Most Sincerely

Wow! The comments on yesterday’s blog were so thoughtful and interesting that I’ve decided to continue this thread for another day. The question I’d like to focus on today is this: What would indeed be a scientifically valid test to distinguish sincere emotion from merely the artful simulation of it through acting?

The first commenter points out that Ekman’s micromovements are not something of which the observer would be explicitly aware. It is plausible, but not at all certain, that one difference between a merely good actor and a great actor is that the latter is actually incorporating these subliminal micromovements into the performance. Detection of micromovements, together with correlative measurement of how convinced people were by a particular actor’s performance, could – and should – be incorporated into the testing protocol.

A later commenter points out that we humans may have a native ability to suss out the fake when we are in the same room with someone, perhaps through smell, that we don’t have when we are looking at a video. So today I’m going to talk about how one would go about testing for the ability to detect sincerity when there is a constraint that everyone is in the same room. Then this ability could subsequently be testing against our sincerity-detection abilities when looking at a video, through the use of a 2×2 study.

By the way, this same commenter also points out that the strength of a democratic system rests largely on its ability to function despite the fact that people cannot truly trust politicians. Point well taken!

This commenter’s first point, about things needing to happen in person, suggests a double-blind study involving two kinds of participants: (i) a volunteer questioner; (ii) a respondent who is either a volunteer an actor. The questioner is the test subject.

The questioner asks a fixed series of questions, and is not informed as to whether the respondent is answering sincerely or merely acting. First the questions are asked by one questioner of a sincere respondent. Then an actor who has viewed a recording and transcription of the first session is charged with trying to duplicate the “performance” of the sincere respondent. For this session a different questioner is given the same questions to ask. After each session, the questioner is asked whether he/she believes that the respondent was actually an actor.

This process is repeated over a number of different sessions, using different participants as questioners and different respondents. The protocol would measure for a systematic ability on the part of a population of questioners to accurately assess the true nature of their respondent.

4 Responses to “Most Sincerely”

  1. manooh says:

    I can’t help but wonder how much influence the abilities of the actors have. I don’t know much about acting, but I’d assume that the degree of identification with the role has some part in it. So on one side, there might be actors who can produce the “perfect fake”, controlling all micro-muscles and knowing how to use them appropriately. On the other side, there might be the “identifiers”, the ones who are able to put their reality into the role, living it instead of playing it. The “perfect identifiers” might even be indistinguishable from people telling the truth.

  2. sally says:

    I thought that the issue with Ekman is that the creator of the micromovements doesn’t have control over them. They are the subconscious movements of his/her true feelings and that the interpreter of the micromovements can interpret them in a subconscious way–but may not always be able to interpret them…

    So, I would say that the actors could try, but I don’t know if they would be able to have that much control over their own micromovements.

    Also…the interpreters could try, but probably could not observe those movements either….

    The other thing that is sort of unique and wonderful and also really funny about human beings, is that we trust. If we like someone, or they tell us what we want to hear, or if they are cute or famous or handsome or rich, it seems that people become more willing to believe –even against their better judgement.

  3. Klipper says:

    It struck me reading yesterday’s post, as well as today’s, just how much the test proposed looks like a Turing test. In this case the computer has been replaced with an ‘insincere’ human.

    But its an interesting experiment. Of course, you’d have to really cover a lot of different subject matter. For example, people might be much better at judging another’s sincerity when making a claim like “I promise not to hit you and steal your woolly mammoth meat”, than “The gazelles two valleys east are much easier to hunt than these” or “lower corporate taxes will mean a bigger paycheck for you.”

    I can’t remember if I read it in Pinker or Dawkins, but there was some discussion of human’s ability to lie. Obviously there are survival advantages that accrue to successful liars. The point was that the ability to lie to others implies an ability to lie to oneself. In essence to believe, at least partially, in the falsehood, so as to overcome the listeners skepticism.

    Since humans are not pathological liars, an ability to detect lies is implicit. We don’t lie because we inherently understand that there’s a good chance we’ll get busted.

    So I am interested to see if its possible to nail down that ability, and see what its limits might be.

  4. Doing this as a study in a lab setting might be too far from realistic conditions to provide useful results. You might need to somehow record people interacting who do not know that they are part of an experiment.

    What you’re talking about here is a lot like “lie detector” technology: is there a way to test whether someone is telling the truth or not? I read an article about this recently, which I can no longer find. It gave the usual summary of what’s known about polygraphs, etc. But its main point was that some scientists are trying to use functional MRI to see what parts of the brain “light up” when someone is lying. The dream of a real, working lie detection machine has been around for a long time. (Imagine political candidates being pressured to submit to such a test!!)

    A great article about the Ekman work:

    When you get right down to it, though, my voting decisions aren’t usually based on whether I think certain statements are lies or not, anyway. What you’d really like to know is about a candidate’s character. I don’t mean “personality” (would I like to have a drink with this person?). I mean their real beliefs and convictions, and why they have them, and how they’d be likely to act on them.

    Although in this particular Democratic primary, electability is everything:

Leave a Reply