However, this is not the case for three reasons: people traditionally clustered in a certain race are not at all homogeneous (read: variation in hair color among Europeans). Indeed, genetics have confirmed that there is far more genetic diversity within groups than between groups. The second reason there isn't considerable homogeneity between groups of people and conspicuous differences between groups is that, since our ancestors spread out of Africa, different human populations have always been in genetic contact with each other, constantly 'mixing' genes and making conspicuous difference between groups impossible. Third, there are a number of 'unclassifiable' peoples (darkly pigmented Indians with European-like faces), which, by their very nature, dispel the idea that there are a specific number of races in the world. By the end of the article, Marks concludes, as he says in the beginning, that race is not a biological factor, but a cultural one. While the flow of the piece,...

His message is relatively simple, and, with the help of modern genetics, relatively easy to prove. Of particular interest, for me, was his observation that racial violence -- between biologically indistinguishable races (i.e. Hutus and Tutsis) -- that uses 'race' as a cause for the violence is simply using an excuse.
Perhaps the only thing missing from Marks' argument was the case of the other side -- an explanation of the genetic differences between, say, African-Americans and Europeans. I was left wondering, often, how he was going to treat that difference (because there must be some, if at least superficial, biological difference), but it never seemed to be addressed.

References

Marks, Jonathan. "Black, White, Other: Racial Categories are Cultural Constructs Masquerading

as Biology."