Burma can thank the British for its current mess


Since my last post on the Rohingya I’ve kept reading up on the topic, mostly in relation to their origins. Google scholar has been of minimal help to be honest, though this draft of a presentation given to Southeast Asia scholars in 2014 has been one of the better analyses I’ve seen, with lots of citations that you can follow up.

For me the biggest hard fact that one can not deny is that there was massive increase in the number of Muslims in Arakan recorded in the censuses of 1871, 1901, and 1911. The number of Muslims tripled in this period, and work out to an annual growth rate of 5.5%; far above anything recorded in South and Southeast Asia at the time (where growth rates were closer to 1%).

The most plausible model seems to be that most of the self-identified Rohingya in Burma today descend from a population which was part of the broader migration of peoples from the Indian subcontinent during the period of the British Raj. Between 1874 and 1917 nearly 100,000 Indians emigrated to Trinidad. In fact millions of Indian peasants were sent to far flung regions of the British Empire in the period between 1830 and 1920 (as well as merchants and traders to East Africa and elsewhere).

So where does this idea of Rohingya being in Arakan for 900 years come from? There are many websites on the internet litigating the Rohingya issue from both sides. Let me quote from one such site, Voice of the Rohingya:

The Origin of Rohingya

Rohang, the old name of Arakan, was very familiar region for the Arab seafarers even during the pre-Islamic days. Tides of people like the Arabs, Moors, Turks, Pathans, Moghuls, Central Asians, Bengalees came mostly as traders, warriors, preachers and captives overland or through the sea route. Many settled in Arakan, and mixing with the local people, developed the present stock of people known as ethnic Rohingya. Hence, the Rohingya Muslims, whose settlements in Arakan date back to 7th century AD are not an ethnic group which developed from one tribal group affiliation or single racial stock. They are an ethnic group developed from different stocks of people. The ethnic Rohingya is Muslim by religion with distinct culture and civilisation of their own. They trace their ancestry to Arabs, Moors, Pathans, Moghuls, Central Asians, Bengalis and some Indo-Mongoloid people. Since Rohingyas are mixture of many kinds of people, their cheekbone is not so prominent and eyes are not so narrow like Rakhine Maghs and Burmans. Their noses are not flat and they are a bit taller in stature than the Rakhine Maghs but darker in complexion. They are some bronzing coloured and not yellowish. The Rohingyas of Arakan still carried the Arab names, faith, dress, music and customs. So, the Rohingyas are nationals as well as an indigenous ethnic group of Burma. They are not new born racial group of Arakan rather they are as old an indigenous race of the country as any others.

The Origin of Rakhine

In the year 957 AD, a Mongolian invasion swept over Vesali, and killed Sula Chandra, the last king of Chandra dynasty. They destroyed Vesali and placed on their throne Mongolian kings. Within a few years the Hindus of Bengal were able to establish their Pala Dynasty. But the Hindus of Vesali were unable to restore their dynasty because of the invasion and migrations of Tibeto-Burman who were so great that their population over shadowed the Vesali Hindus. They cut Arakan away from Indians and mixing in sufficient number with the inhabitants of the eastern-side of the present Indo-Burma divide, created that Indo-Mongoloid stock now known as the Rakhine Arakanese. This emergence of a new race was not the work of a single invasion. But the date 957 AD may be said to mark the appearance of the Rakhine in Arakan, and the beginning of fresh period.

If you didn’t connect the dots here, what is going on is that the Rohingya are being presented as the more indigenous population of Arakan in comparison to the Rakhine majority in the text above! This is almost certainly wrong in any straightforward reading…but imagine how Rakhine react to this sort polemic, even though it is almost certainly a reaction to Rakhine nativism.

The Rohingya extensively cite their Arab and assorted West Asian antecedents in Arakan. The admission of a Bengali contribution is typical, but, it is rarely given outsized influence or importance. This, despite the fact that Rohingya are physically indistinguishable from the peasants of southeast Bengal, and their language closely resembles the Chittagong dialect of that area.

I can make judgments on the issues of physical appearance and language. My own family is from a nearby region (Comilla, and some branches of the family, Noakhali), and I have been to Chittagong. I can understand to some extent the Chittagong dialect, and the Rohingya language is clearly related to it (the peasant Bengali of the region of Bangladesh my family is from is almost certainly closer to Chittagong dialect than standard Bengali because of proximity).

