Many years ago whilst researching my PhD I stumbled on a meta-analysis of cross-cultural studies on facial attractiveness by Langlois et al. They included all the available studies that had been conducted to date on facial attractiveness preferences across cultures (i.e. studies where researchers had shown different kinds of faces to different people across cultures and asked them to rate which ones they found more attractive). They came to the conclusion that attractiveness preferences are “universal”, i.e. that different cultures all across the world seemed to find certain kinds of faces more attractive than others, and these faces were the same whether you were in the UK, Japan, or in the Amazon basin. This was then used to justify the concept of a universal human psychology of attractiveness that is innate in the human mind. Some people are just ugly, some are just attractive, and what determines that has been shaped by hundreds of thousands of years of human evolution, not by the culture you grow up in.
The idea here is that if people all across the world find similar faces attractive, then the best explanation for it is an underlying universal psychology. After all, what are the chances that so many cultures so separate from one another would all just so happen to coalesce on the same standards of facial beauty by accident? Given the sheer number of people and cultures, this would be extraordinarily unlikely. Some sort of universal underlying psychological similarity is far more likely, the thinking goes.
After a brief read through their paper, I was close to conceding that their conclusion was probably (at least partially) correct. The studies they had chosen really did seem to show quite striking agreement across cultures. I had a sneaking suspicion something was off, however, so I did some digging.
I first noticed that one of the studies included in the meta-analysis (Udry, 1965) probably shouldn’t have been included at all as it only compared preferences between American and British people. Yeah, it’s still technically “cross-cultural”, but only in a really trivial sense. The USA and the UK are really quite culturally similar. It’s not even really that surprising that most people in the USA and the UK agree on facial attractiveness. It’s kind of to be expected due to a shared cultural heritage and a close modern day political and social relationship.
The East-West divide
This is only one study, however. Most studies in the Langlois meta-analysis don’t compare such obviously similar cultures. Most of them compare Eastern Asian (Japanese/Chinese/Korean) and Western (North American/European/Australasian) preferences, including Bernstein et al (1982), Cunningham et al (1995), Madden & Hollingworth (1932), McArthur & Berry (1987), and Zebrowitz et al (1993).
There is a large amount of research showing that Westerners and Easterners really do have a large amount of pretty fundamental differences in the way they see the world. This isn’t particularly surprising; people in the East and West lived in relative isolation from one another for thousands of years until modern transport and mass media brought us all much closer together. If cultures this different from one another seem to have similarities in the kinds of faces they find attractive, doesn’t this mean that there must be something innate to all humans about the kinds of faces we find attractive that endures across very different cultures? How else can we explain it?
An alternative explanation: Westernisation
There may be another explanation. In one of the studies listed above – Madden & Hollingworth (1932) – pictures of 40 Caucasian adolescents were judged by 10 white and 10 Chinese judges for physical attractiveness, or “good looks” broadly speaking. What’s interesting is that this is one of the only negative studies in the Langlois meta-analysis: there was very little agreement between the white and Chinese judges. They had different attractiveness preferences. This study is also particularly interesting because of its age – 1932! At this time, the majority of people in the West and the East would never have even seen an Eastern or Western person respectively. Cultural transmission between these two sides of the world would have been almost non-existent.
What’s the main difference between 1932 and today? We now have globalised mass media – with a predominantly Western basis – coupled with the affordability of long-distance transportation, the relative influx of Westerners to the East and vice-versa, and the recent dominance of the internet (a highly globalised data exchange). Given that the West has been dominant in all of this, it is highly likely that there has been a massive amount of Western-led cultural transmission from the West to the rest of the world. For this reason, it’s pretty reasonable to suggest that facial attractiveness preferences could have been Westernised.
How Westernisation works
Note that this isn’t just speculation; there’s actual evidence of this happening. Yu and Shepard (1998) studied the waist-to-hip ratio preferences of two populations of the Matsigenka indigenous people in Peru who were both first contacted by the West 20-30 years prior to the study. By all accounts, they were a relatively recently contacted people, and so shouldn’t have experienced much cultural change since contact.
The first group lived in the area called Yomybato and were the most culturally isolated. They said that “overweight” women with a higher waist-hip-ratio and larger body size were more attractive, healthier, and more desirable as spouses, which was – perhaps unsurprisingly – the exact opposite of the US control population. The second group lived in an area called Shipetiari and had a lot more exposure to Western people and Western media over the past 20-30 years. Some aspects of their preferences were the same as the group in Yomybato: they considered overweight women to be healthier, interestingly enough. For the other two dimensions (attractiveness and desirability as a spouse) they preferred the thinner women with lower waist-to-hip ratios. For these people, the concepts of healthiness and attractiveness were becoming separated from one another, and their opinions regarding attractiveness (but not healthiness) were more in-line with the typical Western preferences. Amazingly, after only 20-30 years of intermittent exposure to the West, their body preferences were changing to reflect the Western standard.
What does this mean for cross-cultural research?
This seems like a pretty difficult problem for cross-cultural research. Isolated, small-scale societies are arguably the best source of cross-cultural data because of their very limited exposure to the West (or, for that matter, any other cultures). Yu and Shepard show that it is possible for very little exposure to the West to have quite a striking impact on the preferences of people.
Basically, if a cross-cultural study finds a surprising similarity in some preference (say, facial attractiveness) between apparently very different cultures (say, the East and the West), it is very difficult to isolate the cause of this. It might be something to do with a universal human psychology, or it could equally be due to exposure to the West. Which conclusion is more likely to be correct? There’s no way to give a confident answer to that. Still, if you see any research about “surprising” cross-cultural similarities and any grand claims being made based on that, bear in mind everything I’ve said above.