After parting from our ape cousins some six million years ago, us humans have evolved easily readable faces. Hairless where we need them to be - the forehead and area around the eyes - we are able to recognise one another, read each other’s emotions and spot intentions, helped by our elongated eyes with their exposed sclera (the white bit) making gaze direction visible. We also use our easily readable faces to emit our own feelings to others. However, we wrongly use people’s faces to make snap judgments about their character, it’s a mistake we can’t help. Our brains have not adapted to our age.
Konner (1982) [1] suggested that our interest in faces originally adapted to unravel kinship. In other words, from an early age, we are on the lookout for the presence or not of shared ancestry. It’s true that babies are drawn to faces, and from two-days-old they can recognise their mother’s face, staring longer at it (and a caregiver’s) than a stranger’s, as demonstrated by Bushnell et al (1989) [2]. Babies are hardwired to be attracted to paired, horizontal dots or circles, an innate skill, according to Givens (2010) [3].
Human brains instinctively attend to other human faces and interpret their messages. The mimicking of expressions is one way that babies learn how behave. Meltzoff & Moore (1977) [4] found that babies as young as two weeks olds are able to imitate adults’ facial movements such as a protruding tongue or an open mouth. And yet, babies have really poor vision, bad enough to be registered blind. When they look at a face, they only see a blurred combination of parts, and respond only to faces as a whole, rather than individual features. This is something that never leaves us. It’s difficult to look in someone’s eyes but not be influenced by their mouth (and vice versa) [5].
Our region of the visual cortex used to identify faces – which Gauthier et al (1999) found to be the fusiform gyrus – works when we see people as they would have been seen for millennia, with normal shading patterns (such as being illuminated from above). Damage to the fusiform gyrus can result in face blindness or ‘prosopagnosia’ (a term coined by Bodamer) [6], sufferers of which find damaging social consequences.
We have evolved to make quick face-based decisions because it’s energy efficient and instantaneous, the latter keeping us from danger. Faces are constantly moving displays of expressions and we need to speed read to keep up with them. Our ability to recognising emotional expressions (in context) and identify faces, all in one tenth of a second, is one of the key skills that we have brain regions dedicated to. But, because we can judge transient moods from faces, we think we can also judge a person’s character.
Take our resting faces, which are not the neutral expressions they should be, our faces typically adopt a ‘look’ that resembles an expression of emotion, and from this, others are inclined to attribute meaning [7]. The static and neutral signals of a face are sending messages and these are flawed.
As far back as Aristotle (Physiognomica) [8] it was thought that humans resembling animals also possessed something of their character; for example, a wide-eyed, flat nosed, ‘cow-faced’ person would be lazy and stupid.
By the 16th and 17th century physiognomy was very popular. This is the art of judging character from facial characteristics. Charles Darwin was nearly prevented from traveling aboard the Beagle because of his nose as captain FitzRoy, a physiognomist, had “made up his mind that no man with such a nose could have energy”. Discredited as it now is, physiognomy has never gone away, and reached dangerous levels in the 19th century when Francis Galton (he of the term Nature v Nurture) brought facial structure into the eugenics debate. He even suggested that each race had an ideal type or form, a bit like dog breed standards in Crufts. With racist implications and no scientific value, physiognomy remains a controversial subject.
It’s easy to see why physiognomy is appealing. Our faces are evolved for efficiency so in different climates we may have evolved differently. Blackford and Newcombe [9] saw human nose differences as evolutionary adaptations, with those coming from hot and humid climates possessing a low and flat nose, and those from cold and dry regions having high and thin noses, both adapting to be more efficient for their environs. But revealing character or intelligence is a different matter.
Let’s equate all human evolution to a 24-hour day. It would only be that in the final five minutes of this day that we have lived in societies largely populated by strangers. Previously we would have been in smaller, manageable groups where we recognised everyone or rarely mixed with anyone whose characters we didn’t know. There would have been no reason to ‘read’ character from faces but every reason for decoding fleeting emotions and intentions. Evolution hasn’t caught up with our dependency on reading faces meaning we read too much into them.
First impressions are the most dangerous of all because of their power and often lasting influence. We can’t help this unconscious response to seeing faces. We even attribute personality to cartoon or toy faces, and even to faces spotted in clouds. There is also evidence that we form the same impressions that others do.
