Wednesday, December 23, 2009

263. The Baseline Scenarios -- 39: The Gap

While the mtDNA "clock" contrasts superhaplogroup M in South Asia with M in "East Asia," it's important to understand that the dynamics of the "southern route" points to South-east Asia as the precursor of modern humans in northern East Asia. In other words, if we find tone languages in China and Japan, it is probably because tone languages were already established in Southeast Asia, long before modern humans migrated north from there.

Further confirmation of the "southern route" hypothesis comes from a very recent paper, just published in the December edition of Science, Mapping Human Genetic Diversity in Asia, by "The HUGO Pan-Asian SNP Consortium":
More than 90% of East Asian (EA) haplotypes could be found in either Southeast Asian (SEA) or Central-South Asian (CSA) populations and show clinal structure with haplotype diversity decreasing from south to north. Furthermore, 50% of EA haplotypes were found in SEA only and 5% were found in CSA only, indicating that SEA was a major geographic source of EA populations (from the Abstract -- my emphasis).
Evidence for important founder effects in South Asia, as suggested by both Oppenheimer and Soares et al., comes from another recent paper, published in Nature last September, Reconstructing Indian population history, by David Reich et al: "Allele frequency differences between groups in India . . . [reflect] strong founder effects whose signatures have been maintained for thousands of years owing to endogamy" [my emphasis]. So strong are these effects that they are believed to have an important influence on the health of the population generally:

The widespread history of founder events in India is also medically significant because it predicts a high rate of recessive disease. In Finland, there is a high rate of recessive diseases that has been shown to be due to a founder event, and that has resulted in a minimum FST of 0.005 with other European groups. Our data show that many Indian groups have a minimum FST with all other groups that is at least as large. . .

By showing that a large proportion of Indian groups descend from strong founder events, these results highlight the importance of identifying recessive diseases in these groups and mapping causal genes [my emphasis].
The large number of Indian lineages with a history of founder effects is consistent with the notion that South Asia could have been subject to a major population bottleneck during the Out of Africa migration, possibly due to a catastrophic event such as the Toba explosion, a Tsunami, major flooding, or a widespread, long-lasting drought. As I suggested in my "Echoes" paper, an event such as this could have had a major impact on the culture of the surviving groups, which could explain both the absence of much trace of P/B musical style, and the absence of tone language in this region. In other words, the "strong founder events" reported in this paper may have been a major factor in human evolution, genetic, morphological and cultural, at a crucial moment in history.

As is well known, certain tribal groups in South Asia have been described as "Australoids," due to their striking resemblance to Australian aboriginal peoples, which has suggested to many an archaic connection between the two regions. As I wrote in "Echoes of our Forgotten Ancestors,"
Is it possible that Australia, populated, at least 10,000 years after the Toba blast, could have been occupied by one of the altered survivor lines out of South Asia? Could that also explain the musical discrepancies between Australia and the rest of the out-of-Africa P/B style singer-players? Could that population have simply lost its polyphony, its hocketing, its panpipes, etc. as
a result of an ages-old natural catastrophe, and been forced to invent itself anew?
While a genetic connection between south Asiatic tribal peoples and Australian aborigines has been debated for years, a recent paper on that topic, Reconstructing Indian-Australian phylogenetic link, by Satich Kumar et al., though modest, does support such a possibility:

Our complete mtDNA sequencing of 966 individuals from 26 relic populations of India identified seven individuals from central Dravidian and Austro-Asiatic tribes who share two basal synonymous mtDNA polymorphisms G8251A and A9156T with
the M42 haplogroup, which is specific to Australian Aborigines. . .

Our results showing a shared mtDNA lineage between Indians and Australian Aborigines provides direct genetic evidence that Australia was populated by modern humans through south Asia following the "Southern Route" (pp. 1-2 -- my emphasis).


The cultural connection would be more convincing if the music of Tribal India were stylistically close to that of native Australia, but it is not. Nevertheless, the absence of both the African musical "signature" and tonal languages in both Tribal India and Australia, whereas both of these important cultural indicators can be found in abundance everywhere else along the Out of Africa path, does seem consistent with genetic evidence pointing to an archaic connection between the two regions. The unusual features attributed to "Australoid" peoples are also consistent with a genetic bottleneck and subsequent founder effect centered in South Asia during the great migration eastward. If the first wave of migrants had already arrived in Southeast Asia by this time, they would have been upwind from the Toba blast and, assuming this as the cause of the bottleneck, relatively unaffected.

