Wednesday, December 30, 2009

268. The Baseline Scenarios -- 44: The Migration

To my knowledge, there are no populations in the region between the Horn of Africa and what is now Pakistan that show any traces of GM (the Great Migration), either culturally or genetically. A recent study of the mitochondrial DNA of Saudi Arabs, for example, concluded as follows:

Although there is evidence of Neolithic and more recent expansions in the Arabian Peninsula, mainly detected by (preHV)I and JIb lineages, the lack of primitive autochthonous M and N sequences, suggests that this area has been more a receptor of human migrations, including historic ones, from Africa, India,
Indonesia and even Australia, than a demographic expansion center along the proposed southern coastal route ("Mitochondrial DNA structure in the Arabian Peninsula," Abu-Amero et al., BMC Evolutionary Biology, 2008).
It is difficult to assess what this absence of evidence might mean. It's possible the migrants bypassed this region, possibly by boat, or that there was simply never any expansion northward from the coast, into what would have been hostile desert country, or that the descendants of any colonies left behind in this area simply died out over time or were possibly wiped out or assimilated by groups that entered the area much later. In any case, we see little or no evidence of their presence until we reach the South Asiatic peninsula (present day Pakistan and India).

The differences are cultural as well as genetic, but cultural issues, aside from those centered on language, are rarely discussed in the population genetics literature. Musically there is an enormous gap, especially with vocal music, which is almost exclusively monophonic, with great emphasis placed on solo singing, often of a highly embellished and even virtuosic nature. In strong contrast to the typical voices of Subsaharan Africa, relaxed and open-throated, the Arabic and Persian peoples of this region tend to sing with very tense, narrowly constricted voices.

Whereas most Subsaharan music emphasizes relatively short, simple and unembellished motives, often tossed back and forth among individual singers or in call and response fashion, the music of the Middle East (and North Africa) favors long, elaborately woven melodic lines, often of great intricacy and subtlety, accompanied by instruments playing variations of the same line in a style usually referred to as "heterophony." The contrast with Sub-Saharan Africa isn't universal, and there are in fact certain types of singing based on short phrases tossed back and forth in call and response style -- but almost always with tense voices and in unison, as polyphonic singing is extremely rare if not entirely absent from the entire region. This type of vocalizing is quite different from the open-throated, highly group-oriented, often elaborately interwoven part singing that would probaby have been characteristic of HMC, and seems likely to have developed during a much later period.

Moving on to the next stage of our hypothetical GM journey, and acknowledging that the point I'm about to make will be hotly contested by some, the genetic evidence for HMP and its descendants throughout the entirety of South Asia appears problematic. And the same is true for the cultural evidence. I've already displayed one phylogenetic map for mitochondrial DNA (mtDNA), based on a revised molecular clock that places all the M clades for South Asia at later date than those for Southeast Asia, a very surprising result considering that South Asia represents, in purely geographic terms, an earlier stage along the presumed GM path (see the illustration from the article Correcting for Purifying Selection: An Improved Human Mitochondrial Molecular Clock, as reproduced in Post 262). The molecular clock is undergoing many revisions these days, so any one version of it cannot be regarded as definitive.

But there are other results as well, also reported in recent posts, that present what can only be described as an equivocal picture for this entire region. For example, as reported in Post 264, Cordaux et al., who could not, in 2003, find "any unquestionable genetic signature of the ~60,000 year old migration from Africa to Sahul following the postulated southern route," suggested that "in India the genetic traces of early migrations along the southern route" might have been "erased by the subsequent migrations which shaped the present-day mtDNA gene pool of India."

On the other hand, a more recent study published in 2009, Updating [the] Phylogeny of [the] Mitochondrial DNA Macrohaplogroup M in India, by Adimoolam Chandrasekar et al., concluded that "the basal diversity (37 nodes) and founder ages (57,000–75,000 years) of macrohaplogroup M in India reveals [an] initial settlement of [the] African exodus in India. . . several Indian mtDNA M lineages are deep rooted and [have an] in situ origin." Since the Chandraseka paper doesn't challenge the findings of the Cordaux group, it's hard to know what to make of the above conclusion, or of any of the genetic findings for this very complex region, which must be regarded as provisional until larger population samplings become available.

As I've already argued on this blog (e.g., Posts 259 et seq.), certain cultural evidence is consistent with the more problematic genetic findings, both pointing to a serious gap, to some extent an extension of the gap we found in the Middle East, continuing throughout the length and breadth of South Asia as well. As far as the musical evidence is concerned, stylistically both the "classical" court and religious-oriented music and the more "folk-oriented" village music of both Pakistan and India, together comprising an extraordinarily homogenous style family, can be seen as an extension of the traditions of Middle Eastern music, as described above. I recall a private conversation years ago with a leading authority on Indian music, Harold Powers, who confided in me his conviction that Indian classical music was heavily influenced by Moslem traditions, a theory he dare not publish, as it would have made it impossible for him to continue his research in India.

