A Near Eastern Origin of the Indo-Europeans – An Internet Hoax? Observations on the Writings of a Confused Genome Blogger

Dienekes has been campaigning for a while for the Indo-European homeland in “West Asia” or the “Near East.” I usually don’t actively discuss the origin of Indo-Europeans and technically this question falls outside of the theme for this blog. But this time around I cannot help but expose the the rare species of nonsense that Dienekes’s claim contains. Razib first trumpeted Dienekes’s “finding” in the loud manner of a good old paperboy (“The Mystery of the Origin of the Indo-Europeans May Be Solved Within the Next Two Years”). In this post he proclaims that “within the next few years we may solve the ‘mystery’ of the demographic origin of Indo-European languages and culture through genetics.” A couple of days later, he, again, applauded Dienekes’s efforts by writing a post entitled “Has Dienekes Pontikos Found the Signature of the Indo-Europeans” in which he “wanted to give him full credit in the public record if researchers confirm his findings in the next few years.” GCochran commented obtusely to the effect that Dienekes is the Galileo of science.

Dienekes and Razib have been in the “You scratch my back, I scratch yours” mode for a while parading the open source approach to genetic samples and the software for their analysis. Amateurs armed with publicly available data and software, they are now eager to prove to professional researchers that they can think, too, and that they may even be able to solve age-old problems that were unsuccessfully tackled by generations of archaeologists, linguists and suchlike. One of these problems, Razib and Dienekes believe, is the origin of Indo-Europeans. As we will see below, however, the question of the origin of Indo-Europeans has been given a compelling answer 50 years ago and this theory is called ‘Kurgan theory.’ But wouldn’t it be great to be able to undo history, begin from a blank slate and wake up tomorrow to the revolutionary headline in the New York Times “A Genome Blogger Solves the Mystery of Indo-European Origins”? Not Gimbutas, not Mallory, but a nameless Greek blogger. This would be the ultimate triumph of genetics over the social sciences (since even a nameless blogger can become a successful “scientist” once he masters the basics of genetics) and of unschooled enthusiasm over academic expertise.

I firmly believe in the power of groupthink; that’s why I’m on the web sifting through pages of mostly confused and ignorant comments from “Nobody45” and other aliased science hobbyists. The labors of the army of nameless genome soldiers each savoring a handful of their own and their neighbor’s genes and then reaching up to the sky in an effort to make an offering to the God of Science will not be forgotten. But just like a machine cannot replicate all the complexity of human intelligence, groupthink cannot replace professionalism.

Dienekes begins his “song of the Near East” from after 16,000 years BP when the 6 populations that he detected at the K=12 level of ADMIXTURE analysis (Mediterranean, North_European, Caucasus, Gedrosia, Southwest_Asian, and Northwest_African) presumably began to emanate from a Near Eastern source driven outwards mostly by the soon to evolve agriculture. He identifies the Caucasus as the refugium that has preserved most faithfully the genetic and linguistic signature of that original population. It occupies the center of a MDS plot and harbors the highest levels of linguistic diversity in West Eurasia.

Once the Lalueza-Fox team published the partial genomes of two Mesolithic Iberians dated at 7,000 BP and determined that they clustered with modern Northern Europeans rather than with modern Southern European populations, Dienekes quickly concluded that “this appears quite consistent with my model of mostly recent origins of European populations from a West Asian womb of nations.” Meanwhile, nothing in the study gave support to this claim – it simply documented the discontinuity between Mesolithic and modern Europeans in Iberia, without indicating where modern Europeans came from. Plus the date for the break-up of the ancestral West Asian community given by Dienekes at 16,000 years BP is on the order of magnitude older than the post-Mesolithic settlement of Europe. About a year ago, in April 2011, in a post entitled “Indo-European origins: Neolithic Anatolia still the best hypothesis,” Dienekes praised the Gray and Atkinson statistical study that challenged the 5,000 YA timeframe for the dispersal of Indo-European languages from the Pontic steppe and supported an Anatolian homeland at a much earlier date, 7,800 and 9,800 years BP. But in the recent post, “The Bronze Age Indo-European invasion of Europe” Dienekes lowers his ambitions down to the Bronze Age and advances the thesis that Indo-European speakers originated from a West Asian source in the post-5kya timeframe, which is “conventionally accepted by linguists for either the dispersal of Indo-European languages, or at least of a significant subset thereof.” In 2012, Dienekes must have realized that the Gray and Atkinson’s approach is questionable, hence he swept their studies under the rug and reverted back to the consensus 5,000 YA timeframe for the spread of Indo-Europeans.

