This got me thinking about how reliable these "they look they're from..." judgements are.
Clearly on a basic level, we can usually tell which continent someone's ancestors were from, in terms of the familiar "races" of Europeans, Africans, East Asians etc. But what about shorter distances?
Could you tell, just from looking at them (and setting aside dress, hairstyle, jewellery etc.) whether someone was from Spain as opposed to France? Korea or Japan? Russia or Germany?
I can only speak for England, but there's certainly a vague but widespread belief that every part of Europe has a distinct 'look'. In the past, people were very fond of talking about that kind of thing; today, we're rather embarrassed by the idea but the belief lives on.
I don't know, but I'd be very surprised if there weren't analogous beliefs in other countries.
But how accurate are these folk beliefs, really?
Supposing you were the world expert on human faces - or suppose you were a supercomputer with face-recognition software and access to Facebook's entire dataset. How accurately could you place someone's origins on the map, on average? To within 1000 km? 100? With what degree of accuracy? In an ideal world, could the ultimate face-placer judge someone as French vs German 75% of the time? 90%? Or only slightly better than chance?
I suspect that if you researched this, you'd find that a supercomputer could do very well, in most parts of the world, but that the majority of actual people are less accurate than they think they are.
17 comments:
My guess would be that we supplement visual recognition with a lot of contextual information about who should be where. Note that would give better results than a supercomputer doing visual recognition alone.
Let me get out of my lurk-mode because I do some work on this. There is much social psych work on this: people tend to lump faces from outgroups together ('they all look a like', called outgroup homogeneity) and they use their stereotypes to guide face categorization. So probably, people's judgments are not very accurate. Whether a machine learning algorithm would do better is, I guess, an empirical question, but it would probably have problems handling faces from multi-ethnic cities in these globalized times.
there are definitely differences: http://thenextweb.com/shareables/2011/02/11/research-creates-the-average-face-for-countries-across-the-globe/
so it's a question about individual variance. Because you categorized a whole group of people, you are more likely to have categorized them correctly.
I for one am always confused for a local person or at least with family ties to the many countries I have visited. Just recently, I spent two months trying to convince Sri Lankans that I am not in any way related to them or other South Asians for that matter. To no avail. It was the same in Indonesia, the Philippines, etc.
My parents who were in Cuba a few months back faced the same situation while my sister was taken for a prostitute servicing her white boyfriend during their visit there. And also in Brazil, actually.
Meanwhile, people from the horn of Africa, where we are originally from, also confuse us for being part of their own ethnic group. My mother and sister are often accused by Somalis of being snobbish/haughty when they deny the connection. I usually get Ethiopian and less often Northern Sudanese.
While it can all be quite annoying, I like the advantages it brings me while travelling...at least until I open my mouth.
People all look the same. One big blur. I tell people I'm black. What does it matter? after a long sojourn consuming and crapping we die.
Hauke: I hadn't seen that, cheers. You're right that the big question is individual variability, and whether individuals vary on those dimensions that define their origin.
Those averaged images do indeed "look like" people from the countries in question - but then think of real people we know from those countries, they "look" very different to the average.
So yeah it's a question of effect sizes.
Ron Dotsch - Yeah, urbanization and migration would throw this off. You'd have to stick to native populations. And I it would probably just fail in say Brazil or the USA.
The French president Nicolas Sarkosy 's has coined recently the term :"The religions of Faces".
He was publicly speaking of a catholic young elite corps French military personel and French national killed by a mass murderer, Mohamed Mearah in Montauban.
The president of France said that the victim had the appearance of a muslim -he thought it politically correct not to say that he looked from Northern Africa.
PS: There would be much to say also about "The geography of foreign accents" and "The sociology of way of speaking".
Food for thought in that "superficially trivial" post Neuroskeptic!Thanks.
Hello,
There is much of the social work...
and the peoples face together... after all that is great post.....
Regarding Hauke's link and neuroskeptic's response: I think the component no taken into account is facial characteristic grouping (some groups of characteristics can lead to certainty while others lead to uncertainty). Inheritance of averaged genes doesn't necessarily lead to averaged faces. It is possible that some characteristics can make you >95% positive about some people and worse than chance for others
I would really like to know if we distinguish between Japanese and Korean (for example)? It seems like an easy experiment. Have people try to determine the race of an assortment of people. You could then determine IF it is possible for people to do and WHAT faces are more easily identified.
