A bad taste in your mouth – moral outrage has origins in physical disgust

By Ed Yong | February 27, 2009 8:30 am

Blogging on Peer-Reviewed ResearchBoth objects and behaviour can be described as disgusting. The term could equally apply to someone who cheats other people out of money as it could to the sight of rancid food or the taste of sour milk. That’s not just a linguistic quirk. Some scientists believe that the revulsion we feel towards immoral behaviour isn’t based on our vaunted mental abilities, but on ancient impulses that evolved to put us off toxic or infectious foods.

It seems that your facial muscles agree. Hanah Chapman from the University of Toronto has found that both physical and moral disgust cause the levator labii muscles, which run from your eyes to your mouth, to contract. The result: you wrinkle your nose and you purse your lips. Nasty tastes, gross photos and foul play all cause the same physical reaction and the same subjective emotions. When people say that moral transgressions “leave a bad taste in your mouth”, it’s more than just a pretty metaphor.

Chapman began by studying disgust in its more primitive forms – reactions to foul tastes. She recruited 27 volunteers and recorded the electrical activity in their levator labii muscles as they drank small vials of various liquids. If the concoctions were unpleasantly salty, sour or bitter, this group of muscles contracted more strongly than if the liquids were sweet or flavourless. These reactions were a good measure of their subjective opinions – the more distasteful they found the drinks, the more strongly their muscles contracted.

Disgusting photos elicited the same reactions even though no taste buds were actually being offended. When the volunteers saw pictures of faeces, wounds, maggots and other unhygienic visuals, their nose-wrinkling muscles contracted strongly, again in proportion to how disgusted they actually felt. Sad photos, or neutral ones, had no such effect.

So far, so predictable, but Chapman went on to show that our faces react in the same way to immoral behaviour. Humans value fair play and we frown at cheats and thieves. But do we wrinkle our noses at them? To find out, Chapman challenged 16 volunteers to an “Ultimatum Game“.

The game’s rules are simple – the players have to split $10, one of them (the proposer) makes an offer to the other (the responder) who can accept or reject it. Rejection means that both parties lose the money. Rationally, people should accept any offer greater than zero, since they’ll get some money out of it. Realistically, most people choose to screw themselves over to deprive their partner of the profits, if they’re offered less than a quarter of the pot.

The volunteers played 20 rounds of the game in the role of the responder, against either an experimenter or a computer. The offers were always randomly generated by a programme and ranged from an even split to a pittance of just $1.

The players always accepted fair offers that split the pot evenly and, as usual, the more unfair the offers became, the more likely they were to be turned down. And sure enough, their levator labii muscles followed suit. As offers became stingier, so this muscle twitched more strongly (B in image below).

To separate feelings of disgust from other emotions, Chapman asked the players after every round to look at photos of seven different facial expressions and rate how well the visuals summed up their feelings.The faces were taken from a standard set and designed to represent either disgust, fear, anger, contempt, sadness, surprise or happiness.

Unsurprisingly, as the offers became more unfair, the volunteers related less to the happy faces, more to the angry and sad faces and most of all to the disgusted ones (A). When Chapman tweaked the seven standard photos to show stronger or weaker variations of the same emotion, the players related most strongly to pictures of strong disgust, moderate anger and only mild sadness (C).

As before, the more disgusted the players felt, the stronger the contractions of their levator labii muscles (D). Neither anger, contempt nor any other emotion was related to the activity of these muscles (E,F). And of all seven emotions, disgust was the only one whose strength could predict a player’s odds of rejecting an unfair offer.

In both their subjective views and their reflexive muscle movements, the players were telling Chapman the same thing: when people are treated unfairly, they emotions they felt were most similar to their reactions to foul tastes or repulsive sights. Even though immorality leaves an abstract “bad taste” and bitter liquids leave a literal one, both trigger the same facial expressions.

Over a century ago, Darwin himself reckoned that our facial expressions are adaptations. The wrinkled nose and pursed lips that accompany physical disgust could serve to close off our senses from offending tastes and smells – a theory that was confirmed just last year by Josh Susskind, who also worked on this study. The new research provides more support for this idea but Chapman suggests that we have co-opted the protective function of disgust to shield us from social offences too.

Her experiments suggest that moral disgust is born from the same reactions that drives us to snub potentially dangerous foods. This impulse guards us and other animals from disease and poisons and it’s an incredibly ancient one – even sea anemones will recoil from bitter tastes. During the course of human (or ape) evolution, this 500-million-year-old system could have been drafted to protect us from offensive social actions too.

In humans, the range of sensations that trigger feelings of disgust has been greatly expanded from the physical to the abstract world, but the reactions are still the same. Feelings of disgust lead to avoidance and withdrawal, so people could be literally treating unfair offers like plates of rotten food. 