But it does seem clear that some Muslims were present in Arakan at a relatively antique date. The romance of Sinbad the Sailor reflects that even in the period before 1000 AD Muslim travelers and traders were a common on the shipping lanes of the Indian ocean, even as far east as the trading entrepot of Guangzhou in Tang China. This can be confirmed by the fact that Muslim conflict with Chinese occurred in 758 and culminated with a well known massacre of foreigners, mostly Muslims, in 875. They were certainly in Arakan by this period in some numbers.

Nevertheless this Arab connection to Arakan is tenuous at best. The Muslims of Arakan are of the Hanafi school of shariah, which is dominant in the Turco-Persian-Indian world. In contrast, Arabs tended to transmit the Shafi school to the eastern shores of the Indian ocean. This is evident among the Muslims of Kerala, who have long had a relationship with southern coastal Arabia, and so adhere to the Shafi school unlike the vast majority of India’s Sunni Muslims (Southeast Asia is Shafi as well).

More concrete and substantive is the association of kingdom of Mrauk U with Islamicate civilization, and in particular the Sultanate of Bengal and later the Mughal Empire. This polity flourished between the 15th and 18th centuries. In the early period the Muslim rulers of Bengal to the north and west patronized the dialect of the region, what was becoming Bengali (elite support for Bengali among Muslims declined during the Mughal period). This is also when large scale Islamicization began on the eastern frontier according to The Rise of islam on the Bengal Frontier. A number of Muslims settled in Mrauk U during these centuries, often in association with the sultan’s court. It seems likely from the literature I have seen that the term Rohingya began to become popular among Muslims in Arakan during this period at the latest. Even before the British conquest of Arakan they noted the existence of a community of Muslims who called themselves Rohingya.

So what is the connection between the Rohingya, who certainly existed as a community in Arakan before British conquest, and the modern Rohingya, who probably descend from peasants who arrived from southeast Bengal in the 19th century?

First, one needs to be reminded that at in 1940 16% of Burma’s population was of recent Indian subcontinental origin. They spanned the gamut from wealthy Chettiar financiers to middle class Punjabi police to peasants from Bengal. In the decades after World War II the majority left the country, especially the prosperous ones. The literature I’ve read indicates that the less prosperous ones, who did not have portable skills or assets, were less likely to leave. Many of them have assimilated to Burmese culture in cities such as Yangon (many Hindus have switched their religious affiliation to Buddhism, while all Burmese with total fluency).

The Rohingya were drawn exclusively from a peasant culture, and exist in concentrations in a particular region where they are preponderant (northern Arakan), and so exhibit cultural critical mass. The Rohingya are by and large not a literate people. At least until recently. Though their language is clearly Indo-Aryan, and closely related to Bengali, they do not use Bengali script.

In Bangladesh there are two regions where Bengali-related dialects are extensively spoken. In Sylhet in the northeast and Chittagong in the southeast the local dialects are unintelligible with standard Bengali. But because Bengali is a dialect continuum (standard Bengali derives from a particular region of West Bengal, in what is today India) I have better understanding of Chittagong dialect that most Bengali speakers, as I also understand to some extent rural Comilla speech, which is nearer to the Chittagong dialect. People in Sylhet and Chittagong can generally speak standard Bengali, and though both groups exhibit some ambivalence they do consider themselves Bengali.

So why are the Rohingya different? The period when the Rohingya migrated to Burma was also the period when European-style conceptions of nationalism, based around a common written language, were starting to take hold in South Asia. South Asians of all religions, at least of an elite background, understood themselves as being part of Hindustan, which was characterized by its own unique traits. But their identity on a national level, bound by language, was weak. This is in part because the languages favored by the elites were Persian or Sanskrit, with the simultaneous emergence in North India of the dialects that later became Urdu and Hindi (South India had its own independent traditions).

In Germany, Italy, and France, the standard national languages spread in the 19th and 20th centuries so that local dialects went extinct. That process in a place like South Asia has been much slower because the illiterate agricultural peasant majority has been more insulated from literacy. Just as Italian is based on the dialect of Florence, so standard Bengali derives from a particular region in modern West Bengal, and spread as an elite language across the Bengali dialect continuum (when I was a very small child in Bangladesh my parents made an effort to speak in very standard Bengali so that that was my native language, as opposed to a lower status dialect).