Dienekes has indeed circumscribed an interesting autosomal component that is widely found among Indo-European speakers but seems to be minimal-to-missing in the linguistically isolated Basques and Uralic-speaking Finns as well as in the currently available sample of pre-Bronze Age Europeans (older than 5,000 years BP). Dienekes calls this component “West Asian” because it “reaches its highest occurrence in the highlands of West Asia, from Anatolia and the Caucasus all the way to the Indian subcontinent.”

At this point, I’m not going to contest the reality of this autosomal component. Other researchers may check into Dienekes’s method to confirm or disprove his results. I will assume they are accurate and they intuitively make sense – Indo-European speakers range across a wide swath of southern Eurasia, whereas Basques and Finns are far removed from West Asia. (But the presence of this component in Finns, albeit at lower frequencies, than among Indo-European speakers still needs an explanation.)

What is interesting is that, while Dienekes reverted back to the 5,000 YA timeframe for the origin of Indo-Europeans, he positions his newly-discovered “West Asian” component as a refutation of the mainstream Kurgan theory of the origin of Indo-Europeans, which is inseparable from this timeframe. He attacks Jim Mallory as the key scholar who “believes” in Kurgan theory and denies the “influence” of West Asian agriculturalists on the steppe. At the same time, he doesn’t specify where exactly in “West Asia” Indo-Europeans originated, not does he offer any archaeological or ancient DNA evidence for this putative homeland.

At closer examination it turns out that his “West Asian” component peaks in quite a few populations of the Caucasus (Georgians 62.2%, Abkhazians 61.3%, Adygei 52.5%, Armenian 53.4%, Balkars 51.4%, Chechens 54.6%) as well as in Balochi (61.3%), Makrani (Southern Balochi) 60.7%, Brahui (62.1%) and Iranians (52.3%). It’s further found in a wide range of West Asian, North African, South Asian and Central Asian populations at frequencies similar to those found in Europe. If at the beginning of his journey of discovering the Near Eastern “womb of nations,” Dienekes highlighted the “Caucasus” component as central to the reconstructed original population, now, when it comes to the origins of Indo-Europeans, he prefers the label “West Asian.” There’re enough reasons in the data to continue to refer to this component as the “Caucasian” or “Caucasus” component but for some reason Dienekes chooses “West Asian.”

The reason for this choice of names is that when it comes to the origins of Indo-Europeans “Caucasus” doesn’t sound as revolutionary as “West Asia,” while when it comes to a population process that presumably began 16,000 years “West Asia” doesn’t sound as sexy as the “Caucasus.” On July 3, 2012, in a post called “Proto-Indo-European and North Caucasian,” Dienekes proudly declares that the new paper by Ranko Matasovic “Areal Typology of Proto-Indo-European: The Case for Caucasian Connections” is “quite consistent with my idea that Proto-Indo-European is related to the West_Asian autosomal component.” We also learn that “This component occurs at a a level  greater than 50% level in modern North Caucasian speakers, is absent in Europe prior to 5,000 years ago, and occurs at levels greater or equal to 10% in most present-day Indo-European speakers from Europe.”

It appears that Dienekes’s genetic data as well as Matasovic’s linguistic data are fully consistent with Kurgan theory. The Pontic steppe is geographically adjacent to Northern Caucasus and the Nalchik and Maykop cultures of the North Caucasus are part of the broad range of Bronze Age cultures at the Europe-Asia border that constitute the archaeological underpinnings of Kurgan theory. The Maykop culture (3700-2500 BC), for instance, is centered in Adygeia, which shows 52.5% of the “West Asian” autosomal component. Matasovic himself (p. 307) believes his data supports the Pontic steppe homeland of Indo-Europeans: “The evidence presented in this paper is in line with the widely-held theory that the PIE homeland was somewhere in the region to the north of the Black Sea (Anthony 2007). We believe that our arguments make it likely that it was in the eastern part of that region.” This statement directly contradicts Dienekes’s attempt to hijack Matasovic’s research to support his Near Eastern fantasy.