As for the Horn of Africa - isn't there a great deal of genetic mixing along all the shores of the Indian Ocean? People have been sailing and trading from Sri Lanka to Zanzibar, and further, for centuries.
Don Cox
Dienekes Anthropology blog does surveys of this type for his readers every now and again,and there are several other physical anthropology bloggers that do as well. Some people have better track records at getting the right answers than others. There are published studies in the literature on this but, I'm not very familiar with it.
Ancestry informative markers from DNA testing in Europe can make predictions that are likely to be correct at scales on the order of something between the geographic size of an American country (typically a few hundred sq. miles) or even township (36 sq. miles). Appearanced based estimates necessarily draw from a narrower set of ancestry informative markers, so ought to be less predictive.
Environmental impacts particular to locality exist (e.g. certain isotype ratios) and some influence appearance. Someone who grows up someplace where you tan as deep as you can every season and who eats mostly fish, will look different in midlife than someone who always stayed out of the sun and eaten ham and beef all their life, even if they were both born to people who had lived in the same village in Europe for generations.
Some people due to greater experience are better at this parlor game than others, and the ability to infer geographical origin from appearance is usually region specific. I know people who have lived in East Asia who state that they can distinguish visually between someone from Southern China and someone from Northern China on sight. I know people from South Asia who claim that they can make fairly local ancestry distinctions within that region. I couldn't reliably do either. Given the accounts I've heard from people claiming to have these abilities, I'd guess that fairly accurate determinations are possible at a scale on the order of areas the size of U.S. states (thousands or tens of thousands of square miles), with some areas being large than others due to variations in population homogenity.
"I would really like to know if we distinguish between Japanese and Korean (for example)?"
For someone who has lots of Korean relatives by marriage, like me, this is easy. I would never mistake a full blooded Korean for a full blooded Japanese person (although a full blooded Northern Chinese person, perhaps a Manchu, might be trickier to distinguish from a Korean). But, I probably wouldn't have been able to make that distinction accurately when I was in high school in a predominantly white small town, before I met my wife.
A little context helps as well. In Denver, the largest black populations are recent immigrant Ethiopians, and African-Americans with origins in the antebellum South, who look different in a reasonably distinguishable way if you're familiar with both. But, in New York City, where there are a lot more reasonably likely possibilities, a good guess is harder.
Between adornment and pure physical appearance, there is also body language. A woman raised in Algeria doesn't have the same body language as one raised in Saint Louis, even if all of their ancestors all come from the same village. Someone raised in Japan has instinctive body language like bowing in certain circumstances that a full blooded Japanese person raised in San Francisco does not. Even the kind of smile or lack thereof someone has an a photograph can be distinctive body language.
A geographically specific look also isn't constant over time. A Japanese or Korean person just reaching full adult height in 2012 is going to be much taller than someone in his grandparents' generation and will also look different in other ways that I would be pressed to articulate specifically. My Swede-Finn relatives in Finland today (a few of whom I've met) look quite a bit different than my ancestors did in the generation when they immigrated to the U.S. (among other things a characteristic gauntness is absent now).
Andrew,
Thanks for an intersting comment.
And what became Korean in your body language after living with and loving a Korean woman for long?
"And what became Korean in your body language after living with and loving a Korean woman for long?"
Not much, as my wife, while having Korean born parents grew up in the U.S. in an area where the community wasn't large enough to cocoon in. But the differences between her, and her more insulated cousins in L.A., or her Korean born cousins is noticeable, e.g., in the level of formality and consciousness of how one moves.
I'm just incorrigable and untrainable, like most husbands.
There's a site called AllLookSame [uh, I didn't name it] that has a test to see if you can distinguish Chinese, Korean, and Japanese faces.
http://alllooksame.com/exam_room.php
Amos: Heh, thanks for the link, that's interesting :D
I really appreciate your effort for sharing a great topic it's wonderful. Thank you for sharing.
Post a Comment