Flickr

  • www.flickr.com
    Abiola_Lapite's photos More of Abiola_Lapite's photos

« Egyptian Smoke and Mirrors | Main | In Praise of Hungarian Notation »

May 13, 2005

Comments

dearieme

A bit surprising that a single band would galivant to far SE Asia and then that members wouldlater treck back to Europe and the Near East? Would it not be more plausible that people would migrate around in various favourable areas along the route, while the growing population fills the niches thus ensuring that part of the group keeps pushing on into new territory? Still, if the evidence says not....

Abiola Lapite

"Would it not be more plausible that people would migrate around in various favourable areas along the route, while the growing population fills the niches thus ensuring that part of the group keeps pushing on into new territory?"

That's just what the evidence suggests, that the population of emigres expanded along the coastline all the way to South-East Asia, and only when the familiar coastal expansion route was filled did populations begin to move inland.

dearieme

Thanks for that: it's clearer to me now.

 razib

that there is absolutely no support for multiregionalism in any shape or form

ok abiola, this depends on how you word it. i'm in a hurry now, but here is my current "position," i think there is the strong possibility of some alleles having an extra-recent-out-of-african (i will post the links here when i can look them up, i have literally 5 minutes), and there are also problems with the model of assumed neutrality for mtDNA (recent work suggests that europeans who carry haplogroup U have higher cancer rates, and there has been research which shows some populations seem to inherit "longevity" through their mother specifically because of differences in mtDNA degeneration over their lifetime).

in any case, if readers are interested in the text of the paper the article is based on, i've cut & pasted it over at my blog:
http://www.gnxp.com/MT2/archives/003982.html

(feel free to ignore the irritable commentary)

additionally, the work in this paper seems to be utilizing a model promoted by stephen oppenheimer in his book "the real eve." i don't really agree with a lot of things in it, but it is pretty well chock full of data, so i think a typical lay person will find it really interesting. in sum, oppenheimer's position (which they allude to in the paper, though not in detail) is this:

1) about 80000 BP moderns leave africa
2) they swing across the southern eurasian coastline
3) just as they do the "toba event" blows up, and wipes out the population of central-east india.
4) this bifurcates eurasia's mtDNA lineages into two broad streams, a west eurasian and east eurasian one.
5) oppenheimer's thesis rests upon the interpretation of indian genetic history by an estonian geneticist, toomas kisivild. i think there is a greater chance than not that kisivild is correct...but there is another camp that has also been producing papers, and arguing that the south asian population is far more admixed with many more recent linages, especially on the Y chromosome.
6) this matters because oppenheimer's thesis is mtDNAcentric, but he wants to make the case that little admixture has occurred after 20,000 years BP, because that messes up his narrative.

ok, in sum, basically oppenheimer's thesis is that africa is the human race's grandmother, but for most of humanity (that is, non-african), india is their mother, with western india giving rise to western eurasians and eastern india giving rise to eastern eurasians.

already overlong, i will add two final things

a) oppenheimer gives a very ancient age for "recent" out-of-africa. i don't think this is untenable, but most out-of-africanists would disagree.
b) i don't think his argument about the southern expansion being singular is that revolutionary or contrarian. i don't think paleoanthropologists have a good idea and the models aren't fixed.
c) if you read oppenheimer's book, keep in mind that the most disturbing aspects was that i noted he left out the examination of a few important papers which strongly contradict his thesis about india (i checked publication dates, the papers were published YEARS before some of the ones he references). i don't want to suggest he had to agree with said papers, but he should have engaged them and put them out there for the lay readership (one of the papers was done under underhill, with the help of spencer wells, so they weren't smallfry, oppenheimer just left it out because it would have added many more pages of explanation to his book).

more later....

 razib

3 final things. no editing :)

Abiola Lapite

"i think there is the strong possibility of some alleles having an extra-recent-out-of-african"

How would one even go about proving this? Just because a particular allele doesn't fit neatly into the common tree for the rest of humanity needn't mean that it came from an extra-human source: it could simply be the sole surviving leaf node a branch which was part of the original "Out of Africa" event but has since gone extinct - with the rest of humanity lying on the other branch.

"there are also problems with the model of assumed neutrality for mtDNA"

I'm doubtful that there's any such thing as a completely neutral mutation anyway, mtDNA or otherwise, but in any case, this doesn't affect mtDNA's mode of inheritance. What do you have in mind, some sort of selective sweep in European populations which just happened to wipe out all evidence of non-African ancestry on that locus? If this is true, it's peculiar that there should also be no evidence of non-African ancestry on the Y-chromosome.