It is to be expected that the Caucasus refugium preserved an ancient genetic signature that got diluted in the spread zone of the steppe. Kurgan theory does not argue for the population continuity between Bronze Age inhabitants of the Pontic steppe and modern Russians and Ukrainians, an argument which would justify the expectation of higher frequencies of the “West Asian” component in the steppe vs. the mountains. It’s understood the Pontic steppe has churned different populations over the millenia.  Methodologically, at the intersection of mountains and the steppe we shouldn’t expect a complete overlap between the distribution of genetic components among modern human populations and the archaeological cultures – the archaeology will recover the traces of the homeland in the steppe, while the corresponding genetic component will better survive in the adjacent mountains. What Dienekes’s data is telling us is that proto-Indo-European bearers of Kurgan culture interacted genetically and linguistically with the non-Indo-European Northern Caucasus. As a result, they apparently absorbed a North Caucasian autosomal component (part of the Caucasian autosomal component, part of the West Asian autosomal component, but now without any relevance to proto-Indo-Europeans) next to a few grammatical properties (consonant-to-vowel ratio, tonal accent, number suppletion in personal pronouns, the presence of gender and the morphological optative and, possibly, the presence of glottalized consonants and ergativity). Kinterm loans between Indo-European and North Caucasian languages also favor long-term marriage exchange and admixture between the two speech communities. Although John Colarusso (“Proto-Pontic: Phyletic Links between Proto-Indo-European and Proto-Northwest Caucasian,” Journal of Indo-European Studies 25 (1997): 119-151) presented some evidence for a genetic connection between Indo-European and North Caucasian languages, others believe it’s unlikely that Indo-European and North Caucasian languages are related. Hence all the genetic and typological linguistic matches observed between these speech communities must result from admixture and borrowing, not common descent. Even if autosomal genetics suggests that Indo-European and North Caucasian languages must be related, this would place the break up of the hypothetical Pontic protolanguage again the Pontic steppe/North Caucasus. Hence Dienekes’s answer to the “origin of Indo-Europeans” question is neither novel, nor based on the right question.

Dienekes attempts to overthrow the mainstream theory of the origin of Indo-Europeans by silencing the fact that his own data actually fully supports it. This bizarre behavior can only be explained by the combination of the lack of professionalism, the irrational belief in the power of genetics to “solve mysteries” and “answer questions” and the desire to become famous. Although there seems to be some substance to Dienekes’s research, the way he serves it up and the way the “news” get promoted on other blogs amounts to spreading a hoax. The moment I alerted him that his data is fully consistent with Kurgan theory, he banned me from his blog. He censored my last comment but let through an incomprehensible rant from some “newtoboard” commenter who decided to support Dienekes’s hissy fit with a witchhuntesque “Thanks for banning this German Dziebel character.” I was banned from amateurs’ science blogs before (no regrets) but every time for bringing up the controversial out-of-America theory or for criticizing out-of-Africa. Dienekes placed a ban on discussing out-of-America on his blog regardless of the repeated emergence of an American Indian component in all over Eurasia (see, e.g., here). The irony of Dienekes’s reaction is that I now got banned for defending a mainstream theory. (It’s my belief that mainstream science can figure out human prehistory up to 5,000 YBP; everything prior to that should be open to widely contrasting theories.) Again, why such a violent reaction to a proposal that actually most parsimoniously incorporates a genome blogger’s data into mainstream science? It’s a good example of how the web can be used not only to promote collaboration but also to mimetically spread hype and hate and to scapegoat dissenters. The nature of the content being contested in this case is not relevant – it could be fringe, it could be mainstream. The rational goals behind research are irrelevant, too, as the only thing that matters is the powerful emotion of being part of a collective illusion that Internet can enable and perpetuate.