Taking emotions at face value

Boston Magazine has a fascinating article on the work of psychologist Lisa Feldman Barrett who has been leading the charge against the idea that we recognise the same facial expression of emotion across the world.

This was first suggested by Paul Ekman whose work suggested that humans can universally recognise six emotions: anger, disgust, fear, happiness, sadness, and surprise.

His research involved showing people from different cultures pictures of faces and asking them to label each expression from a choice of emotional words.

But Barrett has found a simple flaw in the procedure:

She returned to those famous cross-cultural studies that had launched Ekman’s career—and found that they were less than watertight. The problem was the options that Ekman had given his subjects when asking them to identify the emotions shown on the faces they were presented with. Those options, Barrett discovered, had limited the ways in which people allowed themselves to think.

Barrett explained the problem to me this way: “I can break that experiment really easily, just by removing the words. I can just show you a face and ask how this person feels. Or I can show you two faces, two scowling faces, and I can say, ‘Do these people feel the same thing?’ And agreement drops into the toilet.”

The article is on much more than this controversy in cognitive science and also tracks how research on emotion and facial expression is playing an increasing role in law enforcement – with not all of it well supported by evidence.

And if you want links to some of the scientific papers, the always interesting Neuroanthropology blog has more at the bottom of this post.
 

Link to Boston Magazine article ‘About Face’.

8 Comments

  1. Posted July 1, 2013 at 12:50 am | Permalink

    I’m not a very “girly” girl. I frequently have this sort of problem on the rare occasions I read a “ladies’ ” magazine (like Cosmo) and I take one of their “relationship” quizzes. Invariably, there’s at least one multiple choice question (and usually several) where none of the options given match what I would do in a given set of circumstances. Because I score myself a zero for those questions, my overall score drops into the toilet, and the magazine tells me I “must be doing it wrong”, and I don’t really understand the seven signs of a bad relationship or whatever the case may be.

    Therefore, I’m not really surprised at Ms. (Dr.?) Barrett’s results. When the only choice I’m given is “Does this person look happy or sad?”, but I look at the picture and I think that the person looks confused, how am I supposed to answer that question?

    • Posted July 1, 2013 at 4:29 pm | Permalink

      That’s not actually how Ekman’s experiment went. When he was present with the cultures in Papua New Guinea, Ekman was faced with a series of difficulties.

      First was the translation of the concept of the emotions; none of these were easy to convey to they New Guineans. When he tried to have the concepts translated, through a boy translator from the village, the emotions couldn’t have been more vague.

      He tried to give the simple duality of options, “is this person happy or sad” and the mark was still missed.

      He then tried to show them photographs! Which didn’t help at all either because they’d never seen photos or cameras and had no clue what they were supposed to do with them.

      How, then, could this have been fixed? After all this – nothing worked. So Ekman devised a more interesting approach; he would take pictures of them to capture their emotions.

      But, then again, how would Ekman know what they were talking about while the story was translated? How would he have any clue about their emotions?

      He’d show them a picture of a woman and they’d just appear confused, what could they do?

      The cultural problem was that they had these “different experiences” that shaped and created their emotions.

      If you or I heard that a boar or pig was lost; we likely wouldn’t bat a lash – just go get meat from the market.

      Ekman began to let go of his search for emotional recognition and listened to stories from these people, then he collected their concept of emotions and this helped get rid of his “anthropological gaize.”

      Soon he was able to capture images of their faces making emotions which he then was able to translate back to them as – concepts – that were linked to specific cultural experiences with them.

      And – tada – emotional universality as expressed through the human face began to be uncovered.

      At this point he could ask them, “happy or sad,” and then the New Guineans were able, after all this time, to say things like, “neither!” Or, at least, their word for neither.

      Sorry… See earlier comment… I’m slightly passionate about this material; I have much of my own research done on the correlation between this material and facial – grammar – in signed languages. I.e. I’ve read all this material tenfold.

      • Posted July 1, 2013 at 5:32 pm | Permalink

        Understood. It’s very true that frequently things do get “lost in translation”. I used to work in a grocery store that catered to a largely Hispanic clientele. I understood enough basic Spanish to get by, but not enough to carry on a full-blown conversation. Many times I’d be speaking English to my bilingual colleagues, I’d use an English idiom, and they’d look at me blankly. I also took French in school. I’ve always been fascinated at how idioms frequently don’t translate well across languages (“my little cabbage” as a term of endearment in French”, for example–you don’t hear that in English). It doesn’t surprise me that different cultures might have multiple words to describe “happy”, for example, but perhaps only one word to describe “sad”.

  2. Posted July 1, 2013 at 4:25 am | Permalink

    The idea of universality of emotions and facial expressions originated with Charles Darwin in 1878, in his book The Expression of the Emotions in Man and Animals.

  3. Posted July 1, 2013 at 4:13 pm | Permalink

    Lisa Barrett’s poorly composed ideas in a nutshell…

    First, “microexpressions,” are not, nor are they labeled by Ekman as such, indications of a lie. They are indications of an emotional interruption on the part of the speaker; they are leaked emotion which, once noticed, can only be noticed as it being that emotion. See Also: “Othello’s Error,” which Ekman has written extensively about that states, in short, to say you are aware of the lie or of the reason for the emotion you are being fallacious for you can never really know for certain the certainty of either.

    Secondly,”[Ekman's research has been] verified and expanded upon in hundreds of studies.” In multi-blind trials that stand up against the “well this shit better be inevitably repeatable to be proven as scientific,” method of… What’s it? Oh, yes, science.

    Thirdly,”she believes [...] emotions themselves don’t have their own places in the brain…” See also: Limbic system, basal ganglia, temporal lobes etc. etc. “…or their own patterns in the body…” See also: blood rushing to quads and glutes while experiencing fear, blood rushing to hands during anger, simultaneous contraction of the orbicularis oclui and zygomatic major during happiness, see also dilation of pupils during excitement, see also, see also, see also.

    Fourthly, Bartlett states, “each of us constructs [emotional expressions] in our own individual ways, from a diversity of sources: our internal sensations, our reactions to the environments we live in, our ever-evolving bodies of experience and learning, our cultures.” Which is also to say that she hasn’t finished reading all of Ekman’s material because he says the same thing. See also: Cultural Emblems (http://www.eiworld.org/docs/Emotional-And-Conversational-Nonverbal-Signals.pdf)

    Fifthly… Forget it…

    You see where I’m getting with this; I’m not even halfway through this article and I’ve got better things to do than kvetch on here.

    • Posted July 1, 2013 at 4:15 pm | Permalink

      *Better things to do* – trust me – I can rarely find something better to do than read Mindhacks posts.

      I meant that all I’m doing at this point is getting fired up about a science I really care about; I’m spouting from a real biased perspective and that’s not argument worthy at all.

      Blah.

  4. quidnunc
    Posted July 1, 2013 at 6:29 pm | Permalink

    I remember from the one nonverbal communication book I read that interpretations are context dependent so it’s not surprising that some facial expressions can be ambiguous across some class of emotional states

    • Posted July 2, 2013 at 7:30 pm | Permalink

      Absolutely. And this is something that Ekman covers extensively as well.

      Categories of emotional expression on the face include and are not limited to Micro- Macro- Partial- Subtle- Asymmetrical- expressions…

      There are a million categories of vagueness.


Post a Comment

Required fields are marked *
*
*

Follow

Get every new post delivered to your Inbox.

Join 23,128 other followers