« Not A Girl, Yet Not A Woman | Gene Expression Front Page | Hell No, We Won't Go! We're Having Kids! »
January 12, 2003

The Carnival of Race

A month ago a study came out that showed that most white Brazilians had a large number of non-white ancestors (over half their female progenitors in fact). This was shown as proof by some that race is a social illusion. But many ignored this quote:

"First of all, let me stress that the conclusions of this study apply to Brazil, and should not be naively extrapolated to other countries," Dr. Sergio D. J. Pena, the study's lead author, told Reuters Health in an interview.

And almost simultaneously, results were produced by other scientists that indicated that there were five geographical groups of human beings-races.

How do we resolve this? The short answer-it's complicated, and obvious answers are not always correct.

Brazilians are obviously a racial mix-and this mixing has been happening for 300 years. It seems that certain genes associated with selected phenotypes have been assorted over the generations to various segments of the populace to produce the illusion of relative whiteness, despite the underlying genome still attesting to a large quotient of non-white ancestry.

Say what? OK, assume you have two individuals, half-Swedish and half-Nigerian, who marry and have children. Let us say that the grandparents who were Swedes were blue-eyed, and the Nigerians brown-eyed. In the most simple explication of mendelian inheritance, the two individuals of mixed-race would have brown eyes but still carry a blue eyed gene. 1/4 (statistically) of their children would be homozygous for blue eyes, and so would display that phenotype. In fact it seems not implausible that if these two individuals had 10 children some of these ten would look more "white" and others would be more "black" in appearance, while the balance would be somewhere in between.

The ones who looked more "white" expressed more of their Swedish grandparent's genes for easily identifiable phenotypes (for instance, blue eyes, or possibly blonde hair). The opposite would be true for those who looked more black. In Brazilian society people who could "pass" as white but were genetically considerably non-white still might have mostly European genes for phenotypes that are traditionally markers of what it means to be white-light hair, eyes or wavy rather than kinky hair, etc. The light hair and eyes are recessive, as I believe is non-kinky hair (this is up for debate, lots of these traits are polymorphic, with multiple genes contributing to a phenotype), so would "breed true" and their children with other "white looking" people would also be white despite their considerable non-white ancestry (still evident in the Y chromosome and mt-DNA as well as many other genes that are coding for processes not noticed with the eye but might be different between races-for instance, the difference in processing carbohydrates that leads to obesity in black women).

Brazil-where race mixture has been much more common historically than in most parts of the world, is a particular case. For most Europeans in Sweden, their phenotype is a good signature of their overall genome, while most Nigerians are Nigerian, explaining why though there are specific exceptions to mapping phenotype to overall genotype (Brazil, individuals of mixed race), it is still generally a good rule of thumb.

But the key is that it is a rule of thumb. This is why I reject the almost religious appeals to racial solidarity made by race activists, whether white, black, brown or yellow. If someone has many non-white ancestors, but are physically white, are they part of the "racial soul"? But what if someone has white ancestors, but by some chance of nature is more racially ambiguous in appearence:

(Bjork of Iceland)

My final point would be this: many Brazilian whites carry genes that might code for phenotypes that we don't think of when we think "race." I believe elements of behavior are strongly influenced by genotype, and if these factors were not assorted by Brazilian social norms ("good hair" = not kinky, "good features" = European, etc. are easy to see, r vs. K selection a la Rushton's rule is not), than Brazilian whites might not act how we believe whites would act based on adaptation to Ice Age conditions in Europe, but more like their African or Amerindian forbears.

Update: Juan points me to this article debunking the Brazilian study's claims. Not totally convinced-the Brazilian study still surprised me even though I do agree the authors overplayed their hand. But, it makes the same point that I just did above, that it is particular to Brazil and shouldn't be generalized.

Posted by razib at 02:33 PM

What the hell is Bjork anyway? She's always puzzled me. Her ethnicity, I mean.

Posted by: Diana at January 12, 2003 02:57 PM

there are rumors that she's half chinese or eskimo, but i've read several interviews where she dispells those rumors-here parent are icelandic. her schoolmates made fun of her and called her "china girl" though.

Posted by: razib at January 12, 2003 03:51 PM

I think there was some mention of this study when it came out and the ususal races don't exist exclamations. However, I think another study came out at the same time showing actual clustering of groups to fit races that got even more press. The details are a little vague but I think Wade discussed this other study in NYT.

Posted by: -R at January 12, 2003 06:37 PM

yes, wade did. the second link is about the second study.

Posted by: razib at January 12, 2003 07:37 PM

I agree that sexual selection in Brazil (the fair sister marries the rich white guy, while her dark sister marries the poor black guy) has had some effect delinking genes for appearance from other genes. Still, I don't think this study will prove to be the last word on Brazil. I think we'll see that appearance and ancestry is fairly well connected, even in Brazil. We know that in the Northeast where lots of slave were imported to work in the sugar fields and few whites immigrated, the people look awfully African. And in the far south of Brazil, where lots of Europeans immigrated to about 100 years ago, the population looks a lot more European.

Posted by: Steve Sailer at January 12, 2003 09:42 PM

seems like whites from extremely cold lands would start to converge with east asians in appearance, right? that is, the climate would force their genes toward the same solutions, so that their features are similar?

Posted by: jody at January 12, 2003 10:02 PM

jody, that presumes that there is only one adaptation for any given environmental context. for instance, swedes and mongols live in cold climates (one more humid than the other of course), but swedes are larger and rangier. mongols simply changed their proportions to conserve body heat and reduce exposure (shorter arms and legs, longer torso)-while swedes got larger so that volume:surface area ratio increased.

just a thought.

also-sexual selection probably plays a large role in differentiating between races.

Posted by: razib at January 12, 2003 10:09 PM

This "study" about Brazil is a lot of Bullshit, yesterday I finish translating to Spanish a good rebuttal of this study written by Michael Rienzi in Legion Europa (I dont like very much the site but it has a good article about Brazil), you can see the refutación here:


Posted by: Juan ascaño at January 13, 2003 03:18 AM

I mean "refutation"

Posted by: Juan ascaño at January 13, 2003 07:12 AM

If you want to read it, this link have the complete brazilian study:



Posted by: Juan ascaño at January 13, 2003 01:41 PM

I've long been puzzled by Bjork's appearance : I also assumed that she was half greenlandic or something like that, but then i saw some pictures of norwegian saami people which explained a lot. It's quite possible that among the norwegian settlers 1000 years ago there were some people of saami stock. This would be a likely origin for Bjork's features.

Posted by: ogunsiron at April 8, 2003 12:20 AM

well.. i think the most logic would be Eskimo or part-Eskimo at least.. Eskimo's were the original peoples of Iceland.

Posted by: Gaz at May 20, 2003 05:01 PM