Reference: H. A. Chapman, D. A. Kim, J. M. Susskind, A. K. Anderson (2009). In Bad Taste: Evidence for the Oral Origins of Moral Disgust Science, 323 (5918), 1222-1226 DOI: 10.1126/science.1165565

More on metaphors:  

Subscribe to the feed

Comments (13)

  1. Well that explains why the only time I have said “That’s disgusting” (I think it was a total of 2x) to my children was upon an extraordinary loud burp or something to that effect. I wondered myself why I wouldn’t reserve such strong language for something more important. It was just visceral.

  2. I once felt so scornful that my lip curled up away from my teeth, inverting the upper lip, with no volition on my part. It was a very strange feeling. That was when I realized that curling one’s lip was more than just a saying.

  3. As interesting as this study is, shouldn’t it at least consider the possibility that this is a learned association, coming from education?
    Is there significant transculturel data to allow us to discard this hypothesis?

  4. Sam K.

    This reminds me a lot of Paul Ekman’s research; this brings up the bigger question of whether facial expressions are signals or part of our actual hard-wired reaction, i.e. our facial expression *is* our emotion to some extent. Also, in light of Abie’s comment, I wonder how the reaction time of the muscles varied?

  5. Edwin

    Eh, Harris et al put forward the very same idea last year. Why is there no mention? Is Harris too politically incorrect for (so-called) Scienceblogs?

  6. Ms. Aenthropi

    Is this undergraduate work or something? 27 volunteers; 16 volunteers; at least when I did my work in pseudo-science, I had the self-respect to admit my findings were more likely a statistical anomaly than anything finding.
    Tut tut.

  7. Phil Goetz

    Could causation run the other way – from shared language terms, to shared expressions?

  8. Shaun Sperl

    If you are further interested in this topic read Jonah Lehrer’s first chapter in his first book – Proust Was A Neuroscientist. It is a thought-provoking, superbly written book overall, but his first chapter on Walt Whitman, subtitled The Substance of Feeling, addresses this idea directly. Our feelings begin in our flesh. Not to mention his excellent blog – The Frontal Cortex – also on ScienceBlogs. I sing the body electric.

  9. UlrikP

    Ed, thanks for a thought-provoking article.
    Edwin, thank you for the Sam Harris link; he is one of my heroes.
    @ms. Aenthropi: I don’t quite understand your attitude towards the idea behind the study (even though your choice of words are somewhat revealing). Sure, it’s a small study in terms of the number of volunteers, but you must agree that if verified, this poses an interesting link between man’s biology and psyche. No? In that case, let me ask you if you recall your facial expression at the time you read the article or when you commented it? – joking apart, I think the idea sounds plausible enough to deserve further scrutiny.

  10. Aenthropi

    Maybe it is not the study that bothers me so much as the reporting. I could elaborate two trains: How I dislike the study, and how I dislike this story.
    No one should care in a substantive way what the implications of the theory are at this point; for it is basically untested or it needs much more and rigorous testing than this. The implications of The Mythos of H.P. Lovecraft, if verified, would pose an interesting link between man in the universe. So what? No testing; not science.
    I think the study had too many variables, and the data was over fitted. E and F are useless scatter graphs. We can tell because it looks like someone hit it with a shot gun, and if there is a pattern, there are not yet enough samples to determine it. D has pattern more along the line but again too few samples. I will roll two dice of six sides. They tallied five pips and six: a total of 11 pips; an average of 5.5 pips; therefore, six-sided dice average 5.5 pips. This is obviously wrong. Without the aid of mathematics to predict the average, I need to roll the dice many more times to establish a correct average of 3.5 pips in my samples.
    In truth, I fudged the test and rolled twice. My first attempt was two and four pips, which would demonstrate an average of three. Three is close to the average, but not on yet. Even if I got something exactly–say two and five–I could not know I had the correct average until I repeated the sample many more times, and I have nothing worth talking about until I did. Further related to the study in the above article: if we want to them make a claim about the human condition, we need to test all groups of humans to help shift out what is cultural and what is innate, and I think there are at least a few more than 20 cultures in the world.
    I think C is contaminated, but I am not sure to what extent. Did they leave the words on when they sampled the participants? Words are icons and stand for ideas, thus they would contaminate the results. Even facial expressions are icons for emotions, which are (I guess) well established as cultural artifacts. Take theater masks. Take this :) . Is it not a face? Does it it not represent a smile and then some.
    If the Scientist who did this wants to make a convincing and serious argument, she should need to be more rigorous than this. If this is undergraduate work, then good job. She showed all her work, followed the method well enough and did the motions. As graduate work it is not acceptable to me.
    But maybe I am bothered more by the reporting of it. Why is this author lauding this so much as an accomplishment? I know I did not list everything I find wrong. My peers who reviewed this article with me at college agree it is pseudoscience with psychologists amongst us. It does not have predictive powers. It is not well executed. It is not even finished. This is not yet noteworthy, but it is presented as some kind of breakthrough.
    The author wrote this: “The wrinkled nose and pursed lips that accompany physical disgust could serve to close off our senses from offending tastes and smells – a theory that was confirmed just last year by Josh Susskind, who also worked on this study.” The theory (http://scienceblogs.com/notrocketscience/2008/06/15/fearful-facial-expressions-enhance-our-perception) gets the same glorification, but it is similarly feeble: To quote, “Suskind asked 20 students to make fearful or disgusted faces while looking at a large grid.” Twenty? Only Twenty? And just made the faces. Humbug! We are testing a fear response linked to mortal danger. This is a strong response in humans, who for an irrational reason fear death and demise. They can not just make a face and open their eyes and claim they have got it; they have to scare the participants, shake them to the bone and put the fear of God into them, and then test them. Now this is my kind of testing.
    This finding strikes against my intuition. As I understand, people in mortal stress become hyper focused. Tunnel vision is a problem, as well as hyper fixation on specific details. People who train for positions that involve this sort of stress have to work hard to overcome the reflex.
    So, that people who are not scared see more when they open eyes wide is not a surprise at all. If we want to know how frightened people see, we need to test the frightened.
    This is sensationalism, and if this is typical of Ed Young’s reporting, then he is a sensationalist. I understand how academics need to occasionally make things more accessible to the lay, but they is no excuse glorifying mimsy science or faulty results to the public, many of whom are easily duped (sorry lay people). It is even counter productive. It is misrepresentation. It is lying.
    Testimonial: Recently, my friend and I were walking the hall when we passed someone I know my friend finds morally revolting. As we passed, the revolting person greeted us, and my friend smiled kindly. Why should she smile if the finding of this article is true?
    Panglossian indeed.