The problem with multiregionalism is that it can never be entirely ruled out in some very weak form or other, and its supporters can always say that we've just not looked hard enough, or the genetic impact has been too small for samplers to pick up, but this is a classic example of what Lakatos called an immunizing strategy, and the statistical evidence against the multiregional model at this point is so strong that it simply isn't worth taking seriously.

"3) just as they do the "toba event" blows up, and wipes out the population of central-east india.
4) this bifurcates eurasia's mtDNA lineages into two broad streams, a west eurasian and east eurasian one."

This sounds like a pretty radical thesis to me, seeing as I've never laid eyes on a single study to suggest a cleavage of Eurasian populations dating that far back.

Jim

So we Europoids didn't inherit our heavy eyebrows and big noses from the Neanderthals. There must be some adaptive advantage to those features, if they evolved twice.

Is the "Toba Event" necessary to explain the genetic differences between eastern and western Eurasians? Aren't the gigantic distances sufficient to explain genetic drift?

 razib

jim, the argument for total replacement is probably strongest in europe IMO (with the most doubts showing up in east asia). the original out-of-africanists who draw from paleontology tend to demur when asked about eastern asia because of the lack of archaic h. sapiens fossils (though rumor has it that a neandertal specimen was found in china last year, but it has been in press for months it seems), but focus specifically on europe because of the sharp discontinuity. on the other hand, multiregionalists tend to focus on east asia (shovel teeth) and southeast asia and australia (robusticity-though here they often compare southeast asian erectines with modern australian aboriginals, rather than arguing for geographic continuity in southeast asia). i don't buy into old-school multiregionalism because i think the phenetic aspects are very hit and miss (and if you had to ask me, i would still bet that 90-99% of non-african alleles are identical-by-descent to those of individuals who left northeast africa 50-100 BP).

more on abiola's questions later when time permits.

 razib

starting the i.b.d. clock from the moment moderns started leaving africa.

 razib

"The problem with multiregionalism is that it can never be entirely ruled out in some very weak form or other, and its supporters can always say that we've just not looked hard enough, or the genetic impact has been too small for samplers to pick up, but this is a classic example of what Lakatos called an immunizing strategy, and the statistical evidence against the multiregional model at this point is so strong that it simply isn't worth taking seriously."

you are very correct. i think out-of-africa replacement is a good null hypothesis because it is falsifiable. the point though is that there aren't just two alternatives out there: out-of-africa vs. multiregionalism. genuine worldwide "anagenetic" transition is too squishy to really explore with any level of precision. rather, there are subdivisions within the out-of-africa models. the paper above works with the one-southern-exit-route-only. there is also the two-routes-out-of-africa, and within this there are debates about the population of origin for those who participated in the northern route (whether that be "modern" commmunities excavated in what is today the nile delta, or later africans. the question is complicated by the fact that there seem to periodic expansions of african fauna, including humans, followed by die-backs around the levant and they don't have enough granuality to figure out where the successive modern populations have a relationship to each other or are separate re-expansions).

as for proof...well, there won't be definitive proof.

but, there are strong suspicions that moderns and non-moderns fucked:
http://biology.plosjournals.org/perlserv/?request=get-document&doi=10.1371/journal.pbio.0020340
and papers like this are starting to trickle out:
http://mbe.oupjournals.org/cgi/content/abstract/msi013v1
and this article also suggests ancient coalescence times for alternative alleles:
http://www.nature.com/ng/journal/v37/n2/abs/ng1508.html
(this could be balancing selection of course)

there are all sorts of models wandering around out there now that NRY data and autosomal data is getting really rich. one aspect that i am curious about is the possibility of migration and back migration between africa and eurasia before 100-200 BP. the bones & stones people are also suggesting that the transition between moderns and neandertals in europe is more complex than we thought, there might have been 2-3 expansions of anatomical moderns that did not fix, until the last expansion finally pushed the neadertals aside.

"This sounds like a pretty radical thesis to me, seeing as I've never laid eyes on a single study to suggest a cleavage of Eurasian populations dating that far back."

well, i don't buy his thesis. but if you are curious, he is positing the M vs. N mtDNA superclades (both descended from L1, which is shared between africans and non-africans i recall correctly) represent the western and eastern branches of eurasian humans (some back-migration to africa via the YAP haplotype as well). but the archeology seems to lean against this thesis, most of the early chapers were devoted to why there isn't much evidence that is unchallenged of moderns in south-southeast asia around 80 BP during the toba event.

Jim

"one aspect that i am curious about is the possibility of migration and back migration between africa and eurasia before 100-200 BP."

And since then too. There has to be some explanation for the Egyptians being where they are between two obviously Europoid populations in North Africa and Asia.

The comments to this entry are closed.

Notes for Readers