Last week I argued that the impulse to make assumptions about people based on how they look, or the groups to which they belong, is not always irrational. A "rational" generalization is one rooted in empirical evidence. As insurance companies know, young unmarried men are more likely to be involved in car accidents than other drivers.

Reader reaction was interesting. No one denied the predictive value of demographic profiling, though I did detect a plea for wilful ignorance: Yes, African-American men in the United States represent 12 per cent of the U.S. population and 47 per cent of those arrested for murder -- but can't we pretend that crime is colourblind? Is there some knowledge not worth having?

The article was about Bruce Lahn, an emerging star in genetics research. Raised in China, Mr. Lahn is not yet 40 and already a full professor at the University of Chicago.

Lately, controversy has swirled around him. Mr. Lahn discovered that the human brain has undergone evolutionary changes -- that it has grown bigger and more complex -- in a relatively recent time frame. One such "mutation" is only 6,000 years old, corresponding to the time that humans learned to build cities and to develop written language.

Trouble is, certain populations, mainly in Africa, especially sub-Saharan Africa, appear not to have experienced the same mutation. The crude implication is that African brains are not as evolved as non-African brains. As the Wall Street Journal reports, Mr. Lahn and his colleagues had serious internal discussions about the consequences of their discovery. They feared racists would exploit the findings. Nonetheless, they went ahead and published the papers in the prestigious journal Science. Sure enough, racist groups were soon celebrating the research on their websites.

The University of Chicago seemed to take the position that scientific knowledge is neutral. ("Don't worry about the implications," the medical dean told an anxious Mr. Lahn.) The argument seems to be that ethics enter the picture only later on, when it comes to deciding what to do with that knowledge, after the scientist has left the scene.

This argument has merit -- to a point. Canada's aboriginal people are often stereotyped as socially dysfunctional, prone to drug abuse and disease. Is it unethical to investigate whether this generalization is rooted in statistical fact? Not according to Health Canada, which has published data showing, for example, that the incidence of genital chlamydia is seven times higher in First Nations communities than in the general population.

Although this knowledge might well be used unethically by bigots seeking to denigrate First Nations people, it can also be used ethically by aboriginal leaders seeking to understand the pathologies plaguing their communities in order to fix them.

For a long time scientists avoided research on the links between race and disease, especially diseases tied to biology rather than social circumstance. Evidence of innate group differences challenges our belief that humans are made of the same stuff. Eventually, however, the clinical data became overwhelming. We know that sickle cell anemia is more common among blacks, that the breast cancer genes BRCA1 and BRCA2 disproportionately afflict Jewish women and that certain iron disorders tend to surface in Swedes. The benefit of knowing these connections, as measured in preventive medicine, outweighs the social harm of drawing attention to group differences.

This is cache, read story here