The situation with the tribal peoples of India (and also many of the lower caste groups, who probably originated as tribals) is more complex, largely because they have not been subject to anywhere near the degree of anthropological or ethnomusicological attention as tribal groups to the East and Southeast. Many are, nevertheless, represented in the Cantometric database, and the picture presented there is of a fairly uniform mix of relatively simple singing styles, characterized in many cases by call and response patterns, and in a few cases polyphonic, but completely lacking any sign of what I've called the "African signature," i.e., hocket and/or interlock, yodel, counterpoint, etc., and no sign of the sort of hocketing instrumental ensembles so commonly found in both Africa and greater Southeast Asia.

There is also some very interesting, though inconclusive, linguistic evidence consistent with the same gap, which might or might not be significant, as well as a possible series of major population bottlenecks produced by the eruption (ca 74,000 ya) of Mount Toba, which could have produced such a gap in both the genetic and cultural picture. I won't go into these questions any further here, since I've already discussed them at length in various earlier posts. We should nevertheless keep all of the above in mind as we attempt to follow the peregrinations of our pioneering Out of Africa migrants along the southern route.

(to be continued . . . )

11 comments:

German Dziebel said...

"the lack of primitive autochthonous M and N sequences."

A critical quote from this paper is the following: "Although a newly defined clade L6 in Yemenis, with no close matches in the extant African populations, could suggest an ancient migration from Africa to Yemen [30], the lack of N and/or M autochthonous lineages left the southern route without genetic support."

L6 must be that ancient trace that connects West Asia and Africa. The phylogeographic evidence, however, again points to a migration not out of Africa but into Africa (unless L6 will eventually be found in Africa) and suggests that some mutations, not seen on M and N lineages, must have emerged in advance of Africa as a result of bottlenecks.

German Dziebel said...

"The situation with the tribal peoples of India (and also many of the lower caste groups, who probably originated as tribals) is more complex, largely because they have not been subject to anywhere near the degree of anthropological or ethnomusicological attention as tribal groups to the East and Southeast. Many are, nevertheless, represented in the Cantometric database, and the picture presented there is of a fairly uniform mix of relatively simple singing styles, characterized in many cases by call and response patterns, and in a few cases polyphonic, but completely lacking any sign of what I've called the "African signature," i.e., hocket and/or interlock, yodel, counterpoint, etc., and no sign of the sort of hocketing instrumental ensembles so commonly found in both Africa and greater Southeast Asia."

This is absolutely consistent with mtDNA or Y-DNA, as these systems show no "African signatures" in Indian gene pool(s). The very fact that you find "African signatures" in music contradicts genetic evidence, which shows that none of the oldest Sub-Saharan lineages (Y-DNA A and B, mtDNA L0, L1, L2, etc.) are found anywhere outside of Africa. By the standards of preservation of ancient "signatures" in music the whole genetic evidence for an out of Africa migration is one huge gap.

DocG said...

German: "By the standards of preservation of ancient "signatures" in music the whole genetic evidence for an out of Africa migration is one huge gap."

The gap works both ways, German. Whether you see early humans migrating eastward or westward across Asia, the same gap exists and must be taken into account. It does seem consistent with the multiregional model, however, as is the old dual model of musical origins you've been urging me to revive. Have you transformed into a multiregionalist? :-)

I like seeing such a gap because it reminds us that no simple and straightforward theory is likely to be correct when we are attempting to reconstruct such a long and complex history. As I see it, the gap is a huge clue that could tell us a great deal, if we treat it as such, rather than as an obstacle.

German Dziebel said...

"It does seem consistent with the multiregional model, however, as is the old dual model of musical origins you've been urging me to revive. Have you transformed into a multiregionalist."

Revive but only as a way to give a more balanced account to human music variation in which one macrotype doesn't simply devolve into another but rather both (monophony and polyphony) have been retained in modern populations since the original migration (out of America or out of Africa, doesn't matter at this point). Then within the macrotypes you could create scales of complexity.

I believe in the single origin of modern humans. I think genetics is very clear on this. However, out of America does take into account some strong empirical points that have been made by Multiregionalists. One of them is the existence of "oriental" biological and cultural traits that can't be explained as product of an out of Africa migration. I take these traits (as well as (paleo)biological continuities with Asian Homo erectus as an ancient heritage depleted as populations moved from east to west into Africa and Europe. Europe inherited monophony, Africa polyphony but both of them are found interspersed in Asia and America. You could still make an out of African claim but I would keep monophony as the other tradition carried out of Africa.

