Human Lie Detectors

Our court system assumes that people can tell whether other people are telling the truth by evaluating the demeanor of the other people. This is true in a wide range of contexts. Jurors are encouraged to consider the demeanor of each witness in deciding whether to credit the witness’s testimony. N.C.P.I. — Crim. 101.15 (referring to the “manner and appearance” of the witness). When judges must determine facts, their fact-finding is reviewed deferentially because they had the opportunity to assess the demeanor of the relevant witnesses. Yurek v. Shaffer, __ N.C. App. __, 678 S.E.2d 738 (2009). An officer’s assessment of a suspect’s demeanor — like whether the suspect appeared to be nervous — may be a factor in deciding, for example, whether the officer was entitled to detain the suspect for further investigation. State v. Jacobs, 162 N.C. App. 251 (2004).

It turns out that this assumption is mostly false. There’s a wealth of social science research on lying and lie detection, and “[t]he major finding is that people, both lay persons and presumed lie experts working within the legal field, are not very skilled in distinguishing between truthful and deceptive statements, with hit rates generally around the level of chance.” Maria Hartwig, et al., Detecting Deception, Law and Human Behavior (Aug. 2005). As the quoted language suggests, police officers aren’t much better than laypeople at detecting lies based on behavioral cues. Nor do people who express confidence in their prowess at lie detecting perform any better than average.

Part of the reason is that the behavioral cues for which most people look — like nervousness, or the refusal to make eye contact  — aren’t statistically associated with lying. In fact, some research suggests that a principal sign of nervousness, fidgeting, actually diminishes when a person is lying, perhaps because the person’s brain is working so hard to get the person’s story straight that it doesn’t have any energy left for fidgeting.

There are a thousand possible implications from this research, including whether the legal rules referenced in the first paragraph of this post should be modified or eliminated, e.g., whether we should stop encouraging jurors to assess witnesses’ veracity with reference to their demeanor. But one possible implication strikes me as particularly interesting and controversial: should the exclusion of polygraph evidence under State v. Foye, 254 N.C. 704 (1961), be reconsidered? Such evidence was rejected in part because of its perceived unreliability — but its unreliability appears to be dwarfed by the unreliability of the subjective assessments of the human participants in the criminal justice system. In fact, one can make the argument that we should start using computer programs, which have shown promise in detecting lies based on verbal patterns such as a reduction in first-person pronoun use and an increase in words expressing negative emotion — in court to monitor witnesses’ testimony. Or even farther afield, that witnesses should be subjected to the so-called “no lie MRI,” which looks for brain wave patterns characteristic of deception. Your thoughts?

1 thought on “Human Lie Detectors”

  1. From my experience, polygraphs belong in the same category as Ouija boards and tarot cards.

    I did have the privilege of hearing Don Rabon give a presentation on lying and interviewing techniques once at the Justice Academy. I think his interrogation and interviewing techniques should be required training for all law enforcement, criminal justice and corrections personnel. The only problem is, folks would be using it on bosses, co-workers and spouses! I guess the rule of thumb should be… if you don’t want to know the truth, don’t ask!


Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.