Say It. Draw It. Act It.


Check out Vrij et al., Outsmarting the Liars: The Benefit of Asking Unanticipated Questions (PDF), from Law and Human Behavior (June 2008). Vrij notes that

If investigators interview individual suspects once (with no factual information about the case), they tend to rely more on noverbal cues than verbal cues to detect deceit. However, when investigators have access to multiple statements from different persons they change tactics. In such cases, investigators overwhelmingly tend to focus on speech content. In particular, they examine the consistency between different statements, believing that consistency implies truth telling and inconsistency implies lying.

One of the areas of human knowledge that we should be scavenging to become better trial lawyers is that of interrogation. Test this: for “investigators” in the above excerpt, read “lawyers” or “juries”.

The Vrij article is an exploration of consistency between multiple witnesses to the same events.

The only study published to date examining the consistency between statements provided by pairs of truth tellers and pairs of liars found that liars were more consistent than truth tellers. The high consistency amongst liars can be explained by the argument that liars prepare for the interview.

The method that the Vrij article investigates is the “theory-based interviewing strategy” of “asking unanticipated
questions” (other such strategies are “withholding information” and
“imposing cognitive load”) to pairs of witnesses.

Liars’ “agreeing upon a story” strategy has weaknesses that investigators can exploit. Specifically, this strategy is limited in its utility: It may work, but only if liars anticipate correctly the questions that will be asked. If investigators ask questions that they liars did not anticipate, the liars will not be able to use their planned answers. Answering such unanticipated questions may yield inconsistent answers. . . . Asking unanticipated questions about central topics should therefore give rise to tell-tale inconsistencies amongst pairs of liars.”

(Again, read “lawyers and juries” for “investigators”.)

In the experiment, pairs of liars and pairs of truth-tellers were asked unanticipated spatial questions (“e.g., ‘In relation to the front door and where you sat, where were the closest diners?’”), asked unanticipated temporal questions (“e.g., ‘In which order did you discuss the different topics you mentioned earlier?’”), and asked to draw the layout of the restaurant at which they purportedly had lunch.

The questioner (who was not informed of the hypothesis of the experiment) was able to correctly classify 60% of truth tellers and 80% of liars on the basis of correspondence in spatial questions, and able to correctly classify 80% of truth tellers and 75% of liars on the basis of correspondence in their drawings.

My experience is that if drawings are good, then reenactments are better. I found this quote here:

Alfred Adler, a psychologist who collaborated with Sigmund Freud, said:
“Trust only movement. Life happens at the level of events, not of words. Trust movement.”

Two liars who can relate an untrue story convincingly verbally are unlikely to be able to be able to draw it and seem truthful. I’ve found that they are even less likely to be able to reenact the events and seem truthful.

Putting events into action works to unmask the lying single witness as well as the inconsistent pairs of witnesses — the nature of human memory is such that putting things into action often reveals the truths that people are able to avoid speaking aloud. We don’t have to be conscious of this to apply it, so jurors are much more likely to be able to detect the lying witness through watching her actions than through listening to her words. They’re also much more likely to correctly classify the truthful witnesses through their actions than through their words.

So have witnesses draw you a picture and put their stories into action, but only if you want the jury to share your opinion of their truthfulness.

(Thanks to Defending People reader, law student, and future Houston criminal-defense lawyer Josh Zientek for his research assistance.)


0 responses to “Say It. Draw It. Act It.”

  1. Mark,

    You’ve gone into type I and type II errors before, and the success rates for those interrogation techniques practically demand the same treatment.

    It’s tempting to read 80% and 75% as “pretty good” and 60% as “better than nothing”, but that’s really only true if about half of people are liars. If either lying or truthtelling is even a little rare, those tests may be worse than useless.

    For example, if only 10% of people lied to the police, then spatial questions would accurately identify 10% * 80% = 8% of people as liars, but would also inaccurately call 90% * 40% = 36% of people liars: i.e., if they think on the basis of spatial questions that you are lying, there’s a 36/44 = 82% chance they are wrong.

    —Aaron.

    • Good point. People are still woeful at detecting liars and truth-tellers.

      Among other things, this has got to be factored in when deciding whether to put a defendant on the stand — a truthful defendant who is perceived as a liar is worse off than one who doesn’t testify.

Leave a Reply to Gritsforbreakfast Cancel reply

Your email address will not be published.