If you look at the phylogeography of mtDNA and Y-DNA (forget about the dates for a sec), you'll see some very strong "oriental" traits such as haplogroups A, D, C, B, X that are frequent in Asia and America and then slowly "deteriorate" as one moves west. India still has A, B, C and D (as well as Y-DNA Q), North Africa and Middle East have X (as well as Y-DNA Q), while Europe and Sub-Saharan Africa barely so (X comes to Europe late). Certain "western" lineages were formed along the east-to-west gradient, including M and U, with offshoots into North and East Africa. Somewhere in the Middle East L lineages began to evolve, and those were taken deep into Sub-Saharan Africa.

German Dziebel said...

"As I see it, the gap is a huge clue that could tell us a great deal, if we treat it as such, rather than as an obstacle."

Victor, what happened to your love for Occam's Razor? Again, you're falling into extremes: first you celebrate the fact that beehive huts have survived unscathed, now you see a potential in seeing whole linguistic and musical traditions once common in whole (sub)continents go missing. My method is very simple indeed: monophony and polyphony are widely distributed around the globe, both are found in areas with deep archaeological records. Why not assume that they've been competing with each other throughout human history, with some populations generalizing and deepening monophony and others generalizing and deepening polyphony? I'm not saying this is exactly what happened but why do you not consider it a viable alternative?

DocG said...

German: "Victor, what happened to your love for Occam's Razor?"

Occam's Razor is a double edged sword -- er, razor. :-)

Simplicity, yet, but ONLY insofar as compatible with the evidence. The evidence presents us with a gap, so we have to find the simplest solution consistent with that gap.

I don't know how much of a fan of cosmology you are, but the bottleneck proposed by Oppenheimer is very much like the notion of cosmological inflation proposed by Alan Guth. There may not be much direct evidence for it (as yet) but its explanatory power is so great that we can't do without it (at least for the time being, till something better comes along). This is how science works, not on the basis of certainty and proof (as with math) but on the basis of what seems most simple and reasonable at the time, given the current state of our knowledge.

"monophony and polyphony are widely distributed around the globe, both are found in areas with deep archaeological records. Why not assume that they've been competing with each other throughout human history,"

But they are distribugted according to very specific patterns that cannot be ignored and must be accounted for. You seem to want me to fabricate some thery that suits you out of thin air. I can't do that.

German Dziebel said...

"You seem to want me to fabricate some thery that suits you out of thin air."

Monophony is dominant in Australia, North America, all of Eurasia, including South Asia and West Asia, North Africa, as well as sporadically elsewhere. Air doesn't get thicker than that.

DocG said...

German: "Monophony is dominant in Australia, North America, all of Eurasia, including South Asia and West Asia, North Africa, as well as sporadically elsewhere. Air doesn't get thicker than that."

You make an interesting point and it's useful to refute it, so thanks.

The most archaic survivals are NOT necessarily the most widespread, because there are many relatively recent factors that could lead to such a distribution. For example the wide distribution of plough agriculture, which we know is a relatively recent development.

The signs of archaic survival were outlined by Edward Sapir many years ago. See the end of Post 242 for the full quote:
"In a general way, a culture element whose area of distribution is a broken one must be considered as of older date, other things being equal, than a culture element diffused over an equivalent but continuous area."

It is traditional vocal polyphony, not monophony, that has such a distribution, and on a worldwide scale. It is almost always found in isolated pockets, refuge areas, surrounded by vast seas of monophony, just as hunter-gatherers are surrounded by vast seas of agriculture.

German Dziebel said...

Victor,

This is another occasion when you expose the flaws of your method.

"It is traditional vocal polyphony, not monophony, that has such a distribution, and on a worldwide scale. "

Africa has no monophony, it has only polyphony. It means there's no "gapped distribution" here. There must have been plenty of cases of diffusion in African history. Sapir's quote can therefore be used against you.

The serious problem that Sapir's quote poses for your scenario is that BOTH polyphony and monophony have gapped distributions vis-a-vis each other. We can see it along the putative "southern route": polyphony (Africa), monophony (West and South Asia), polyphony (Southeast Asia and PNG), monophony (Australia, North America), polyphony (South America).

"It is almost always found in isolated pockets, refuge areas, surrounded by vast seas of monophony..."

So, why do you think that what's in the "pocket" is somehow ancestral to what's outside of the "pocket"? It's like saying that a pocket can exist without a pant. They are both equally old, only monophony has put some geographic pressure on polyphony in the last 10K years. (In Sub-Saharan Africa, the opposite may have happened.) Australia is a monophonic pocket. North America can be seen as a refugium in its own right.

