So, Do “Pollyannas” Watch “Lie to Me?”
I never watched the show “Lie to Me” because I had enough police procedurals on my plate (read: one—Castle) and because I have never bought into the idea of a fool-proof or even mostly-proof method of lie detection. I have especially never thought that there was any decent proof for microexpressions—the idea that there are split-second facial muscle “tells” that can be analyzed for lie detection. I’ve mostly never bought into it because I can’t imagine a reasonable explanation for such an adaptation.1 Now, it turns out, there’s another reason not to bother watching the show (other than it’s cancellation): watching it makes you worse at lie detection!
In a study I first encountered via Mindhacks, authors, Timothy Levine, Kim Serota, and Hillary Shulman show that individuals who watch an episode of “Lie to Me” and then attempt a lie detecting task perform much worse than people who watch something else or watch nothing at all (Levine et. al 2010). I’m not terribly surprised by this result given an additional study on individual trust-levels and how that affects lie detection. In their paper “Not Pollyannas,” Nancy Carter and Mark Weber show that individuals who have high-levels of trust (so-called Pollyannas) are better at detecting liars that individuals with a low-trust threshold of strangers (Carter and Weber 2010). So, it seems clear to me that what the television show accomplishes is inducing cynicism. After being saturated with story and imagery that focuses on liars and their “tells” people walk out into the real-world retaining some sense that A) a lot of people are liars and B) It’s really easy to catch them. I imagine these people focusing in on meaningless gestures or speak-os, mostly disregarding any compelling reason for a person to be lying to them. I believe the effect is not dissimilar to what Marshall MacLuhan pointed out, which is that if you only watch the local news, you’d think the world was a horrific place.
Mind you, I would’t consider a good evolutionary explanation as any kind of proof; just that it would give me a reason to look more closely at the evidence for microexpressions. The fact is, we aren’t bad natural lie detectors without using facial expressions and only using information available to us (see Montague et. al 2011). ↩