The Unz Review - Mobile
A Collection of Interesting, Important, and Controversial Perspectives Largely Excluded from the American Mainstream Media
 Gene Expression BlogTeasers
It All Began at Ararat, and Esau's Revenge

ag

51IZQjMbVlL._SY344_BO1,204,203,200_ About four years ago the genome blogger Dienekes Pontikos published a post, The womb of nations: how West Eurasians came to be. The argument was that the genetic variation we see around us across western Eurasia and northern Africa has its ultimate roots in the the structure that was extant in the ancient Near East, near to the zone of initial agricultural innovation in the hills above the modern nations of Syria and Iraq. It was similar to the thesis of Peter Bellwood in First Farmers: The Origins of Agricultural Societies. The model as outlined by Dienekes suggested that local substrate was absorbed as these initial agriculture societies rapidly swept outward from their initial focal zone. In the broadest strokes I do think that he, and Bellwood, were correct from what we now know.

Ancient DNA in Europe strongly indicates massive replacement. But, there is also suggestion of admixture with the local substrate. And, unlike the stylized model of Bellwood, it seems that there were multiple migrations after the initial pulse which reshaped the genetic and cultural landscape of human societies in the wake of agriculture. Here is the abstract for Lazaridis et al. at ASHG:

It has hitherto been difficult to obtain genome-wide data from the Near East. By targeting the inner ear region of the petrous bone for extraction [Pinhasi et al., PLoS One 2015] and using a genome-wide capture technology [Haak et al., Nature, 2015] we achieved unprecedented success in obtaining genome-wide data on more than 1.2 million single nucleotide polymorphism targets from 34 Neolithic individuals from Northwestern Anatolia (~6,300 years BCE), including 18 at greater than 1× coverage. Our analysis reveals a homogeneous population that is genetically a plausible source for the first farmers of Europe in the sense of (i) having a high frequency of Y-chromosome haplogroup G2a, and (ii) low Fst distances from early farmers of Germany (0.004 ± 0.0004) and Spain (0.014 ± 0.0009). Model-free principal components and model-based admixture analyses confirm a strong genetic relationship between Anatolian and European farmers. We model early European farmers as mixtures of Neolithic Anatolians and Mesolithic European hunter-gatherers, revealing very limited admixture with indigenous hunter-gatherers during the initial spread of Neolithic farmers into Europe. Our results therefore provide an overwhelming support to the migration of Near Eastern/Anatolian farmers into southeast and Central Europe around 7,000-6,500 BCE [Ammerman & Cavalli Sforza, 1984, Pinhasi et al., PLoS Biology, 2005]. Our results also show differences between early Anatolians and all present-day populations from the Near East, Anatolia, and Caucasus, showing that the early Anatolian farmers, just as their European relatives, were later demographically replaced to a substantial degree.

A somewhat different abstract was submitted for a meeting in Germany:

We study 1.2 million genome-wide single nucleotide polymorphisms on a sample of 26 Neolithic individuals (~6,300 years BCE) from northwestern Anatolia. Our analysis reveals a homogeneous population that was genetically similar to early farmers from Europe (FST=0.004±0.0003 and frequency of 60% of Y-chromosome haplogroup G2a). We model Early Neolithic farmers from central Europe and Iberia as a genetic mixture of ~90% Anatolians and ~10% European hunter-gatherers, suggesting little influence by Mesolithic Europeans prior to the dispersal of European farmers into the interior of the continent. Neolithic Anatolians differ from all present-day populations of western Asia, suggesting genetic changes have occurred in parts of this region since the Neolithic period. We suggest that the language spoken by the homogeneous Anatolian-European Neolithic farmers is unlikely to have been the same as that spoken by the Yamnaya steppe pastoralists whose ancestry was derived from eastern Europe and a different population from the Caucasus/Near East [Haak et al. 2015], and discuss implications for alternative models of Indo-European dispersals.

