The intersection of science and the law is a confused place: Lawyers used to bending facts and law to suit their clients often cannot tell real science from quackery and many scientists get their understanding of the law and court proceedings from TV detective shows. A long-running example of the nonsense on the science-law border revolves around the “lie detector” or, by it’s pseudo-scientific name, the “polygraph,” from the Greek for much or many writings — even it’s name is nonsense. Lie detector interpretation is entirely subjective, with high error rates — both false positives and negatives.
The polygraph’s efficacy is a myth — it’s “gizmo theater” — not real science. The show of the various inputs for pulse, electrical resistance on the skin (for sweating), respiration rate — with the little scratchy pens tracing in parallel on a roll of graph paper — is supposed to impress you that it’s “scientific.” The script followed by the person administering the test is part of the game, to make the subject nervous and believe in the magic powers of the machine. About the only time the polygraph “works” is when the subject spontaneously (but under duress) admits guilt or something probative of guilt. In that situation, the machine is just a prop, along with the patter of the operating con artist.
I know quite a bit about lie detectors and the fools who use them because I used to be a trial attorney at the US Department of Justice, which lets federal agencies use them, mostly as part of security clearance procedures, where it’s all about the theater — no one goes to jail, but they do cost people their jobs. I also worked on a well known case brought by a quasi-religious sect which uses a crude version of a lie detector to gather information about its members. And I brought perhaps the only, and apparently the first at least, civil case for lie detector malpractice. When the lie detector company engaged by my client failed to catch a manager who was actively stealing from the cash register, my client sued. The lie detector operator’s defense (raised by his insurance company) was that lie detectors are fake — so any client relying on them should have known they cannot be relied upon. Continue reading