By design, we make rapid verdicts on character, our neural mechanisms kicking in to determine if a person is friendly or hostile, argues Todorov et al (2009) [10]. On seeing a face, associated feelings are triggered, based in part on the memories of faces seen previously. If a person resembles someone we like, ourselves or a family member, we are more likely to view them favourably [11]. Yet, despite our experiences of seeing faces being unique to us, we often agree with others. Our knowledge of faces is so generalised and influenced by evolution (and often shared culture) that we arrive at the same judgments even if we know, consciously, that it’s wrong to do so.
Before looking at those areas of agreement, I must expand upon why we fail so badly when it comes to inferring character. Images are snapshots, and whilst these may express emotion, emotions don’t reflect true character. In the case of photographs, the effects of head posture or the angle of the camera or lighting or context are all examples of influences beyond personality. There will be hundreds of images of an individual, showing all kinds of expressions, and when we see one of them it’s likely to have been chosen for a reason, be it a profile on a dating website or negative article in a tabloid. Seeing people live, in the flesh, has similar problems. We are “victimized by what we see,” in the words of Billy Beane [12].
Billy Beane was a pro baseball player, who looked the part. His playing career never matched the standards his face suggested they would but he stayed in baseball. He joined the Oakland Athletics’ front office as a scout in 1990, was named general manager after the 1997 season. Beane’s success was based on removing bias opinions and judgments from player recruitment. Instead of picking players who looked like baseball players, he followed the stats. Impressions about capabilities made from appearance are usually wrong. Take the interview illusion (Nisbett and Ross) [13]. It’s much better to rely on references from people who know a candidate well than to go with your gut.
Faces are quickly rated for attractiveness, trustworthiness and dominance. We spot who is a potential mate, competitor, threat or benefit to us. Age, gender/sex, and expressions are fast determined and if ambiguous we veer on the side of caution. If, for example, we are unable to determine the sex of a person at first glance, we hold our gaze a little longer or glace back. Trust often follows familiarity. The more we see certain types of faces, the more we trust people with that look. Atypical faces are less trusted [14] and, in many cases, so are ethnic minorities (if outgroups). First impressions, the unconscious, superfast responses, are racist by design or at least pro ingroup, going back to evolved preferences for kin and one’s own tribe. Atypical faces are not as trusted, and yet many brains don’t trigger a negative response to seeing a different race. This is either due to the positive experiences of seeing others with similar faces, or the conscious brain kicking in to overrule the caution. By seeing a broad range of faces, ideally from birth, we remove the perceived ‘threat’ of the unfamiliar. Sadly, we live in a prejudiced world where those who look untrustworthy often receive greater prison sentences and are more likely to be wrongly sentenced to death. Along the same lines as the atypical, more distinctive faces are also perceived as being untrustworthy or criminal. Police line ups have even been shown to avoid atypical/’criminal’ faces in their line ups.
So, other than atypical and unfamiliar, what does an untrustworthy face look like? A key feature is the eyebrows (which have been shown to be more important for recognition and expression than the eyes themselves). Furrowed/angry eyebrows look untrustworthy, especially if they are close together.
The eyebrows of more feminine and youthful (baby faced) faces – and therefore perceived as less threatening - have a wider glabella (the space between the brows) [15]. Studies by McArthur & Apatow (1983, 1984) [16] showed that when faces were manipulated to be seen as baby-faced – smaller chin, larger eyes etc - the people’s characters were deemed to be weaker, socially submissive and intellectually naïve, but potentially more trusted. Trustworthiness is linked with attractiveness which, in turn, is linked with a more ‘feminine’ face [17].
The ‘attractiveness halo’ we place on people means we are more likely to trust and invest in the good looking. And, to generalise, younger looking adult faces are more attractive than older looking ones. Older faces look more dominant and this can signal threat and therefore mistrust in the receiver. Perceived health is also attractive and it’s not just a sexual selection preference (healthier offspring), it’s also the case that think our own health might be threatened by hanging out with someone ill looking.
The look of health – which equates to attractiveness and trust – is seen on well-rested faces that has a healthy glow or yellow hue aided by a good, carotenoid rich diet [18]. An increased contrast between skin colour and features is also seen as more feminine, submissive and trusted (women’s eyes and mouths are darker, make up playing a role). Smoking however affects the production of collagen, harming health which harms trust.