And now, David has a post up, Yamnaya’s exotic ancestry: The Kartvelian connection. You can read the post yourself, but he presents TreeMix results which strongly imply that the Near Eastern gene flow into the Yamnaya population derives from one with a relationship to modern Kartvelian groups, the most prominent of which are Georgians. This is not entirely surprising. I’ve seen similar things, though I’m not sure what to think of this. Going back to Dienekes’ post he suggests that “It is, perhaps, in the ancient land of the Colchi, protected by the Black and Caspian seas, and by tall mountains on the remaining sides, that something resembling the ur-population survived.” That is, the people of the trans-Caucasian region may preserve elements of the deep ancestral heritage of the Middle East which was to some extent effaced by later migrations. I have some Assyrian genotypes, and this population has strong affinities with Armenians and Georgians. Additionally, I believe that much of the west Eurasian heritage in South Asians is from this same region and source. In Genetic Evidence for Recent Population Mixture in India the authors note that “For all 45 Indian groups on the Indian cline…we find that Georgians along with other Caucasus groups are consistent with sharing the most genetic drift with ANI [Ancestral North Indians -Razib].” Of course in South Asia there was a second intrusion of this sort of ancestry, but in this case as part of the admixed compound brought by Indo-Europeans, who interleaved Near Eastern, ancient European, and Central Asian, all in one.

There are two elements here which need to be noted. First, a genetic one in regards to the Middle East. In Europe over thousands of years the heritage of the first farmers waned, as that of the hunter-gatherers experienced some resurgence, and eventually the Indo-Europeans from the steppe overran vast territories. In the Middle East analogous groups which expanded out of the ancient hillocks and swept south were overwhelmed by a later Arabian reflux. These are perhaps prefigured by the drifting of Amorites into the cities of ancient Mesopotamia 4,000 years ago, and more recently the arrival of Arabs who had been outside of the limes of civilization into the worlds of eastern Rome and western Persia.

Finally, this is about the nature of culture and the advantage which a particular toolkit does, or doesn’t, provide a people. The rollover of European hunter-gatherers, or the fact that “Ancestral South Indians” (ASI) don’t exist in pure unadmixed form, point to the advantages which a fully elaborated agricultural cultural system provided the farmers. But, during the initial stages of the development of this toolkit it does not seem that there was any particular advantage to single group in a narrow delimited zone. To be more concrete, the linguistic diversity of the Caucasus region, or what we see and know from the edges of history in the Near East, may be close to the reality of the period of the early Neolithic in the Near East. Different polities with radically different languages, and likely divergent genetics, were all crystallizing the new lifestyle in a cheek by jowl fashion in the hills between the Tarsus and Zagros mountains. At some point though the system became powerful enough that it was portable and extendable in space. Rather than engaging in inter-group competition against populations which were comparable, the most successful route was to expand outward into the vast “unoccupied” zones of hunter-gatherers.

Because of the nature of climate west and east migrations were probably easiest and more rapid. Likely this resulted in a near total translocation of the ancestral cultures on far shores or distant horizons. But as the groups pushed north or south the agricultural toolkit was less well suited, and so synthesis with the native substrate was necessary. I believe that Indo-European and Afro-Asiatic may both be language families of hunter-gatherer populations who absorbed migrating farmers into a radically different ecology where the competitive playing field was much more level than in the initial zone of expansion.

 
• Category: History • Tags: History 
Email This Page to Someone

 Remember My Information



=>
Commenters to Ignore...to FollowEndorsed Only
[Filtered by Reply Thread]
  1. “I believe that Indo-European and Afro-Asiatic may both be language families of hunter-gatherer populations who absorbed migrating farmers into a radically different ecology where the competitive playing field was much more level than in the initial zone of expansion”.

    Interesting idea. I think you are onto something there.

    Read More
    • Replies: @LevantineJew
    I have my own speculation about ethnogenesis of Armenians and Kurds, which happened at the end of the Neolithic period.

    Armenians are descendants of the eastward migrating Indo-European commoners, who end-up being ruled by the Urartian (Caucasian) nobility.

    Kurds are to the contrary descendants of the native Hurrian (Caucasian) commoners who were conquered by Indo-Aryan Mitanni and later by Median nobility.


    "Founded by an Indo-Aryan ruling class governing a predominantly Hurrian population" [1]
     
    Interestingly Mitanni in Hittite written with the cuneiform KUR [2] (cognates with Kurd) which usually referred to the Zagros mountains.