From what I can tell the Rohingya were untouched by the changes triggered by the Bengali Renaissance in the early 19th century, which first captured the imaginations of the Bengali Hindu upper castes, and then spread over the decades to the middle classes of all both religions and all regions. In lieu of a Bengali identity the Rohingya seem to have co-opted the identity of the earlier Muslim community of Arakan, whom they likely absorbed. This is not entirely fantastical. I have some experience talking to peasants in the countryside of Bangladesh, and even now the lower classes are vague about their national and ethnic identity. Rather, they focus on their village and locality, and exhibit little sense of scale of difference in relation to outsiders. In 1990 when I was visiting rural Bangladesh I remember being introduced to a woman from Bogra district as a “fellow foreigner.” Bogra was less than 200 miles from where I was at the time, but for these peasants it was another world, and we were interchangeable in our alienness.

There is a clear analogy to what might be happening with the Rohingya, and that exists in the Central Asia. In the 19th and early 20th centuries a broad group of settled Turks and Iranians in Central Asia under Chinese nationalist and Russian rule transformed their ethnic identities into those we know today. Groups like Uyghurs and Uzbeks have tenuous connections to the historical groups which were called Uyghurs and Uzbeks. The decision of the various Turkic speaking groups of the oases of Xinjiang to call themselves Uyghur, and so established a connection to the Turkic confederation which flourished more than 1,000 years earlier, occurred in the 1920s. Of course today we don’t say that the Uyghurs aren’t real Uyghurs; the assent of the Turkic speaking population to the term Uyghur has resulted in that becoming their new self-identity.

Similarly, a reshaping of the self-identity of the people who became Rohingya in Burma likely occurred in the 20th century. Without an educated upper class which was literate they were totally detached from the emergence of a modern Bengali identity. Rather than become Bengali the more upwardly mobile members of the Rohingya elevated their own dialect into its own language (and adopted a non-Bengali script as well). Additionally,  they synthesized various aspects of Islamic history in Arakan and integrated it into their own identity, and so establishing their bona fides as sons of the soil.

To the question of whether Rohingya have been present in Arakan for 900 years, I think it is clear I believe that the answer there is no, not at all. But, to the question of whether they actually Bangladeshi, I would also have to say at this point, no. This is a subtle and nuanced issue, but the existence of a Bangladeshi national identity is relatively new, just as the nation is new. But it is has likely been on the order of 100 years or more since most of the ancestors of the Rohingya left the districts of southeastern Bengal to which they were originally native. Not only have the ancestors of the Rohingya never lived in Bangladesh, but they never lived in Pakistan. It would be somewhat similar to suggesting that an Indo-Trinidadian would be comfortable in India, though the gap here is larger because most Indo-Trinidadians lack fluency in any South Asian language today.

The non-Bengali identification of the Rohingya has also made their Islamic identity more salient. If you Google image Bangladeshi women as opposed to Rohingya women, the latter look far more Muslim. I’m no expert, but the Rohingya seem partial to the headscarves of the sort common in places like Malaysia, again attesting to their attempt, conscious or not, at indigenization. Without a connection to Bengali elite culture, which is multireligious, and of a secular bent due to that fact, the Rohingya will naturally gravitate toward an Islamic identity when they attempt to transcend their peasant origins.

Why does any of this matter? Obviously the truth matters. And if the world community is to foster peace in Burma it will not help its cause by promoting false narratives created by the Rohingya as a counterargument against those Burmese who deny their national status. The Rohingya have established their non-Bengali identity, as they have created a different one in Burma. But attempts to the deny their relatively recent origins in South Asia will almost certainly inflame and agitate the Rakhine majority and the Burmese state even more than is the case now. It will also undercut any credibility that outsider have in fostering moderation and peace.

The Ghosts of the European Pleistocene

2011’s Cave of Forgotten Dreams was a strange film. I went to watch it in the theaters mostly to see the paintings of Pleistocene peoples in an immersive manner, but the director and narrator, Werner Herzog, used the film as an instrument to forward his thesis that humanity as we understand it emerged during this period in the European Ice Age.

Whether he knew it or not Herzog was channeling the “Great Leap Forward” model of the origin of our species. That in an almost punctuated manner the cultural proteanism which we take to be a defining hallmark of our species emerged at some point deep in what we call the Ice Age. In The Dawn of Human Culture Richard Klein localized this burst of humanity ~50,000 years ago in Africa, and hypothesized that it was triggered by a biological change which enabled language fluency. In The Humans Who Went Extinct Clive Finlayson posits that cultural changes associated with the Gravettian people in central Eurasia eventually defined what he meant to be human, and explained the marginalization of Neanderthals.