17 comments:

Maju said...

SEA was a major geographic source of EA populations

This is surely correct (at least is my opinion, regardless of whatever Neolithic backmigrations). However it means mainland SE Asia, Indochina and not Sundaland, where the various groups of Negritos seem isolated since the very initial expansion.

The large number of Indian lineages with a history of founder effects is consistent with the notion that South Asia could have been subject to a major population bottleneck during the Out of Africa migration, possibly due to a catastrophic event such as the Toba explosion...

I don't read it that way. What I read is that due to persistance of endogamy (via the transformation of tribes and clans into jatis or endogamous castes), the founder effects of Paleolithic have been preserved to the present, what has not happened elsewhere to the same extent because endogamous castes do not exist elsewhere.

It has nothing to do with ancient bottlenecks but with "modern" (post-Neolithic) social organization. It is highly suggestive though of the caste system fundamentals being pre-Aryan, probably as old as Neolithic itself, with tribes transforming themselves into castes (jatis) naturally as they made their transition into Neolithic and civilized "modernity". A very curious peculiarity of South Asia, indeed.

Re. the Australian connection, it is very weak. The akin lineage has not been even seriously proven (they share two polymorphisms but not a clear phylogeny yet) and would anyhow only affect minor lineages in both continental regions. The "Australoid" tag for various Asian peoples' morphology (not just Indians but also SE Asians, Ainu and I'd dare say many Caucasoids) has also been contested often and is generally considered a mis-interpretation of widespread archaisms that rather belong to the common founder population of all Greater Eurasia.

German Dziebel said...

M42 is a minority lineage in Australia. Australian N lineages seem to be key to its peopling. And here we find no matches with South Asia. (Unless, of course, the expansion of the N-carrying Pama-Nyungan family replaced all pre-existing M lineages within the past 10K.)

"Tribal" kinship systems in India are most reminiscent terminologically and sociologically (endogamy, marriage practices such as sister's daughter marriage, bilateral cross-cousin marriage) of Amazonian ones. Fst levels are also similar between tribal India and Amazonia. In specialized literature, "Amazonian" kinship type is considered to be a purer version of "Dravidian" type. At the same time, Australian kinship systems ("Kariera" being the root type, with "Aranda" and others to follow) are quite unlike South Asian ones (in my "Genius of Kinship" "Kariera" and "Dravidian" systems are contrasted as two different types with evolutionary possibilities of conversion of one into another discussed) but are similar to some North American Indian patterns.

Linguistically, there're no special matches between Australia and South Asia either, with all grammatical similarities of Australian languages located in Papua New Guinea, North and South America. (I detected one additional match with a Khoisan language.) In late 19th century, however, the alleged Dravidian-Australian connection was explored by a number of linguists. These days these attempts are considered futile. But since you, Victor, tend to consider superficial linguistic similarities as potentially reflecting ancient migrations, here is a reference to one such early work: An Australian language as spoken by the Awabakal, the people of Awaba, or lake Macquarie (near Newcastle, New South Wales), by Lancelot Edward Threlkeld (1892).

German Dziebel said...

"It has nothing to do with ancient bottlenecks but with "modern" (post-Neolithic) social organization. It is highly suggestive though of the caste system fundamentals being pre-Aryan, probably as old as Neolithic itself, with tribes transforming themselves into castes (jatis) naturally as they made their transition into Neolithic and civilized "modernity". A very curious peculiarity of South Asia, indeed."

As the kinship structure similarities between Indian and Amazonian "tribes" are rather striking (again, marriage forms rare on a worldwide scale, similar Fst values, endogamy, kin terminological similarities), I would give this "social organization" a much deeper pedigree. Y-DNA Q and mtDNA D in India could be genetic signatures of this deep connection. Also, Berezkin found very striking matches in the Earth-Diver mythological motif between Munda and Muskogeans in North America. Some details are only uniquely found in these two groups and are missing in the whole of Siberia. This is not a perfect match, as Earth-Diver is not found in South America, but still worth noting.