    [1] https://en.wikipedia.org/wiki/Mitanni

    [2] https://en.wikipedia.org/wiki/Kur

    ReplyAgree/Disagree/Etc.
    AgreeDisagreeLOLTroll
    These buttons register your public Agreement, Disagreement, Troll, or LOL with the selected comment. They are only available to recent, frequent commenters who have saved their Name+Email using the 'Remember My Information' checkbox, and may also only be used once per hour.
    Sharing Comment via Twitter
    http://www.unz.com/gnxp/it-all-began-at-ararat/#comment-1173584
    More... This Commenter This Thread Hide Thread Display All Comments
  2. So are you saying that Indo-European was the language of European hunter-gatherers, later adopted by farmers who moved in (from Anatolia?)? Did I get that right?

    Read More
    • Replies: @Razib Khan
    sort of. i actually believe that the most likely, albeit speculative and without much confidence, that IE and uralic languages are central eurasian in provenance, one of the language groups of the "ancestral north eurasians" (who contribute ~50% of the ancestors of eastern hunger-gatherers who mixed with the arrival near easterners).
  3. Indeed, I’ll be extremely surprised if it turns out that Indo-European didn’t come from North Eurasian foragers.

    Btw, rumor has it that Broad MIT/Harvard have some interesting new ancient samples. You might be able to get something out of them in person, if they’re feeling generous.

    Read More
  4. It could have happened like that, or you could have had the HG communities adopting one the farmer languages, at the same time as fusing into their communities, which would have been a radical change for both.

    Ultimately it seems like we can only really reconstruct the place (climate, time) in which a language was spoken prior to its expansion.

    Read More
    • Replies: @Razib Khan
    It could have happened like that, or you could have had the HG communities adopting one the farmer languages, at the same time as fusing into their communities, which would have been a radical change for both.

    to be clear, i think the ethnographic record that HG -> pastoralist transition can be pretty easy/fluid is the dynamic i'm alluding to. both IE and afro-asiatic groups are fusions of HG+farmer. but i posit that the HG had much more influence than would otherwise be the case because of the fact that they could take to pastoralist economies far more easily than they would have to sedentary agriculture.
  5. Davidski comments on his Kartvelian post that Adyghe and generally contemporary North Caucasian groups don’t work as well as Mingrelians because there has been so much Steppe migrations-related introgression North of the mountains in the recent millennia. So the Colchi conjecture may just reflect the limitation of the modern populations … it may be the closest of the surviving populations with the least historic change, rather than the closest population of the right time period in the past?

    Read More
    • Replies: @Razib Khan
    it may be the closest of the surviving populations with the least historic change, rather than the closest population of the right time period in the past?

    yes, of course. i assume that we're seeing the closest tip of many extinct branches.
  6. @Asya Pereltsvaig
    So are you saying that Indo-European was the language of European hunter-gatherers, later adopted by farmers who moved in (from Anatolia?)? Did I get that right?

    sort of. i actually believe that the most likely, albeit speculative and without much confidence, that IE and uralic languages are central eurasian in provenance, one of the language groups of the “ancestral north eurasians” (who contribute ~50% of the ancestors of eastern hunger-gatherers who mixed with the arrival near easterners).

    Read More
    • Replies: @Asya Pereltsvaig
    If PIE is a language of hunter-gatherers, before the arrival of the farmers (~10KYA), that would make PIE "too old", based on the linguistic evidence.
  7. @Dmitry Pruss
    Davidski comments on his Kartvelian post that Adyghe and generally contemporary North Caucasian groups don't work as well as Mingrelians because there has been so much Steppe migrations-related introgression North of the mountains in the recent millennia. So the Colchi conjecture may just reflect the limitation of the modern populations ... it may be the closest of the surviving populations with the least historic change, rather than the closest population of the right time period in the past?

    it may be the closest of the surviving populations with the least historic change, rather than the closest population of the right time period in the past?

    yes, of course. i assume that we’re seeing the closest tip of many extinct branches.

    Read More
  8. @Matt_
    It could have happened like that, or you could have had the HG communities adopting one the farmer languages, at the same time as fusing into their communities, which would have been a radical change for both.

    Ultimately it seems like we can only really reconstruct the place (climate, time) in which a language was spoken prior to its expansion.

