while she became convinced of her guilt, and it took him several hours of further testing and conversation before he managed to get her to withdraw her confession.

Adler has produced a meticulous history of lie detection, focusing especially on its birth and florescence between the 1920s and 1940s. It is told mainly through the stories of its central figures. Hovering above it all is August Vollmer, for many years chief of police in Berkeley, California and a ground breaker in bringing more humane and scientific methods of police work. John Larson and, slightly later, Leonarde Keeler, both Vollmer protégés in Berkeley who subsequently decamped to Chicago, were the pioneers in developing the polygraph machine and testing techniques. Their's is the tragic part of the story. After initial close collaboration they fell out. The methodical Larson wanted to develop lie detection as scientifically as possible. He recognized its limitations and was determined that it be applied only by the most carefully trained practitioners, with the exoneration of the innocent as an important function. He came to resent what he saw as the ambitious Keeler's commercialization of the process and its use primarily to extract confessions. Comic relief is provided by William Marston, who popularized lie detection in every way imaginable and loudly claimed success rates well over 90%. Marston was convinced that it was only a matter of time before the natural superiority of females would put them in control of society, and he developed the comic book character "Wonder Woman" as a foretaste of the female prowess he anticipated. He lived in a contented ménage à trois with his wife and assistant/mistress. He fathered two children by each of them, and they continued to live together after Marston's death.

Curiously for a study as fine-grained as this one, Adler describes very little about the polygraph machine itself—how it measures blood pressure, breathing, galvanic skin response (perspiration) and other bodily fluctuations—and how it is used in the testing procedure. He mentions, for example, the important "peak of tension" pattern more than once, but doesn't explain what it is. (With repetitions of the same questions in the same order, subjects' responses rise as the "hot" question—having to do directly with the matter under investigation—approaches, peak on that question, and then diminish with the irrelevant or "control" questions that follow it. The increasingly clear pattern is interpreted as evidence of guilt … although, of course, alternative explanations are not difficult to concoct.)

From an analytic perspective, the most interesting part of the book is the Epilogue, in which Adler tackles the question...