German Dziebel said...

"the transformation of tribes and clans into jatis or endogamous castes."

Endogamous castes emerged through a different evolutionary path. It's the absence of strong unilineal descent groups such as clans that allowed for the development of castes from earlier bilateral groups such as kindreds in response to an interethnic conflict between incoming Aryans and indigenous Indians. This is one of the main differences between Australia and South Asia. India has a long history of endogamous bilateral kinship groups, whereas Australia is one of the classic areas of exogamous clans and moieties.

DocG said...

Maju: "However it means mainland SE Asia, Indochina and not Sundaland, where the various groups of Negritos seem isolated since the very initial expansion."

There are negritos in Malaysia and the Philippines.

"What I read is that due to persistance of endogamy (via the transformation of tribes and clans into jatis or endogamous castes), the founder effects of Paleolithic have been preserved to the present, what has not happened elsewhere to the same extent because endogamous castes do not exist elsewhere."

The endogamy preserved them, yes, but that doesn't alter the fact that what is being preserved are founder effects. In other words, we don't see a simple steady migration from Africa into South Asia, which would not have produced founder effects. Something broke them up into small groups and the ones that survived then either migrated or stayed put. The question is: when did this happen? And why?

But regardless of why, a bottleneck, or if you prefer, founder effect, reflects an event that greatly reduced the population, and under such circumstances certain cultural elements can be lost or change. Why is that so difficult to accept? If simple cultural drift were the engine of change, we'd expect to see a variety of different cultural norms in this region. Some languages might have gradually lost their tones and others might have kept them, so you'd see a mix. Same with musical practices. But this is not the case, that isn't how it happened.

DocG said...

Maju: "The "Australoid" tag for various Asian peoples' morphology (not just Indians but also SE Asians, Ainu and I'd dare say many Caucasoids) has also been contested often"

Yes, and it might not mean much. On the other hand, it might, and that could be an important clue. Why dismiss it? Certain tribal peoples in southern India certainly resemble Australians. And the look is very distinctive, not at all typical for Africans or even other Indian tribals. The genetic evidence demonstrated that a connection of some kind exists. It's not definitive or complete, but again, it could be a significant clue. So, again, why simply dismiss it? Why does it make you uncomfortable to consider such possibilities, Maju?

DocG said...

German: "Linguistically, there're no special matches between Australia and South Asia either"

Really? I just pointed to one: lack of tonally distinctive markers. For you this might be superficial, but it certainly seems important when we look at the world map, which is so clearly divided, at least in the Old World, between tonal and non-tonal regions. If there were such a clear division between different kinship systems, I hope you'd take notice of that.

DocG said...

German: "This is one of the main differences between Australia and South Asia. India has a long history of endogamous bilateral kinship groups, whereas Australia is one of the classic areas of exogamous clans and moieties."

What is your point, German? That India and Australia lack a common ancestor? Even according to your theory that would be the case.

India has a very different history from Australia altogether. India was in the middle of all sorts of complicated movements of all sorts of different peoples, India experienced the neolithic and also the development of stupendously complex and powerful civilizations. During all this time Australia was isolated and the Australians were wandering around in a continent consisting mostly of a vast desert.

So what does it tell us when you point out that they have such different kinship systems? Is that supposed to somehow refute the point I was trying to make, because if so, I don't get it.

German Dziebel said...

"The genetic evidence demonstrated that a connection of some kind exists. It's not definitive or complete, but again, it could be a significant clue."

I would like to emphasize that the very focus on finding similarities at all cost between India and Australia stem from Old-World-centrism. There're plenty of very specific pairwise connections (genetic, kinship, myths, languages, music) between India and America and between Australia and America, but nothing of this sort of specificity exists between India and Australia. Since American Indians have traditionally been thought of as a recent offshoot of a marginal Siberian group, these wider connections have been overlooked. They do fill in the bigger "gaps" in human evolution pretty well, as far as I can see.

German Dziebel said...

"lack of tonally distinctive markers."

Victor, 50% of world languages have tones. It's a very blanket category. Linguists attest to many cases of parallel evolution of tones. And you want me to consider the ABSENCE of tones a definitive sign of a specific historical link?