    It could have happened like that, or you could have had the HG communities adopting one the farmer languages, at the same time as fusing into their communities, which would have been a radical change for both.

    to be clear, i think the ethnographic record that HG -> pastoralist transition can be pretty easy/fluid is the dynamic i’m alluding to. both IE and afro-asiatic groups are fusions of HG+farmer. but i posit that the HG had much more influence than would otherwise be the case because of the fact that they could take to pastoralist economies far more easily than they would have to sedentary agriculture.

    Read More
  9. p.s. one implication is that dravidian languages may be the largest group of “first farmer” langauges left.

    Read More
  10. @Razib Khan
    sort of. i actually believe that the most likely, albeit speculative and without much confidence, that IE and uralic languages are central eurasian in provenance, one of the language groups of the "ancestral north eurasians" (who contribute ~50% of the ancestors of eastern hunger-gatherers who mixed with the arrival near easterners).

    If PIE is a language of hunter-gatherers, before the arrival of the farmers (~10KYA), that would make PIE “too old”, based on the linguistic evidence.

    Read More
    • Replies: @Razib Khan
    what i am imagining is that there were a set of PIE-like languages. one of them became dominant among a group of HG who took up pastoralism and absorbed near eastern agriculturalists. i don't date the arrival of farmers to as early as 10 K in eastern europe. closer to 5 K. someone who knows the archaeology etc. of northeast europe and the trans-volga region can inform.
    , @Davidski
    It doesn't look like early farmers from Anatolia had a big impact on most of Eastern Europe (east of the Dnieper).

    Near Eastern admixture probably spread from the Caucasus across the Caspian steppe mostly via female mediated gene flow (exogamy?) 5,000-4,000 BC, which , as far as I know, is about the right time for the appearance of Proto-Indo-European.

    It's well after this admixture event between Eastern European foragers and a Caucasus population that we see a massive series of migrations from the steppe to Europe and Asia, probably carrying Indo-European daughter languages.

    The identity of the Caucasus population is a mystery, but it was probably very similar to present-day Georgian Mingrelians.
  11. @Asya Pereltsvaig
    If PIE is a language of hunter-gatherers, before the arrival of the farmers (~10KYA), that would make PIE "too old", based on the linguistic evidence.

    what i am imagining is that there were a set of PIE-like languages. one of them became dominant among a group of HG who took up pastoralism and absorbed near eastern agriculturalists. i don’t date the arrival of farmers to as early as 10 K in eastern europe. closer to 5 K. someone who knows the archaeology etc. of northeast europe and the trans-volga region can inform.

    Read More
    • Replies: @Dmitry Pruss
    A possibility that a population native to the Caucasus Mountains took part in Yamnaya culture fusion / events followed by radiation of IE languages is pretty fascinating. Everyone tends to consider the Caucasus Range to be a migration barrier, but it would have been far less of a barrier to a population connected by kinship and kinship-related alliances with the tribes controlling the mountain passes. Today's highlanders are largely pastoralists, often moving herds between summer and winter pastures and across grasslands, which aligns their lifestyle quite well with the hypothesized early Indo-Europeans.

    Interestingly, it might even reconcile the "Steppe" and the "Anatolian" hypotheses to some degree, since migrating pastoralists have been known to maintain ties with their brethren who stayed behind, and to back-migrate when political conflicts in their new lands or political opportunities in their ancestral lands called for it. Most famously, the Kalmyks who grew disillusioned with life on the Caspian Steppe and saw opportunities back in Dzhungar mass-migrated to their former lands in 1771, a century and a half (and many intermarriages) later after their original thrust from Mongolia into the Volga Steppes. In these 150 years the Kalmyks allied themselves with Russia, traded with it and fought in its wars with the Muslims but never ceased following the events back in the old home and communicating with the tribal leaders there, apparently eager to back-migrate when an opportunity calls. Could the Yamnaya leaders have been just as attuned to the events South of the Caucasus, ready to give a burst of back-migration across the terrain which one would consider hard to traverse, but across which they were culture-bound with their distant kin tribes?

