The Machine That Measured Fear: Marston, Larson and Keeler

The polygraph has three inventors and a disputed lineage, and the disputes are instructive.

William Moulton Marston, a Harvard psychology student, developed a systolic blood pressure test for deception between 1915 and the early 1920s. He was an energetic promoter of it, and he is also the creator of Wonder Woman, whose lasso compels truth — the connection is not coincidental. In 1923 a federal appeals court refused to admit his test in Frye v. United States, and in doing so produced the general admissibility standard that governed American scientific evidence for seventy years: a technique must have gained general acceptance in its particular field.

John Larson, a Berkeley police officer with a doctorate in physiology, built the first instrument in 1921 that recorded blood pressure, pulse and respiration continuously on a moving paper strip, and used it in real cases. He spent the later part of his life regretting it, describing the instrument's use as having become a Frankenstein's monster and objecting to how it was applied.

Leonarde Keeler, who worked with Larson, added electrodermal response, patented the Keeler Polygraph, commercialized it and trained a generation of examiners.

What the machine measures is not in dispute. It records sympathetic nervous system activation: blood pressure, respiration rate and depth, and skin conductance. These are reliable indices of arousal.

The problem, stated once and applying to everything in this part, is that arousal is not deception-specific. Fear of being disbelieved produces the same autonomic profile as fear of being caught. So does anger at being suspected, so does anxiety about a job, so does the general condition of sitting in a chair wired to a machine that will determine your future.

The Control Question Test, developed by John Reid, is the attempt to solve this. Alongside the relevant questions about the crime, the examiner asks probable-lie control questions — broad questions about minor past transgressions that almost anyone would answer dishonestly, such as whether you have ever taken anything that did not belong to you. The theory is that an innocent examinee will be more disturbed by the control questions, having lied on them, while a guilty examinee will be more disturbed by the relevant ones.

The theory requires assumptions about the examinee's internal comparison of question categories that cannot be verified in any individual case, and that is the enduring scientific objection to it.

There is a further element that the technical literature tends to treat as a confound and that this guide treats as the main event. The pre-test interview, in which the examiner explains that the instrument is highly accurate and demonstrates it with a card trick, is not preparation. It is the persuasive core. The examinee's belief in the machine is what produces admissions, both before and after the test.

Which makes the polygraph, in a great deal of its actual use, an instrument of interrogation wearing the costume of an instrument of measurement.