There're plenty of kinship patterns that are as superficial as tones and they arose by parallel evolution. Lewis H. Morgan in fact called a family of such patterns "Turano-Ganowanian" which was supposed to capture the historical connection between Iroquoian and Tamil kinship systems. These similarities later proved to be parallel innovations. Same for "Hawaiian" and Lineal patterns. In the "Genius of Kinship" I was working with deep structures that seem to break down in a lineal fashion, and this linearity suggests migrations.

German Dziebel said...

"So what does it tell us when you point out that they have such different kinship systems? Is that supposed to somehow refute the point I was trying to make, because if so, I don't get it."

Yes, South Asian and Australian kinship systems are very distinct. The earliest kinship forms (which spawned many derivaties) that are referred to in "The Genius of Kinship" as "Dravidian" and "Kariera" (following a convention) correlate with, respectively, endogamy and exogamy, the two broad types of marriage arrangements, again rooted in deep antiquity. Imagine we were talking about polyphony vs. monophony. Once you have very specific matches between North American Indian and Australian kinship systems, on the one hand, and South American Indian and South Asian kinship systems, on the other, the differences between Australian and South Asian kinship systems become too drastic to even think about deep historical kinship between these populations.

Of course, somehow these broad prototypes coalesce in the past (and they are most frequent and highly developed in the Americas) but in my "philosophy" this original unity (HBC, in your terminology) is harder to establish by reference to extant populations than more recent derivations from these broad prototypes. In my world, as time goes by, kinship fades away. In your world (fueled by suspect genetic phylogenies and dating techniques), some populations are timeless, others rootless.

Maju said...

There are negritos in Malaysia and the Philippines.

That's what I meant: Sundaland (and Wallacea) and not Indochina.

The endogamy preserved them, yes, but that doesn't alter the fact that what is being preserved are founder effects. In other words, we don't see a simple steady migration from Africa into South Asia, which would not have produced founder effects.

Steady migration? I have always considered the OoA as an outburst of sorts, not steady. The OoA was a single migrating episode, or at most a group of very few single migrating episodes. There's nothing steady about the OoA, otherwise Eurasians would have much greater genetic diversity.

Any migration, at least classical ones when the colony is founded by a bunch of people, produces founder effects. More so: demic expansions produce founder effects, at least initially.

A founder effect is the signature of a migration, not a steady flow probably but a simple single episode outburst, as most migrations were surely in the past, before "modern" technologies improved communications so radically, allowing for steady flows. A founder effect might also be caused by a hypothetical bottleneck but in that case, I'd expect the diversity to be much lower and the founder effect affecting not particular groups but huge areas - the whole subcontinent probably.

The only founder effects of subcontintental dimensions in South Asia are that of the M explosion, which also poured into Eastern Eurasia. This is not what you and Oppenheimer claim: but you say that localized FEs are signatures of a single massive bottleneck, and that doesn't make any sense to me.

If there was a bottleneck, it was before M, with M and N being the product of such a massive bottleneck (if it happened at all).

But regardless of why, a bottleneck, or if you prefer, founder effect, reflects an event that greatly reduced the population, and under such circumstances certain cultural elements can be lost or change. Why is that so difficult to accept? -

Because I do not see that bottleneck signature anywhere, except in music arguably. There's no genetic bottleneck in South Asia and there's no apparent correlation between tonal languages and P/B survivals in Greater Eurasia as whole.

At this stage, your reasoning, apparently so brilliant to detect musical signatures, has lost its critical ability and is just looking for any sort of supposed justifications for the music-based hypothesis, even if badly assembled. This is wishful thinking, not scientific reasoning.

Sorry for my bluntness but I must say what I see.

I understand that there is no clear evidence anywhere in genetics or linguistics of the bottleneck hypothesis, much less one that would support your musical model. To be efficient and produce a valuable theory you should, I believe now, look for another better (though maybe less hyper-simple) explanation.

Which one? I don't know at the moment. But the bottleneck model fails when checked against non-cantometric data.

(continues)

Maju said...

(continued from above)

Why dismiss it? [The "Australoid" tag].