To a rough approximation I’m skeptical of both these models. I don’t think humanity emerged fully formed like Athena over the last 50,000 years. Rather, humanity we understand humanity is deeply primal, and a feature of the root of our lineage, millions of years in the past. If Homo erectus populations were still around they deserve all the rights of humans, despite their numerous differences.

I suspect that we’ll found out that ‘behavioral modernity’ is a cocktail of soft selection on standing variation and cumulative cultural change. But that doesn’t mean that the Pleistocene history of Europe is not important or interesting. And recently we’ve obtained enough ancient DNA to sketch out a general picture of demographic, if not cultural, change.

Everyone should read The genetic history of Ice Age Europe. But I suspect the impact is going to get deeper when more archaeologists are familiar with the implications. Here is the abstract:

Modern humans arrived in Europe ~45,000 years ago, but little is known about their genetic composition before the start of farming ~8,500 years ago. We analyze genome-wide data from 51 Eurasians from ~45,000-7,000 years ago. Over this time, the proportion of Neanderthal DNA decreased from 3–6% to around 2%, consistent with natural selection against Neanderthal variants in modern humans. Whereas the earliest modern humans in Europe did not contribute substantially to present-day Europeans, all individuals between ~37,000 and ~14,000 years ago descended from a single founder population which forms part of the ancestry of present-day Europeans. A ~35,000 year old individual from northwest Europe represents an early branch of this founder population which was then displaced across a broad region, before reappearing in southwest Europe during the Ice Age ~19,000 years ago. During the major warming period after ~14,000 years ago, a new genetic component related to present-day Near Easterners appears in Europe. These results document how population turnover and migration have been recurring themes of European pre-history.

I modified the model of demographic turnover to the left, adding labels for the primary paleoanthropological cultural groups. Instead of starting with the archaeology the authors let the genetic results guide them. What they discover is that there were roughly four turnovers in population defined by four “clusters”:

– the first Europeans who succeeded the Neanderthals, who seem to have left no descendents

– the Goyet cluster, associated with Aurignacians

– the Vestonice cluster is associated with the Gravettians

– the El Miron cluster with the Gravettians

– the Villabruna cluster with various late Pleistocene cultures, and is the direct ancestor of Mesolithic hunter-gatherers present in Europe when the first farmers arrived

A quick calculation suggests that very little of the ancestry of modern Europeans has deep roots across much of the continent going back to before the Last Glacial Maximum ~20,000 years ago. The “Pleistocene” ancestry of Europeans mostly derives from the last group, the Villabruna cluster. In the paper the authors note that this group is unique for several reasons:

– some individuals in this cluster have an affinity with East Asians (earlier Pleistocene groups do not)

– more universally, individuals in the Villabruna cluster have a notable affinity with Middle Eastern populations which was not evident in earlier Pleistocene clusters

Recall Middle Eastern populations can be modeled as a mix of a West Eurasian group similar to European hunter-gatherers, and, “Basal Eurasians,” who are an outgroup to all non-Africans (European hunter-gatherers to Oceanians to Amerindians!). The authors posit that the gene flow is more likely from the Middle East, because earlier European clusters have affinities with Villabruna, but they share nothing with the Middle East. The Villabruna cluster does not have Basal Eurasian ancestry though. So we might be looking at complex population structure.

Two general issues that crop up in this paper are sampling limitations and population expansions into Europe from the east. The disappearance of Goyet ancestry, only to reappear as part of the El Miron cluster, is curious. Perhaps the post-Goyet people occupied ecologies less likely to be fossilized? It reminds us of the resurgence of hunter-gatherer ancestry in Europe of the Middle Neolithic. As for why there seems to be an eastern bias into intrusive populations into Europe, these groups may simply have had a larger population, and so been more likely to avoid meta-population extinction events?

Finally, the authors point out that Gravettian culture in Siberia and Europe does not seem to be genetically related. This suggests that these people were very modern, because a hallmark of the modernity is that ideas can move between groups without too much genetic exchange.

Europe has the best coverage of ancient DNA. What we find here are repeated population turnovers and a lot of complexity and contingency. Is Europe peculiar? There is circumstantial evidence from modern DNA that Australians are descendants of first settlers. And to some extent this is also true in the case of East Asians if the work reported out of the Fu lab holds up. South Asians though are more likely to be like Europeans, and Middle Easterners show some of the same dynamics during the Holocene.

Totality, 2017

We saw totality in Independence, Missouri. It was a pretty long car trip with the kids, all of whom are under school age (though just barely in the case of the eldest). We had a big fright because there was a massive thunderstorm in the morning…but it cleared out for a few hours. And that was when the total eclipse occurred.

