Lie detectors

Two underlying assumption in discussions of detecting lies are that the person speaking knows that the statement is a lie and that there is an unambiguous physiological measure of lying when a person is doing it.

The first assumption is belied every day in the newspaper. People say things that contradict what other people say. Often, both believe that they are telling the truth (so both would pass a lie detector test even though the truths that they are telling are incompatible.)

Second, in order to calibrate the physiological measure you first have to prove that such a response exists. It is not clear that it does. Then you have to show that this response is consistent across people and times. It is not clear that it is. You also have to show that the response works for the liars, the confused, the delusional, and the ignorant. I have no idea how to pull this one off.

As a specific example, if the examiner thinks that taking home classified materials is always bad and illegal, but the examinee thinks that, when the pressure is high enough, it is OK to take some materials home, then a lie detector will be confused. The answers given will have more to do with situational ethics and less to do with some external standard of what a lie is.