Racism in science

December 25th, 2002 Comments Off on Racism in science

In a visitor‘s blog, I found a pointer to a The New York Times article about a study that supposedly identified five main human populations, based on scans of the human genome. According the scientists that made the study, those five population groups broadly correspond with popular notions of race. I think we already have sufficient trouble fighting racism, and now a study is claiming that races do really exist.

I see some problems with the conclusions the study reached. I don’t know how well the article presents those conclusions, but I will try to question some of the results below.

Firstly, the study found patterns in the DNA that correspond to five major geographical areas. From this, the scientists jumped to the conclusion that each of this groups is a race. I can understand what it’s implicit here, but it simply doesn’t follow. From what we know of history and biology, it’s obvious that cohesive population groups would slowly develop common traits across the centuries. Natural selection, genetic drift and social constraints can account for those common characteristics. Nonetheless, classifying those similarities as properties of “race” is just a semantic construct. Take the Americas, for instance. Their population is a result of heavy intermixing by people of Asian, European and African descent. But once the population stabilized, the common traits aforementioned started developing. As the study identified the Americas as one of the major population groups, or race, we would have to conclude that races are not static patterns, since the America’s population is composed of people of every conceivable descent. That fact alone would invalidate the conclusion that those groups are races.

Secondly, the study claims that, by using “self-reported” ancestry, it would be easier to identify the possibility of occurrence of some diseases, like sickle cell anemia in African populations. Sickle cell anemia is common in Africa because its carriers are immune to malaria, another common cause of death there. The rates of death by malaria and sickle cell anemia balance around 18%, because of natural selection. So, even with heavy mortality rates, sickle cell anemia is preserved in the population when it would tend to naturally select itself out of the gene pool. We can see that the environment is a factor in the genetic development of the population. In other cases, like the Amish population in the USA, the common traits developed are the result of intermarrying, driven by social constraints. It would be dangerous to presume too much on the basis of a broad generalization like the study proposes. More so if we consider two additional facts: genetic drift happens very fast in small populations and population mobility is higher now than in past centuries.

Thirdly, the articles says the markers analyzed have no particular function, and are free to change. I’d say they have no know function and nobody knows how changes in them would affect an individual. It’s unsafe to make such assumption and base conclusions on it.

Fourthly, it has been demonstrated that even among phenotypically related populations, there can be enough genotypical differences to make their components seem like they belong to far removed biological groups.

There are other points I could consider, but the article doesn’t give enough information to allow comments.

In short, even if the study conclusions are valid (considering the points above), it’s unfortunate that those scientists decided to associate the genome variations within the human species to races. There are a lot of people who would be delighted to have their prejudices about race confirmed in that way. To the credit of the article, it points the ethic problems with the study, and challenges the notions presented. I just wish scientists would be more careful when dealing with science where it touches the social problems of our world.

Comments are closed.

What's this?

You are currently reading Racism in science at Reflective Surface.

meta