Really words don’t do the experience justice.

Chase a total eclipse if you get a chance.

The past was not PG

The Week has published a screed against the low moral quality of Game of Thrones, Game of Thrones is bad — and bad for you. Obviously there is something to this insofar as one can see a coarsening of entertainment, or at least a decline in the stylized aspects of the depiction of reality.

But one of my initial reactions is that much of the narrative that we value from the past was not particularly PG. If you read The Harlot by the Side of the Road: Forbidden Tales of the Bible you see that the “Good Book”, in fact the only book many read front to back by many after the Reformation in Protestant Europe, has some quite unsavory tales. The story of Judah and Tamar in particular is hard to digest from a modern Western perspective because many of the elements are understated and workaday. Greek mythology is no better obviously. From Zeus raping Leda, Achilles throwing a fit because his sex-slave was taken away, to the tradition of Agamemnon sacrificing Iphigenia.

In some cases the shocking aspect of ancient stories is because moderns have different values. Slavery and concubinage were taken for granted during the period that the Hebrew Bible and Classical mythology crystallized into the forms which came down to us. In other cases I presume that it was unlikely that small children were going to ever read the original stories themselves, so sexual elements that might confuse were probably omitted in some oral tellings.

This is not to say that Game of Thrones is a modern masterpiece. But some of the disquieting, and frankly perverse, aspects of the narrative are only shocking if your standard is the relatively antiseptic literary fiction which one finds between the Regency and the cultural revolution of the 1960s. That is the aberration in human history, while gritty genre fiction much closer to primal human storytelling.

Open Thread, 07/09/2017


I’m a sucker for the aesthetics of Norden. Why? I wonder if part of it is that the fringe of Northern Europe is a science fictional setting. The long dark nights during the cold winter, and the twilight during midsummer. The sun may be bright, it never gets too high in the sky. The 13th Warrior wasn’t the best movie, but it was evocative. One of the problems with the film depiction of the Lord of the Rings trilogy is that New Zealand seems too bright and airy (and also not decayed enough).

Because of the SMBE meeting I haven’t made much progress on The Enigma of Reason. Much of it has been reviewing the literature in cognitive psychology and reasoning which I’m familiar with (system 1 vs. system 2, Wason reasoning task, etc.). Though it is leading me up to the main thesis.

I remember years ago Matthew Yglesias mentioned he was going to do a bit more reading of books, as opposed to news, to differentiate himself from other pundits. Today he admitted he wasn’t going to make a show of having an informed opinion about the Frankfurt School. I suggested he take time out to read The Dialectical Imagination: A History of the Frankfurt School and the Institute of Social Research, 1923-1950. The modern campus Red Guards don’t know anything about Adorno, Marcuse, or Horkheimer. But the outlines of contemporary project toward cultural revolution and exaltation of the marginalized are all there. Rather than being the origin modern radical movements, I suspect that the Frankfurt School simply provides a useful tool and framework to go about its project.

I do know that some politically moderate scientists who read The Dialectical Imagination, and saw campus politics in a totally different, and more intelligible, light.

Joe Pickrell’s new company, Gencove, Improving ancestry estimates in South Asia.

I said on Twitter that “easiest way to make housing affordable for non-rich is to build more houses for the rich so they won’t buy houses built for non-rich.” What do I mean? It’s all about supply. The well-off will always be first in line for any supply of housing. If you allow for copious development, vertically and horizontally, then the rich can purchase the luxury condos and mansions that they crave, while the middle class and lower class can buy up the more normal housing stock.

The Austin skyline, then and now.

Yazidi sex slave had to choose between son and escape from Isis.

Bangladeshi students test into elite schools. This story is about the entrance examinations for the elite public high schools of New York City. In 2010 the average Bangladeshi family in New York City had a household income of $37,000. I believe in the near future the entrance exams will not be the only criterion for gaining admission. The reality is that Asian American students lack “leadership” and are not “well rounded,” and all the Asian American applicants “look the same.”

A large-scale genome-wide enrichment analysis identifies new trait-associated genes, pathways and tissues across 31 human phenotypes.

Racism Is Everywhere, So Why Not Move South? This article is written in the context of black Americans. But the insights are general. Houston has a cost of living that’s at the national average. It’s the fourth largest city in the United States, and there is a lot of good phở because of the large Vietnamese community.

Patrick Wyman, who sometimes comments on this blog, has a great Fall of Rome podcast.

