The First, Flawed Machine

The quest for such a machine has roots in the early 20th century, when the first modern lie detector, a rudimentary polygraph, was introduced. The man often cited as its inventor, William Moulton Marston, was a Harvard-trained psychologist who went on to make his mark as the creator of the comic-book character Wonder Woman. Not coincidentally, one of Wonder Woman's most potent weapons was her Magic Lasso, which made it impossible for anyone in its grip to tell a lie.

Marston spent 20 years trying to get his machine used by the military, in courts and even in advertising. After the success of Wonder Woman, however, he used it mostly for entertainment. His comic-book editor, Sheldon Mayer, recalled being hooked up to a polygraph during a party at Marston's home. After a few warm-up questions, Marston tossed him a zinger, "Do you think you're the greatest cartoonist in the world?"

As Mayer wrote in his memoir, "I felt I was being quite truthful when I said no, and it turned out I was lying!" What an interesting reaction — even if, as was likely, Mayer was just trying to be funny. Because how prescient, really, to joke that the machine must have been right, that the machine knew more about Mayer than he did himself. It's the power of a simple mechanical device to make you doubt your own concept of truth and lie — "It turned out I was lying" — that made the polygraph so alluring, and so disturbing. And it's that power, combined with the idea that the machines are peering directly into the brain, that makes the polygraph's modern counterparts even more so.

Today, the polygraph is the subject of much controversy, with organizations devoted to publicizing "countermeasures" — ways to subvert the results — to prove how unreliable it is. But the American Polygraph Association says it has "great probative value," and police departments still use it to help focus their criminal investigations and to try to extract confessions. The polygraph is also used to screen potential and current federal employees in law enforcement and for security clearances, although private employers are prohibited from using it as a pre-employment screen. Polygraphists are also routinely brought in to investigate such matters as insurance fraud, corporate theft and contested divorce.

But there is little scientific evidence to back up the accuracy of the polygraph. "There has been no serious effort in the U.S. government to develop the scientific basis for the psychophysiological detection of deception by any technique," stated a report issued by the National Research Council in 2003. Polygraph research has been "managed and supported by national security and law enforcement agencies that do not operate in a culture of science," the council said, suggesting that these are not the best settings for an objective assessment of any device's pros and cons.

The polygraph has many cons. It requires a suspect who is cooperative, feels guilty or anxious about lying and hasn't been educated to the various countermeasures that can thwart the results. Polygraph results can be more reliable in investigations in which the questioners already know what they're looking for. This allows investigators to develop a line of questioning that leads to something like the Guilty Knowledge Test. This is a multiple-choice test in which the answer is something only a guilty person would know — and only a guilty person's polygraph readings would indicate arousal upon hearing it.

The history of polygraphs is a cautionary tale, an example of how not to introduce the next generation of credibility-assessment devices. "Security and law enforcement agencies need to improve their capability to independently evaluate claims proffered by advocates of new techniques for detecting deception," the National Research Council said. "The history of the polygraph makes clear that such agencies typically let clinical judgment outweigh scientific evidence."