The face of the thinker

This post was first puvlished in 2006 on the Alphapsy blog.

The hindsight bias is the tendency to say after an event happened that “we knew it all along”, that it’s not really surprising. This is a bias because when asked before the event, we wouldn’t have predicted it, but after it happened we think we would have (and because it can be quite irritating). A 2006 paper in Current Directions in Psychological Science (August issue) that reviews the role of metacognitive thoughts and feeling in this phenomenon, and among the effects mentioned one is quite surprising: people asked to make the face we make when engaged in deep thinking were actually more doubtful towards their answers, as if they had had to think hard to come out with them. Explanation of the experiment.



Lawrence Sanna and Norbert Schwarz have done quite a lot of work on metacognitive processes, and they have just reviewed the part related to the hindsight bias here. An example of such metacognitive process is the feeling of ‘easiness’ that comes with an answer: was it hard to find, or did it spring to mind automatically? In some cases, the size of the hindsight bias can depend on this feeling, as in the experiments reported in Sanna et al. 2002.



In their first experiment, they told people that the British had won a war against the Gurkha. They then asked people to generate reasons either why the British had won or why the Gurkha could have won. Some people were asked to generate 2 reasons and other 10. Participants were then asked the classical question in hindsight bias research: “If we hadn’t already told you who had won, what would you have thought the probability of the British winning would be?

Among the participants who had to generate reasons for the victory of the British, you might expect that those who had to generate more reasons would give a higher probability: since they were forced to represent all these reasons for why the British had won, they should think that the British were sure to win. However, the opposite was observed: people who only had to generate two reasons gave higher probabilities of the British winning than those who had ten reasons to generate. Why is that so? Presumably because finding the reasons was not easy: people had to think hard to come up with ten reasons. They then must have reasoned that since it was hard to come up with a lot of reasons for the British victory, then it must not have been very probable. On the other hand, those who only had to generate two reasons did it quite easily, and they must have reasoned that since it is easy to come up with reasons for the British victory then it must have been quite probable.

An inverse pattern of results is found for those who had to generate reasons of why the Gurkha could have won the war. Those who had to generate ten answers found it hard and thus thought that it was quite probable that the Gurkha would have lost anyway. Those who only had to generate two found it much easier and answered that the Gurkha actually had good chances of winning the war.

The ‘thinker’s face’ makes its appearance in a second experiment. In this experiment some participants were asked to contract their brow, so that they would look like someone who is engaged in some hard thinking. Those who did so while trying to generate reasons answered as if they really had had to think hard in order to come out with the list of reasons. So in this case, it’s not thinking hard that makes you frown, it’s frowning that makes you think that you think hard. (Tip: When trying to solve a problem, contract your brow, and if you come up with an answer you will be very happy since you felt it was such a hard problem!)

Julien Dutant and I discussed the experiment:


Did they have a control where people were to raise their eyebrows instead? How did that played out? (And does a blasé face makes you answer quicker?)


they didn't control for other brow movements. i don't quite see who you would predict this specific effect with other brow movements, but it would have been methodologically cleaner anyway
as for the blasé face, not as far as i know (you could try to do the experiment, if it works you get a publication in a nice journal!)


I've been trying the blasé face all day, no noticeable effect so far.

Well that's a big worry if they didn't have a control with other face expressions! There's a big alternative hypothesis then: people are slower simply because they keep remembering that they have to frown! (Try reading Trends in Cognitive Neuroscience with your hand up in the air for instance!)


too bad for the blasé face, it might have come in handy!
as for the control, people were not exactly slower to come up with their answers, they though it was harder for them to come up with these answers (it's not the same thing)
anyway, i don't think that frowing drains a lot of your cognitive resources. then again it's true that they should have controled for the specificity of the task (by e.g. raising their eyebrows as you suggested)


My mistake then!

By the way, this might be an instance of a general associative mechanism: when one makes the face associated with an emotion, they're more likely to feel it (caeteris paribus).

(Sorry if I'm reinventing the psychological wheel – and I know this gets us in the kind of things…)


actually there is a very funny experiment in which the participants had to hold a pencil with their teeth (like a pirate and his knife) while doing an unrelated task. holding a pencil in such a way makes you (kinda) smile. and they were more happy after the experiment than people who had had to hold a pencil with their lips, and thus making a 'sader' face.
it means that there is some bottom up process going on here. the question could be: is it a byproduct of the 'normal' top down mechanism (you make the face that goes with you emotion) or is it a specialized mechanism (having to do with coordinating your appearance and you emotions for some social reasons)? the first seems to be much more likely, but who knows?


No comments yet