  11. Alex Deam

    Aenthropi said:

    I think the study had too many variables, and the data was over fitted. E and F are useless scatter graphs. We can tell because it looks like someone hit it with a shot gun, and if there is a pattern, there are not yet enough samples to determine it.

    Er, that’s exactly what graphs E and F are supposed to show. See the part where it says, “As before, the more disgusted the players felt, the stronger the contractions of their levator labii muscles (D). Neither anger, contempt nor any other emotion was related to the activity of these muscles (E,F). And of all seven emotions, disgust was the only one whose strength could predict a player’s odds of rejecting an unfair offer.” The point is that E and F do look “like someone hit it with a shot gun”, so that shows that anger and sadness (see E and F’s x-axes) aren’t correlated with the movement of these muscles. The only thing that shows any sort of correlation is graph D, which is about disgust, the point of the whole paper.

    Testimonial: Recently, my friend and I were walking the hall when we passed someone I know my friend finds morally revolting. As we passed, the revolting person greeted us, and my friend smiled kindly. Why should she smile if the finding of this article is true?
    Panglossian indeed.

    Firstly, your friend is talking to someone they don’t like morally, not seeing some immoral or disgusting action. The paper is about the responses to actions, not people. I mean if I saw Hitler I wouldn’t be disgusted, but I would if I saw Auschwitz. You should be looking at your friend’s reaction when this other person actually does something “immoral” or “disgusting”. Secondly, the paper isn’t talking about “smiling”, it is talking about the movement of particular muscles. Until you measure the response of your friend’s levator labii muscles, you can’t say anything about her response to this person compared to this study. Also, your story is an anecdote. Not only is your story not scientific, but in it, you do exactly what you accuse this study of doing. Namely, using to small a statistical sample. You complain that there were only 27 people studied, yet you give n example of just a single person. You know full well that if what person doesn’t fit a trend doesn’t mean there’s no correlation.
    As it happens, I agree with you that the sample size tested was probably too small in this study to be meaningful, but I don’t agree with the rest of your points.

  12. Aenthropi

    I should concede that I mistook the scatter graphs, but I should concede nothing of the testimony. I specifically labeled it as non science, an anecdote, something that occurred near me; whereas, I accused the author of labeling non science (or insignificant science) science. The panglosian I suppose is a bit of an obscure reference to a comment made on another article written by Ed Yong.
    Clearly a difference of circumstance between my petty report and this petty science exists.

  13. Tom

    if a cop said about another cops… he sure left a bad taste in your mouth.. what should i think?

NEW ON DISCOVER
OPEN
CITIZEN SCIENCE
ADVERTISEMENT

Discover's Newsletter

Sign up to get the latest science news delivered weekly right to your inbox!

Not Exactly Rocket Science

Dive into the awe-inspiring, beautiful and quirky world of science news with award-winning writer Ed Yong. No previous experience required.
ADVERTISEMENT

See More

ADVERTISEMENT
Collapse bottom bar
+

Login to your Account

X
E-mail address:
Password:
Remember me
Forgot your password?
No problem. Click here to have it e-mailed to you.

Not Registered Yet?

Register now for FREE. Registration only takes a few minutes to complete. Register now »