Less important than current health but still significant is the shape of a person’s face, for example a square, larger chin and longer than average face (height of) portray dominance. This is linked to testosterone and the change in our skulls which begin at puberty [19]. But remember, there is no solid evidence that masculine faces are more dominant or aggressive, it’s only perception.
Also affecting perception is fWHR ratio, the ratio between the width (between the cheekbones) and height (distance between upper lip and eyebrows). Typically, male faces have a higher fWHR [20] and are deemed more aggressive/dominant. This could be because men are, on average, physically stronger than women, and more aggressive, and, in a first impression, we are reading a person’s ability to cause us harm. Whilst fWHR doesn’t actually show dominance it is linked with BMI, with wider faces likely to belong to heavier and therefore stronger people. In evolutionary terms, the wide-faced may have been more likely to eat first/more and be the alpha males. We can’t judge general muscle mass from a face but face mass we can.
The fWHR is even a predictor of how people vote. In wartime or in selecting right wing candidates a higher ratio is preferred. In peacetime or when selecting from left wing candidates a lower ratio is favoured. The look of ‘competence’ – an agreed rating - predicts 7 out of 10 election results [21]. When voters are fairly unfamiliar with candidates and their policies, looks play the most important factor is determining who gets the vote. Appearance has the greatest effect on swing voters who, if primed or inclined to distrust – as a result of a negative first impression – have their bias confirmed when they read or hear negative press coverage. These biases have a greater effect if voters see faces on the ballot paper, as they do in Ireland. The unknowledgeable always go for shortcuts. It’s easier for the brain’s efficiency to simply go with the gut.
Our faces are transient, changed by light, make up, Botox, surgery, expression, angles, hairlines, and more, but there are some permanent features, otherwise facial recognition wouldn’t work. Can these static signals reveal anything about character? No. But they can hint to it. When we regularly or habitually express emotions the muscles used are worked and, like exercising any muscles, they can become more defined. So, it’s possible that emotional life histories drive impressions. Long term partners may even share similar expressions and begin to look more alike than newly acquainted couples. I knew an elderly woman who told me she didn’t wear a wedding ring because people could see she was married by looking at her face.
The big takeaways from this post. Reading faces is an evolved skill helping us to detect emotions and recognise people but not to infer character. But we do it anyway so be careful or biases will get the better of you. Oh, and if you want to look more dominant, feminine or attractive, consider changing the shape and look of your eyebrows.
References and further reading:
I recommend Face Value (2017) by Alexander Todorov, which covers the snap judgements we make from seeing people’s faces. Not cheap but worth the price Face Value is packed with research and images.
[1] Konner, M (1982) The Tangled Wing: Biological Constraints on the Human Spirit. Holt.
[2] Givens D. (2010). Your Body at Work: The Why of Human Behaviour.
[3] Meltzoff, A N & M K Moore (1977) Imitating of facial and manual gestures by human neonates. Science 198, 75-8.
[4] Gauthier, I. Tarr, M J. Anderson, A W. Skudlarski, P and Gore, J C (1999) “Activation of the Middle Fusiform ‘Face Area’ Increases with Expertise in Recognising Novel Objects,” Nature Neuroscience 2: 568-73.
[5] Topffer, R (1965) Essay on Physiognomy. Lincoln: University of Nebraska Press.
[6] Bodamer, J (1947) Paper on prosopagnosia.
[7], [11], [14], [17], [18], [19], [20], [21] Todorov, A (2017) Face Value. Princeton University Press.
[8] Works of Aristotle: Volume VI. (1913) Trans. W. D. Ross. Oxford University Press.
[9] Blackford & Newcombe (1917) The Job, The Man, The Boss. New York: Doubleday, Page and Co.
[10] Todorov, A. Pakrashi, M and Oosterhof, N N (2009) ‘Evaluating Faces on Trustworthiness After Minimal Time Exposure” Social Cognition 27: 813-3.
[12] Lewis, M (2004) Moneyball: The Art of Winning an Unfair Game. W. W. Norton & Company.
[13] Ross, L & Nisbett, L. E. (1991) The Person and the Situation: Perspectives on Social Psychology. New York: McGraw-Hill.
[16] McArthur, L Z & K Apatow (1983/4). Impressions of Baby-faced adults. Social Cognition, 2, 315-42.
Comments