Sunday, August 24, 2008

The Myth of the Myth of Race

Over at Greg Laden's blog, we've gotten into one of those discussions on race that tend to crop up whenever a large sporting event like the olympics is going on. Some people start making racist comments about the natural ability of black athletes, as if one could be a world class sprinter merely by walking onto the track with such superior genes, and there is much merit to the position that blacks usually get short shrift in these conversations. Black athletes are described as natural athletes, whereas the white winners are often credited with their hard work. Black athletes are described as "articulate", white athletes are "disciplined". Things are improving, but we still have much work to do.

However, the arguments that there is no racial component to performance in sports at all, or worse yet, that race is merely a social convention, are nonsense, a classic example of political and social bias trumping science. These arguments make many logical mistakes, the most prominent is an ironic black/white view that attempts to paint all arguments as "100% of this group has trait A", straw men easy to knock down. However, the real argument made by those of us who believe there is a racial component, and who would not register as a racist by any other measure, are much more complicated and use sound statistical analytical principles.

Take the position of NFL cornerback. For you nonyanks, this is the position whose job it is to keep players from the other team from catching passes. It is the football position for which foot speed is most essential. And while the NFL is very racially diverse, this position is completely dominated by blacks. The question is: Why? The answer: the exaggerated effect of small changes in mean ability at the extreme end of the curve.

Take all the Americans that 99% of Americans would categorize as either "white" or "black" and separate them (Greg and his ilk can come up with whatever names offend them least). Then imagine a bell curve graph of their average "natural foot speed". Now if there was no genetic advantage, aside from the difference in population sizes, we'd just see two overlapping graphs. However, if one population has just a tiny genetic difference that made a difference in the ability to run fast, say slightly longer limbs, lower body fat, or some change in muscle structure, it would shift their curve a bit. This shift would hardly be noticeable when viewing the entire graph, and indeed, it would be unfair to make a generalization that one group was faster than the other. However, at the extreme end, the difference would be dramatic. And professional cornerbacks, like 100 meter dash runners, come from that extreme end.

The same argument could be made as to why certain racial groups dominate swimming, wrestling, gymnastics, or a whole host of other events. Tiny differences in body structure can make all the difference in elite athletic competition, and these tiny differences exist in all sorts of social/racial groups. This is why we have never seen, nor will we ever see, a Kenyan heavyweight weightlifting champion, or a Swedish 100 meter dash champion. That is not to say there are no fast Swedes or big strong Kenyans. Of course there are, and their abilities are no doubt impressive when compared to the norm. But that doesn't cut it at the elite level.

What do those who oppose this theory say? The contention is often made that "the variation within races is less than the variation between races", which is true but unimportant. I translate that to mean that the range within a race, say the 95% confidence interval, is greater than the difference between the mean of the races. Of course it is. This is why saying "Joe is surely going to be faster than Bob, because Joe is black" puts you on shaky ground. The difference in the means is just not that big. But for my argument to work it doesn't have to be. A tiny difference is enough when you are discussing a handful of people out of a population of millions.

Some claim reflexively that somehow it is the environment that is responsible, leaving the pathetic details of how exactly this was the case to someone else to figure out. This is clearly a political/ideaological statement, not a scientific one. They want it to be about the environment, because they find the idea of genetic differences offensive, and potentially socially problematic. Tough. Reality is what it is, and as with the criticisms of evolution claiming it leads to Social Darwinism, the dangers of acknowledging the tiny genetic differences among human groups are also greatly exaggerated.

Some claim that a racist bias causes the Cornerback Effect. Racist coaches, the theory says, assume black kids are stupid, and steer them into speed positions like cornerback, and outfield in baseball, while steering white kids racistly assumed to be intellectually superior, into thinking positions like quarterback and catcher. However, this theory completely breaks down when tested in basketball, where the position that demands speed, point guard, ALSO demands intellect. Black guards are far more prominent than black centers in the NBA. There is a bias all right, but it is FOR speed, not against intellect.

And then there are those who claim the concept of race is a mere social convention. When confronted with the objective data however, they either deny it outright, change the subject, or of course, accuse people who recognize the genetics as racists. They are simply denying data in lieu of theory, intellectually in the same camp as the good Christians who claim atheists cannot have morals, while ignoring the actual data of real atheists living real moral lives.

Racial groupings, however rough, are readily identifiable, and produce statistically significant deviations from what we'd expect were the entire concept of race fraudulent in some way. The fact that an overwhelming majority of NFL cornerbacks and 100 meter dash finalists are of West African descent and can be independently identified as "black" demands a scientific explanation. Denying reality through philosophical games that try to pretend races don't exist are no more effective in explaining anything than was Berkley, the King of Dodges, when he claimed there was no external objective reality.

These silly denials also damage science in the public eye, giving fuel to those who claim scientists are out of touch with the real world. We can't win the creation/evolution battle, and persuade so many people that what their minister and Bible tell them are scientifically wrong, if they see us claiming that "blacks" and "whites" are merely social conventions, and that what they see with their own eyes isn't real.


Harriet said...

Interesting post. But I'd also like to offer the following:

1. There have been white sprint champions; Warner (USA) is one of them and in 1972 a Russian won the 100 meter dash.

Also, did you see how close the Russian women came to winning gold in the 4 x 400 meter relay? The US lady won in the last 5 meters or so.

2. On the other hand, those who deny a black athlete credit for being a champion are being silly; after all they beat everyone including other blacks. :-)

3. It will be very interesting to see if Chinese track runners eventually excel in the sprints.

4. But to bolster your point, look at the dominance that Jamaica has had in the sprints and Kenya and Ethiopia have had at distance running.

Thanks for a thought provoking post.

Anonymous said...

Thanks for this post and your rebuttals to Greg Laden. I posted a reply to Greg's post about fast twitch and slow twitch muscle fiber distributions between the races and he slapped me down with the 'race is a myth' idea. I am not well read enough in Biology to push my point further, so I am glad you did. He then double-backed me with the "If you were doing real science, you would get fired for using such categories as sampling units" line, but I have read *real* science, especially in the evidence based medicine fields, particularly target studies to certain races to develop better treatments for those particular affected races.

ScienceAvenger said...

Since the statistical edge would be small between the groups, we'd expect some outlier results, although I have considerable skepticism concerning the doping level of many of the Soviet Bloc athletes.

As to the vitriol we get in resonse to raising these fairly straightforward points, that's typical of left-wingers when their psedoscience is called out. You can see similar reactions when their bogus gun death analyses are criticized.

Nada Platonico said...

In 2004 a woman by the name of Yulia Nesterenko from Belarus won the 100m.

You do know that Liu Xiang, the Chinese hurdler, won Olympic gold at the 110m hurdles in 2004 (equaling the world record time; he would later break it -- his record stood until this year, when it was broken by a Cuban).

The hammer throw in 2004 was a Japanese thrower.

And the decathlon -- the event that determines "the world's best athlete"; it combines sprints, throws, jumps, and a middle-distance event (1500m) was won by Roman Sebrle of the Czech Republic?

Very few African women (actually, only two) have ever held the world record in the marathon (one broke her own record).

Russian women have dominated the 1500m in recent years (not a good showing at the Olympics though, although a couple Ukranian women finished second and third.

I have a post coming out tomorrow on the topic; I agree that race is more than a social construct. I am not convinced (as you can tell, I'm sure) that race is a proper category for this.

My blog's address is:

ScienceAvenger said...

All interesting facts, and completely consistent with my thesis. Women's athletics in particular is much more highly influenced by politics and social factors than are the men. After all, all the natural ability in the world does one no good if women are not allowed to compete in one's country.