    , @Asya Pereltsvaig
    With all due respect, there's no such thing as "a set of PIE-like languages", by definition: it's either PIE or some other language. Just as "a set of English-like languages" is so vague as to be meaningless. Nor could PIE be spoken by hunter-gatherers and then centuries (millennia?) later by agriculturalists, then centuries (millennia?) later by pastoralists... It's complete nonsense.
    , @Asya Pereltsvaig
    Nor does it make sense that PIE speakers were pre-agricultural and pre-horse peoples. Read our book, we talk about this in detail...
  12. @Asya Pereltsvaig
    If PIE is a language of hunter-gatherers, before the arrival of the farmers (~10KYA), that would make PIE "too old", based on the linguistic evidence.

    It doesn’t look like early farmers from Anatolia had a big impact on most of Eastern Europe (east of the Dnieper).

    Near Eastern admixture probably spread from the Caucasus across the Caspian steppe mostly via female mediated gene flow (exogamy?) 5,000-4,000 BC, which , as far as I know, is about the right time for the appearance of Proto-Indo-European.

    It’s well after this admixture event between Eastern European foragers and a Caucasus population that we see a massive series of migrations from the steppe to Europe and Asia, probably carrying Indo-European daughter languages.

    The identity of the Caucasus population is a mystery, but it was probably very similar to present-day Georgian Mingrelians.

    Read More
    • Replies: @Razib Khan
    if it's female mediated, then they took them i would bet.
  13. @Davidski
    It doesn't look like early farmers from Anatolia had a big impact on most of Eastern Europe (east of the Dnieper).

    Near Eastern admixture probably spread from the Caucasus across the Caspian steppe mostly via female mediated gene flow (exogamy?) 5,000-4,000 BC, which , as far as I know, is about the right time for the appearance of Proto-Indo-European.

    It's well after this admixture event between Eastern European foragers and a Caucasus population that we see a massive series of migrations from the steppe to Europe and Asia, probably carrying Indo-European daughter languages.

    The identity of the Caucasus population is a mystery, but it was probably very similar to present-day Georgian Mingrelians.

    if it’s female mediated, then they took them i would bet.

    Read More
    • Replies: @Davidski
    I think they probably did in many cases, or at least at the very beginning. But something like this might have happened too. This is in Bronze Age Germany, but it's probably a population of steppe origin to some degree...

    On the other hand, local population continuity was a prerequisite for the accumulation of wealth, the establishment of enduring social differentiation, and the formation of regional elites.

    ...

    The results indicate both local genetic continuity spanning the cultural transition, and, following the onset of the Early Bronze Age, a major influx of mtDNA types previously not found in this region. Integrating stable isotope data with the genetic data reveals a picture of a patrilocal society with remarkable mobility in women.
     

    Mittnik et al., Ancient DNA reveals patterns of residential continuity and mobility at the onset of the Central European Bronze Age

    http://smbe2015.at/program/program-and-abstracts/

  14. @Razib Khan
    what i am imagining is that there were a set of PIE-like languages. one of them became dominant among a group of HG who took up pastoralism and absorbed near eastern agriculturalists. i don't date the arrival of farmers to as early as 10 K in eastern europe. closer to 5 K. someone who knows the archaeology etc. of northeast europe and the trans-volga region can inform.

    A possibility that a population native to the Caucasus Mountains took part in Yamnaya culture fusion / events followed by radiation of IE languages is pretty fascinating. Everyone tends to consider the Caucasus Range to be a migration barrier, but it would have been far less of a barrier to a population connected by kinship and kinship-related alliances with the tribes controlling the mountain passes. Today’s highlanders are largely pastoralists, often moving herds between summer and winter pastures and across grasslands, which aligns their lifestyle quite well with the hypothesized early Indo-Europeans.

    Interestingly, it might even reconcile the “Steppe” and the “Anatolian” hypotheses to some degree, since migrating pastoralists have been known to maintain ties with their brethren who stayed behind, and to back-migrate when political conflicts in their new lands or political opportunities in their ancestral lands called for it. Most famously, the Kalmyks who grew disillusioned with life on the Caspian Steppe and saw opportunities back in Dzhungar mass-migrated to their former lands in 1771, a century and a half (and many intermarriages) later after their original thrust from Mongolia into the Volga Steppes. In these 150 years the Kalmyks allied themselves with Russia, traded with it and fought in its wars with the Muslims but never ceased following the events back in the old home and communicating with the tribal leaders there, apparently eager to back-migrate when an opportunity calls. Could the Yamnaya leaders have been just as attuned to the events South of the Caucasus, ready to give a burst of back-migration across the terrain which one would consider hard to traverse, but across which they were culture-bound with their distant kin tribes?