The Sad, Sexist Past of Bengali Cuisine. Really upper caste Bengali Hindu cuisine.

Utilities fighting against rooftop solar are only hastening their own doom. Not surprised. I have been following Ramez Naam’s commentary on this for years. He’s been on this.

Islamophobes are attacking me because I’m their worst nightmare. Linda Sarsour.

I thought Hillary would win the election. But I told a long-time reader of this weblog who is a Democratic operative that BLM activists getting in Bernie Sanders’ face did not presage well for the direction of the party. Linda Sarsour as the face of progressivism is a massive boon for the Right and Republicans.

Sarsour has left a trail of obnoxious and offensive comments on Twitter. So have many people. For me personally the biggest issue is possible solidarity with Rasmea Odeh. The PFLP is the literal definition of a terrorist organization (though a Marxist, not Islamic, one). But the reality is that her enemies on the Right know that she and her compatriots in the “woke” movement would never exhibit charity toward their political opponents, so they are attempting to destroy her because they know she would do the same to them. That’s where we are in American politics today. You destroy your enemies, or they destroy you. Let’s have fun until the last battle though!

A combined analysis of genetically correlated traits identifies 107 loci associated with intelligence. I guess I’ll start paying attention to when they can explain ~25% of outgroup sample variance. They’re already further than the 7% in this preprint, though that will take a little longer to publish.

Procedures for enumerating and uniformly sampling transmission trees for a known phylogeny.

A reanalysis of Schaefer et al. does not indicate extensive CRISPR/Cas9 mediated off-target editing events.

Evolutionary Action of de novo missense variants across pathways prioritizes genes linked to autism and predicts patient phenotypic severity.

The search for Eden opens up new vistas

The end of Eden

A particular conception of the “Out of Africa” model of human origins died in this decade. This model hooked into preexistence narratives about “Adam” and “Eve”, utilizing Y and mitochondrial DNA lineages passed down through direct male and female lines respectively. Its most extreme manifestation could be exemplified by Richard Klein’s ideas in the early 2000s outlined in his book The Dawn of Human Culture.

For Klein the chasm between Homo sapiens sapiens, humans, and other hominins was vast. A physical anthropologist who surveyed with skill the rapid expansion and proliferation of modern human cultures over the past ~50,000 years, Klein relied on a particular evolutionary model to explain how this occurred. He posited that humankind emerged in East Africa as a punctuated speciation event, triggered by a mutation which allowed for the development of fully elaborated recursive language.

The difference between our own lineage and our relatives in this framework was huge. To not put too fine a point on it, Neanderthals and other archaic humans were animals. We, Homo sapiens sapiens, were humans qua humans.

Though Klein was a paleoanthropologist, he gained great support from a school of molecular evolution which arose in the 1970s and 1980s under Allan Wilson. Wilson’s initial fame arose because he utilized a “molecular clock” analysis of primates to contend that the divergence of our human lineage from great apes was much more recent than paleontologists had believed. Eventually new fossil finds confirmed the molecular phylogeny. After this event Richard Leakey has stated paleoanthropologists were reluctant to challenge molecular results.

Wilson later focused on recent on human origins, utilizing mitochondrial DNA, which is passed down directly through the maternal lineage. In this way they found that African mtDNA lineages were very diverse, and that non-African lineages were nested within the broader tree of African lineages.

The conclusion from this finding was that modern humans arose in Africa and spread to other parts of the world. This conclusion in general has been confirmed.

But over the years more and more evidence has accumulated that the story is more complicated than the original narrative that all modern humans descend from a small bad of East Africans who populated the whole world ~50,000 years ago.

Dissenters from Eden

There were always geneticists who were skeptical of the neat Out of Africa with total replacement model. In Origins Reconsidered Richard Leakey recounts a conference in 1992 where he was pigeon-holed by geneticists who thought there was no reason to accept without dispute the mitochondrial Eve narrative. Over the years talking to some older geneticists I can say that Leakey was reporting a real undercurrent of irritation with the confidence that Allan Wilson’s group and their fellow travelers projected in relation to their model. Nordborg 1998 On the probability of Neanderthal ancestry reflects some of the technical objections to inferring too much from one locus when it came to the possibility of other components of ancestry.

When genome-wide analyses in the middle 2000s became feasible, a visible counter-culture within genetics argued that total replacement was not supported by the data. In 2006 Wall & Hammer published Archaic admixture in the human genome. They concluded that “Recent work suggests that Neanderthals and an as yet unidentified archaic African population contributed to at least 5% of the modern European and West African gene pools, respectively.” They were not that far off with European populations. As far as Africa goes, that is a question that will be explored in detail in the next few years.

