The Use of Racial, Ethnic, and Ancestral Categories in Human Genetics Research. You can find the PDF in gnxp files in the forum (I'm not trying to make you register, but I am noticing that google is sending a lot of people to the few PDFs I'm saving on our servers and they obviously don't see the rest of the weblog and just gobble up bandwidth). There is not much that will be new to GNXP readers, the first half of the paper ranges over the biological aspects of the topic, going from palaeoanthropology to population genetics. The literature review is pretty deep. A pretty thorough overview of the current "Out of Africa" orthodoxy is covered, but some space is given to dissenters who assert there might be archaic autosomal sequences in the genomes of some populations, which surprised me (John and Henry were both cited, but Lynn Jorde of the U of Utah was a part of the group that came up with the paper, so perhaps I shouldn't be that surprised). Eventually they bring up the 85-15 within group-between group variance on a single locus that Richard Lewontin popularized, but they don't touch upon the fallacious aspect of that statistic. Nevertheless, there are literature pointers to Neil Risch, so one can find the other viewpoints if inclined (the precision of identification as one increases the number of loci brought into the analysis, etc, etc.). The second half of the article on social and ethical aspects of the issue was less interesting, and seemed rather platitudinous, but then that might be my own bias in what aspects of a topic generally engage me.