Claims re Autosomal DNA and Jewish Origins in Khazaria

[This page was written in about 2014 and has not been updated since then.]

In 2012, Eran Elhaik published an article arguing that autosomal DNA analysis demonstrated that Eastern European Jews are descended from Khazars who converted to Judaism in the 8th century and moved to Eastern Europe following the collapse of the Khazarian empire.

In 2013, Doron Behar et al. published an article that criticized Elhaik's article on the ground that, inter alia, the population selected by Elhaik was not representative of the Khazar population of a millennium ago. Behar et al. concluded that Ashkenazi Jews share the greatest genetic ancestry with other Jewish populations and, among non-Jewish populations, with populations from Europe and the Middle East; they found no indication of Khazarian descent. In 2014, Elhaik posted a response to the article by Behar et al.

In 2014, Shaul Stampfer published a critique challenging Elhaik's article on both scientific and historical grounds.

In 2013 Marta Costa et al. published an article on Ashkenazi mtDNA, concluding that the direct female ancestors of most Ashkenazi Jews - including the four major founders who are the direct female ancestors of 40% of the Ashkenazi population - had ancestry in prehistoric Europe rather than in the Near East or the Caucasus.

Anatole Klyosov has provided this website with an open letter commenting on Stampfer's article and, indirectly, on Elhaik's article:

* * *

I can only regret that my good friend Eran Elhaik has published said paper. It is a dramatic illustration of a level of misunderstanding by many (most?) geneticists and in particular population geneticists in DNA genealogy, and DNA genealogy was essentially the gist of Eran's paper.

Let me explain. There are many bearers of haplogroups J1, J2, G2a, R1a, R1b, etc. on the Caucasus, and of the same haplogroups among Ashkenazim. In other words, both the Caucasians (who commonly have nothing to do with Jews - by religion, ethnicity, etc.) and Ashkenazim have common ancestors, who lived some 15,000 to 20,000 years ago. This has nothing to do with Khazars. However, those "common ancestors," or rather their ancient specific mutations (SNPs), are "seen" in the DNA. There is nothing surprising in this, since we see tons of SNPs in our DNA from our common ancestors with chimpanzees, for example. Now, geneticists see the same mutations (SNPs) in Caucasians and Jews (not necessarily in Ashkenazim, but in most of the Jews) and say - incredibly! - that Ashkenazim are descendants of the Caucasians, hence, they are Khazars(!). This is essentially the methodology of Eran's paper.

It is, as I have said above, a typical mistake of geneticists. Recently it was found that excavated bones from the Baikal area (South Siberia), dated 24,000 years ago, showed haplogroup R, and also revealed a good deal of similarity of their DNA with that of Native Americans. An "interpretation" of this finding, published in Nature, Science, and other high-profile journals, was that it points at the migration of the Europeans to America (since it was taken that haplogroup R is a European haplogroup). However, lack of knowledge in DNA genealogy prevented the authors from realizing that South Siberian haplogroup P (which arose some 40,000 years ago) produced haplogroups Q and R, that later bearers of haplogroup Q for their part left for America, and that haplogroup R along with its descendants R1, R1a, and R1b much later (about 10,000 to 15,000 years ago) left westward, and arrived to Europe about 5,000 years ago. As you see, haplogroup R left "left," while haplogroup Q left "right," but both shared a common ancestor, haplogroup P. That is why geneticists see similarities in the DNA (again, in SNPs) in Europeans, South Siberians, and Native Americans. Not because haplogroup R went to America; it never happened in those times, 10,000 to 30,000 years ago, or at least we do not see any evidence for it.

Another example, even more amazing, is that geneticists found that many European populations contain a certain fraction of the DNA (SNPs) of Native Americans. There have been published "data" that Irish people, for example, contain 11% of Native American DNA, and the Russians contain 12% of Native Americans, etc. As you already understood, this is because the Irish mainly contain R1b haplogroup, the Russians mainly contain R1a haplogroup, and Native Americans mainly contain Q haplogroup. All of them have, again, haplogroup P as their common ancestor, and geneticists see those SNPs in the Irish, tthe Russians, etc. And certainly it would be observed in Jewish bearers of R1a and R1b. The problem is also in that the geneticists rarely consider haplogroups in their "findings."

So, my friend Eran fell into the same trap.

"What happened here," Professor Stampfer asks, referring to reputable academic journals which actively publish those, well, ridiculously distorted "findings." Acrobatics is how I have defined this practice in some of my publications. In fact, it is a tragedy of present-day population genetics. There are literally dozens and maybe hundreds of trashy publications in academic literature in that area. They employ principally incorrect "population mutation rates," which produce inflated dates - increased by 250 to 350%! - for common ancestors.

As a result, they invent ridiculous "historical events," based on "obtained" wrong dates. They manipulate with the "Cohen Modal Haplotype," which in fact is often observed among Arabs with a common ancestor of about 9,000 years ago. They "calculate" a timespan to a common ancestor of the "CMH," using the same inflating "population mutation rate," and instead of the actual 1,000 years for the "CMH" common ancestor among the Jews of haplogroup J1, they (the creators of the "CMH") "obtain" for some 3,600 years ago, and they claim that this is historically correct data.

They claim that R1a haplogroup among the Jewish population came from Slavic men, while their haplotypes are principally different, and deviate by about 4,000 years. Which Slavs 4,000 years ago? Unfortunately, those papers are gladly published in leading academic journals. Why? Because of the rotten referee system in the area of population genetics, representing a "vicious circle." And I cannot see any means for the circle to be broken; it is too tight right now.

It is something to think about. However, a step was taken ahead by establishing an open access journal, "Advances in Anthropology," which publishes articles in the area of DNA genealogy, free from those wrong "population mutation rates" and "acrobatics" by population geneticists.

Anatole A. Klyosov

Boston