That analysis though has only 44 citations. I have had debates on Twitter with how exotic and marginal these ideas were. In general it is safe to say that they were not exotic and marginal in the community of human evolutionary population geneticists. But that’s not a large set. Both John Hawks and Milford Wolpoff have indicated a lot of marginalization for models outside of the narrow window of Out of Africa with total replacement. From everything I’ve heard about the run up to the 2010 publication of the Neanderthal genome many of the principal researchers, including Svante Paabo, were totally surprised by the evidence of admixture into modern lineages. Wolopff even emailed me after I reviewed the paper to suggest that it felt so good to come out of the wilderness and have some of his views accepted.

Anagenesis and punctuation?

But were Wolpoff’s views accepted? The revised model actually kept much of the Out of Africa framework in place, except it added the wrinkle of assimilation of some archaic lineages. The dominant signal in the non-African genomes seems to have come from an African lineage which left around ~50,000 years ago.

The classical multi-regional model that Wolpoff was associated with, whereby modern humans evolved across the whole world from local archaic lineages, but maintained species cohesion through gene flow, was not supported. Rather, the archaic admixture of Neanderthals and Denisovans into Oceanians pointed to local continuities, which was a broader position of multi-regionalism. But this is not speciation without branching, anagenesis.

Nevertheless, there was another aspect of Out of Africa with replacement that needed revision. Though not explicitly outlined in many framings, one aspect implicit is that the dynamics that Africa and Eurasia were subject to during the emergence of modern humans were the same.

But that doesn’t seem to be the case. The ancestors of all non-Africans went through a major population bottleneck. On the order of ~1,000 individuals (this is a very large bottleneck actually, and I’ve seen numbers as low as 100, though that seems on the small side; calculating effective population size ~50,000 years ago can be tricky). The same is not true of African populations. Though many of them show signals of population declines during the Pleistocene, the extreme uniform bottleneck which characterizes all non-Africans, from Iberia to Australia to Patagonia is just not evident in Sub-Saharan African populations.

In other words, the Out of Africa event did not apply within Africa. Here’s an excerpt of an email I sent to Carl Zimmer in December of 2010 (he was updating the second edition of The Tangled Bank):

…it may be that there was no rapid antique population expansion in Africa which was analogous to [the] out of Africa migration. IOW, non-Africans are just a branch of Northeast Africans, and the Bushmen and other groups were already differentiated by that point. So you could theoretically remove the arrows within Africa! I think this is a subtle and tendentious point, so probably best to leave that as it is. But remember how deep the basal branching of the Bushmen was in the Denisova paper? It WAY predates any possible out of Africa migration by multiples.

Which brings me to the current year and the present time. The recent paper which utilized an ancient genome from South Africa to push back the date of the diversification of African lineages to about ~250,000 years before the present was not entirely surprising to me. Every time I talked to people who had access to African whole genomes their dates kept getting pushed back further and further into the past.

And of course we now have fossil confirmation that human populations which seemed to be anatomically modern (or close) were already present ~300,000 years ago in Morocco. The New York Times has a good overview of the work, Oldest Fossils of Homo Sapiens Found in Morocco, Altering History of Our Species. I read the papers and the commentaries and don’t have much to add, nor do they add much for non-specialists in my opinion (since we can’t really judge the morphology too well, nor do we have a detailed understanding of the fossil record). In one of the Nature letters the authors conclude in the abstract that “The emergence of our species and of the Middle Stone Age appear to be close in time, and these data suggest a larger scale, potentially pan-African, origin for both.”

This suggest to me anagenesis. Has multi-regionalism come back, but no within Africa?

Parameters, not paradigms

John Hawks has put in his two cents, and it’s always worth paying attention. My major take home is that we don’t know a lot even though we know more, and we need to be careful here. The genome blogger from the 2000s who has been relatively quiet over the last five years, Dienekes, resurfaced, dismissing the idea of pan-African anagenesis and asserting an Out of North Africa viewpoint. He’s been talking about this model since 2011, so there’s nothing new here. In January of 2011 he asserted that “Africa was home to a structured population.” That is what we are seeing today.

The publication of the Nature letters triggered a lot of discussion on Twitter. When I was involved it mostly consisted of Aylwyn Scally and Pontus Skoglund, with John Hawks, Chris Stringer, and others jumping into the stream. Here are some points which are of note:

1) Most people now suspect that large scale population structure within Africa over the past few hundred thousand years is a major story.