With your reasoning, you have to postulate, first, the conversion of a polyphonic population into monophony; and then the encroachment of a more recent tradition on an older tradition (which originally has been more widely spread, hence better entrenched areally. (Incidentally, genetics makes a similar mistake: not only that a population that supposedly left Africa didn't take the most widely spread, areally entrenched and oldest lineages, it waited until they would evolve into M and N and then would scrape off all of the M and N lineages out of Africa only to redeposit them back there 20K later, namely U6 and M1).

I can clearly see how the two musical traditions may have competed with each other over the past 50K, and the past 10K have redrawn the maps, but why do you assume that polyphony is older than monophony? And I mean it on other grounds than just gene trees.

To sum up, it's a matter of the glass being half empty or half full. You always show a bias toward one of the two options, and then when I start interrogating the reasons for your choice, you only have volcanoes behind your word.

DocG said...

German: "Africa has no monophony, it has only polyphony. It means there's no "gapped distribution" here."

You want to see rigid rules everywhere, German, but there are no such rules, only guidelines, which must be applied differently in different historical and cultural contexts. This is why Sapir included the phrase "other things being equal . . ."

Africa is an interesting case, so I'm glad you brought this up as it does require clarification. Actually there IS monophony in Subsaharan Africa -- there are certain tribes whose singing is consistently in unison rather than polyphonic, and I've been investigating the distribution of unison in SSAfrica to see if I can figure out why certain groups sing this way and others don't.

But the context of the distribution is clear: we do NOT find pockets of unison singing among marginalized groups living in refuge areas, as is the case for vocal polyphony in Europe and Asia. And while it's true that we don't find isolated pockets of polyphony in Africa either, the history of Africa is very different from that of the rest of the world in this and many other respects.

It's in Asia where a very active and agressive culture emerged that became almost exclusively monophonic musically, and the evidence points to the ultimate expansion of this culture to dominate much of the "civilized" world in Europe and Asia, a culture that spread agriculture but also marginalized the older populations of Europe and Asia, so that these older groups, still singing polyphonically in many cases, can now be found almost exclusively in refuge areas.

It was only in the North that Africa was affected by this development, and it is in North Africa that we find monophony almost exclusively.

As far as the Americas are concerned, the history is very different, of course. But here too we find evidence that the earliest immigrants, possibly the farthest reach of the Out of Africa migration, sang polyphonically, since here too polyphonic vocalizing (along with hocketed wind ensembles) are found only sporadically, in certain pockets, surrounded by monophony, which is dominant in South America as well.

German Dziebel said...

"It's in Asia where a very active and agressive culture emerged that became almost exclusively monophonic musically, and the evidence points to the ultimate expansion of this culture to dominate much of the "civilized" world in Europe and Asia, a culture that spread agriculture but also marginalized the older populations of Europe and Asia, so that these older groups, still singing polyphonically in many cases, can now be found almost exclusively in refuge areas."

You also believe that the original "loss" of polyphony happened as a result of the Toba eruption. You seem to connect everything from a volcano eruption at 70K to the spread of agriculture in the past 10-5K into one single anti-polyphony plot. On top of a catastrophic idea you put a conspiracy idea.

Plus how did this "aggressive" Asian culture end up in Australia?

Australia is a refugium, and its non-polyphonic. On a more micro-scale, the mountains of South Siberia is a refugium but the music of Tofa, Tuva and Altaians is monophonic. Na-Dene have probably re-expanded in the Subarctic out of an Ice Age refugium, and again their music is monophonic (simple monophony, according to Nettl).

You didn't really answer any of my criticisms. Specifically, why would a tradition that survived in a pocket be older than a tradition that survived outside of the pocket? Especially, since you claim that the tradition found outside of the pocket is the result of the loss of a tradition found inside the pocket. This is a circular argument. Your reliance on "geographic distribution" without an evolutionary model leads you into a circular argument that's designed to favor an African tradition over the rest.

Finally, empirically polyphonic traditions show quite a bit of resilience to the encroachment of monophony. This is evident in the Caucasus, for instance, where Christianity attempted to impose monophonic hymns onto traditional polyphony, succeeded for a couple of centuries but then polyphony rebounded and began to dominate the liturgical process.

Also, in Western Christian music polyphony expanded from 14th century on. Prior to that it was considered impious. Empirically, again, we observe that polyphony is a resilient tradition. It's dramatic loss throughout human history would be hard to prove.

Why couldn't two traditions co-exist throughout human history and occupy different niches in cultures? Men vs. women, instrumental vs. vocal, secular vs. religious, etc.