Because if the various Negritos groups have been essentially isolated since the Great Eurasian Expansion, they are as distant to any other Eurasian population as they are among them (notwithstanding recent admixture). This also applies to the other so-called "Australoid" groups, whether in South Asia, Australia, Melanesia, Japan or Yemen: they are each a totally different population dating from the Great Eurasian Expansion, at least in essence.

The genetic evidence demonstrated that a connection of some kind exists.

No. One study suggested (not demonstrated yet) a connection (allegedly more recent than the Great Eurasian Expansion) between some South Asians and some Australian Aborigines might exist. Even if this would be proven it would affect only such minor lineages in both regions that it cannot account for any fundamental link, just an accidental one.

Nothing in that paper makes me change my opinion that, essentially, Australian Aborigines and South Indians are two totally different populations since at least 50 Kya.

Why does it make you uncomfortable to consider such possibilities, Maju? -

It does not make me "unconfortable" to consider any possibilities. Just that I consider them and have to dismiss them after the corresponding reality check.

German Dziebel said...

"Nothing in that paper makes me change my opinion that, essentially, Australian Aborigines and South Indians are two totally different populations since at least 50 Kya."

Agree.

"Because I do not see that bottleneck signature anywhere."

Agree.

"If there was a bottleneck, it was before M, with M and N being the product of such a massive bottleneck (if it happened at all)."

That's my gap scenario no 1, with the exception of the fact that I'd rather see African L lineages as products of rapid expansions after a bottleneck (with some additional demographic factors such as high prereproductive mortality coming into play, which would increase the rate of loss of M/N lineages and the survival rate of mutations occurring in migrating adults) and M and N as smoothly transitioning out of Africa (M and N in India, M and N in America, M and N in Oceania, sublineages are very widely spread geographically, monophony and polyphony coexisting, etc.).

DocG said...

Maju: "Nothing in that paper makes me change my opinion that, essentially, Australian Aborigines and South Indians are two totally different populations since at least 50 Kya."

Duh! Australia has apparently been totally isolated from the rest of the world (except N. Guinea) for at least that length of time, so why would I want to argue that? (Actually there is evidence of relatively recent contact during the Holocene, when the Dingo was introduced, but that need not concern us here I don't think.)

My point is that the ancestors of the modern Australian aborigines could be products of the same pop. bottleneck that caused the loss of tone language in South Asia, an event that surely would have taken place prior to 50,000 ya.

Maju said...

Duh! Australia has apparently been totally isolated from the rest of the world (except N. Guinea) for at least that length of time, so why would I want to argue that? (Actually there is evidence of relatively recent contact during the Holocene, when the Dingo was introduced, but that need not concern us here I don't think.)...

Essentially isolated, not 100% probably, yah. Anyhow, and just for the record, the dingo might well have been introduced in the Paleolithic, as we know dogs to be domesticated since at least 30 Ky BP.

The case is that such a time length is the same (roughly) as that of Eurasian spread and diversification, so this clashes with your suggestion of any kind of closer relationship with India or anywhere else.

My point is that the ancestors of the modern Australian aborigines could be products of the same pop. bottleneck that caused the loss of tone language in South Asia, an event that surely would have taken place prior to 50,000 ya.

The problem is that genetically (at least to some extent: haploid lineages) they are closest to East Asians (mtDNA N(xR) and Y-DNA C dominate). So it's difficult to argue that they belong to such population on these grounds. On the other hand Asutralian Aborigines, are the only Oriental population where blood group A is rather strong, as happens in West Eurasia (but it's a weak line of evidence, as blood groups have been argued to have evolved independently several times - in fact there are two A alleles and three 0 alleles).

German Dziebel said...

"On the other hand Asutralian Aborigines, are the only Oriental population where blood group A is rather strong, as happens in West Eurasia (but it's a weak line of evidence, as blood groups have been argued to have evolved independently several times - in fact there are two A alleles and three 0 alleles)."

The other "oriental" area is North America, where such populations as Blackfoot (unadmixed) show very high frequencies of blood group A, very similar to the frequencies of such a West Eurasian benchmark as the Saami. This has always been a puzzle, but in the light of high concentration of macrohaplogroup N in Australia, West Eurasia and North America it doesn't look odd anymore.