    Read More
    • Replies: @Razib Khan
    this is relevant https://logarithmichistory.wordpress.com/2015/09/20/uruk-and-the-empires-before-history/
    , @Asya Pereltsvaig
    "it might even reconcile the “Steppe” and the “Anatolian” hypotheses to some degree" -- not sure how these two hypotheses can be reconciled at all, as there is about 3000 years between them, so regardless of WHERE PIE was spoken, it certainly wasn't spoken in one place and then 3000 years later in another.
  15. @Razib Khan
    if it's female mediated, then they took them i would bet.

    I think they probably did in many cases, or at least at the very beginning. But something like this might have happened too. This is in Bronze Age Germany, but it’s probably a population of steppe origin to some degree…

    On the other hand, local population continuity was a prerequisite for the accumulation of wealth, the establishment of enduring social differentiation, and the formation of regional elites.

    The results indicate both local genetic continuity spanning the cultural transition, and, following the onset of the Early Bronze Age, a major influx of mtDNA types previously not found in this region. Integrating stable isotope data with the genetic data reveals a picture of a patrilocal society with remarkable mobility in women.

    Mittnik et al., Ancient DNA reveals patterns of residential continuity and mobility at the onset of the Central European Bronze Age

    http://smbe2015.at/program/program-and-abstracts/

    Read More
    • Replies: @Dmitry Pruss
    Shift to the far-distance mobility of the women in Bavaria around 2000 BCE (in Alissa Mittnik's abstract) is remarkable but still fades in comparison with your proposed wholesale change of all mtDNA earlier in the Steppe. Could you describe in a few more words what you see in Khvalynsk and Yamnaya, please? Is it the mtDNA spectrum completely matching the agricultural South in Khvalynsk already? What HG groups does it contrast to?

    I'm tying to understand if the mtDNA transition may have been an earlier, separate phenomenon from the gradual increase of the agriculturalist autosomal DNA in the 4th and 3rd millennia BCE (which as I understand went from about 1/4 in Khvalynsk to less one half in early Yamnaya to more than 50% in late-period Yamnaya?)
  16. @Dmitry Pruss
    A possibility that a population native to the Caucasus Mountains took part in Yamnaya culture fusion / events followed by radiation of IE languages is pretty fascinating. Everyone tends to consider the Caucasus Range to be a migration barrier, but it would have been far less of a barrier to a population connected by kinship and kinship-related alliances with the tribes controlling the mountain passes. Today's highlanders are largely pastoralists, often moving herds between summer and winter pastures and across grasslands, which aligns their lifestyle quite well with the hypothesized early Indo-Europeans.

    Interestingly, it might even reconcile the "Steppe" and the "Anatolian" hypotheses to some degree, since migrating pastoralists have been known to maintain ties with their brethren who stayed behind, and to back-migrate when political conflicts in their new lands or political opportunities in their ancestral lands called for it. Most famously, the Kalmyks who grew disillusioned with life on the Caspian Steppe and saw opportunities back in Dzhungar mass-migrated to their former lands in 1771, a century and a half (and many intermarriages) later after their original thrust from Mongolia into the Volga Steppes. In these 150 years the Kalmyks allied themselves with Russia, traded with it and fought in its wars with the Muslims but never ceased following the events back in the old home and communicating with the tribal leaders there, apparently eager to back-migrate when an opportunity calls. Could the Yamnaya leaders have been just as attuned to the events South of the Caucasus, ready to give a burst of back-migration across the terrain which one would consider hard to traverse, but across which they were culture-bound with their distant kin tribes?

    Read More
    • Replies: @Dmitry Pruss
    Razib, you mean the unidentified massive fighting force which sacked Tell Hamoukar in Northern Mesopotamia about 3500 BCE? Usually the attackers are thought to be Uruk invaders from the South...
    http://archaeology.about.com/od/mesopotamiaarchaeology/a/hamoukar.htm
  17. @Razib Khan
    what i am imagining is that there were a set of PIE-like languages. one of them became dominant among a group of HG who took up pastoralism and absorbed near eastern agriculturalists. i don't date the arrival of farmers to as early as 10 K in eastern europe. closer to 5 K. someone who knows the archaeology etc. of northeast europe and the trans-volga region can inform.