2) But there is an assumption that collapsing of that structure through gene flow was not reciprocal. That is, some populations likely expanded at the expense of others. The arguments are whether the assimilation of the secondary groups is on the order of a few percent, as seems to be the case in Eurasia, or a much higher fraction.

3) Because the phylogenetic distance between within African lineages is likely smaller than between Neanderthals and modern humans, as well as the likely similar census sizes and technological toolkits, I contended that it is not unreasonable to guess that as much as 20% of the ancestry of a daughter population of an expanding group could be from the local substrate. There was no great objection to this guess.

4) Remember, even simple mtDNA phylogenies as far back as the 1980s, as well as paleontological analyses of fossils, indicated that an Out of Africa movement into Eurasia. This was such a strong signal in the data that it was clear with even relatively little to go on. The situation for within Africa is not analogous, suggesting to me that an extreme model of replacement or gene flow across persistent demes in local regions is not tenable.

5) Ultimately, the issue will resolve on parameters of admixture and the nature of demographic expansion in the details. Instead of a tree, we will conceive of this as a graph, a trellis with lengths of different thicknesses.

6) Hawks brought up the fact that one reason classic multi-regionalism did not work is that the Fisher wave of expansion of favored genes is slower than the migration of humans. When I suggested it does not seem that the genetics of gene flow in plants, which do resemble classical multi-regionalism, were a good analogy for humans, Skoglund contrasted the sessile nature of the taxon in contrast to mobile humans. I did point out though that after favored alleles moved through migration into a population, there was often in situ selection. He agreed.

7) A key issue that both Hawks and Dienekes emphasize is that we don’t know the role that extremely diverged lineages from our own ancestors play in our story. That is, were there many modern human populations across Africa, interspersed with other human species? Or was there one modern human population that mixed with other species? We don’t really know the details of all of this.

8) I expressed skepticism of the idea of “behavioral modernity.” My reason for being skeptical is that the origin of modern humans is not as neat as we like to think, and the origin of “behavioral modernity” is also not as neat as we like to think. When the consensus was that humans emerged as a punctuated de novo event, ensouled by the Lord God on High 50,000 years ago (or, coming down from the skies as in Battlestar Galactica or in Larry Niven’s Ringworld), the idea of behavioral modernity kind of made sense. But it’s all more confused now (in any case, in Clive Finlayson’s The Humans Who Went Extinct he seems to be arguing that much of modern culture was invented by Gravettians, well after the Out of Africa event).

The consensus seems to be that rather than focusing on a set of human universals as behaviorally modern, we should look at the demographic patterns of the past to infer when our own lineage came into its distinctive being. Those of you who have read me for a while know this is already congenial to me, Luke Jostins’ plot of the encephalization of all hominin lineages over the past million years was suggestive to me long ago that our own lineage is not so special. Rather, something like us was probably inevitable so long as an asteroid didn’t wipe out large mammals once Homo erectus spread across the globe. Humanity is a destiny, not a lineage.

At an inflection point of archaeology and genetics

People always ask me what to read in relation to the field of historical population genetics. In the 2000s there were a series of books which focused on the mtDNA and Y results from modern phylogeographic analysis. Journey of Man, Seven Daughters of Eve, The Real Eve, and Mapping Human History. But there hasn’t been much equivalent in the 2010s.

Why? I think part of the issue is that the rate of change has been so fast that scholars and journalists haven’t been able to keep up. And, the change is happening right now, so it would likely mean that any book written over a year would be moderately out of date by publication.

I noticed today that Jean Manco has an updated and revised version of her book, Ancestral Journeys: The Peopling of Europe from the First Venturers to the Vikings. This was needed, because the original book was written before some major recent findings, though after some preliminary ones. As Manco has observed herself it was feasible to replace speculations with facts.

Since it seems likely that George R. R. Martin’s next book will be published before David Reich’s, I think that’s all you got. Any suggestions would be welcome.

As for the flip side for history that might be useful to understanding the genetics results, J. M. Roberts The History of the World is the best cliff notes I can think of. It’s obviously a high level survey, but frankly that would improve the interpretation I see in some papers. The fact that much of the history has no contemporary relevance is pretty unimportant, since you want to focus on the older stuff, which is where ancient DNA really shows its metal.

At some point ancient DNA will start to exhibit diminishing returns. Then the long hard slog of interpretation and synthesis will have to begin in earnest.