    With all due respect, there’s no such thing as “a set of PIE-like languages”, by definition: it’s either PIE or some other language. Just as “a set of English-like languages” is so vague as to be meaningless. Nor could PIE be spoken by hunter-gatherers and then centuries (millennia?) later by agriculturalists, then centuries (millennia?) later by pastoralists… It’s complete nonsense.

    Read More
    • Replies: @ryanwc
    Hmm. "A set of Algonquian languages" or "a set of Iroquoian languages" certainly makes sense. And in some cases they recoalesced in the aftermath of colonial contact, as elements of various tribes were thrown together. If, say, modern Ojibwa could grow out of an old Ojibwan core along with multifarious effects from various sister languages, I'm not sure why one couldn't refer to the Algonquian languages then extant as "a set of Proto-Ojibwan languages" in, say, 1492, . "A set of PIE-like languages" seems pretty plausible - related bands of hunter gatherers, speaking similar languages with varying degrees of mutual comprehension, coalescing into whatever tongue was spoken by the Yamnaya.

    The early divergence of the Anatolian branch of IE remains a bit of a quandary, no? Yamnaya develops expansive tendencies, and the first thing they do is head south to claim a patchwork of territories in Anatolia, in a lingering sea of non-IE. And only much later send their wagon trains out to exploit the softer targets scattered west and east of them. They bring with them to Anatolia their words for draft technology and cattle, technologies which have a long history in Anatolia prior to the new IE speakers. But they fail to establish their words for horse, wheel and cart, which are the very new technologies that give them their superiority?

    Linguistically, it still seems more convincing to highlight the IE roots for cattle and for draft tools shared across Anatolian and the rest of IE, unlike the horse/wheel terminology shared only post-Anatolian. This seems to locate the earliest PIE in one of the dairy cultures of Anatolia. Kartvelian genetics might be our only proxy today for the genetics of a dairying people who crossed the Caucasus or circled the Black Sea, carrying their language into the Pontic steppe, where their skill with cattle gave them the advantage in domesticating a more unruly species, the horse. Something which steppe peoples hadn't accomplished in the previous millennia.

    I recognize how one of your points, Razib, cuts against what I'm writing. If all the technologies came with the southern peoples, what advantage accrued to the HG that allowed them to contribute so much of the genetics? Is it not the affinity of HG for a pastoral lifestyle that allows them to steal cattle, tame the horse, and take the upper hand in the relationship with whatever other peoples contributed to the genetic pool where PIE was bubbling up? You may be right.

    I'm just trying to play devil's advocate, while highlighting one weakness with the linguistic argument that has been advanced, often naively, based on an assumption that the only possible counter-narrative would be an expansion with the first farmers. There are still possibilities for later out-of-Anatolia expansion. .

    Is there Hittite DNA coming our way anytime soon?
  18. @Razib Khan
    what i am imagining is that there were a set of PIE-like languages. one of them became dominant among a group of HG who took up pastoralism and absorbed near eastern agriculturalists. i don't date the arrival of farmers to as early as 10 K in eastern europe. closer to 5 K. someone who knows the archaeology etc. of northeast europe and the trans-volga region can inform.

    Nor does it make sense that PIE speakers were pre-agricultural and pre-horse peoples. Read our book, we talk about this in detail…

    Read More
    • Replies: @Razib Khan
    everyone was at some point.
    , @notanon
    Isn't PIE considered to be a point of transition - if so wouldn't it have to be both pre and post agricultural/horse?
  19. @Dmitry Pruss
    A possibility that a population native to the Caucasus Mountains took part in Yamnaya culture fusion / events followed by radiation of IE languages is pretty fascinating. Everyone tends to consider the Caucasus Range to be a migration barrier, but it would have been far less of a barrier to a population connected by kinship and kinship-related alliances with the tribes controlling the mountain passes. Today's highlanders are largely pastoralists, often moving herds between summer and winter pastures and across grasslands, which aligns their lifestyle quite well with the hypothesized early Indo-Europeans.

    Interestingly, it might even reconcile the "Steppe" and the "Anatolian" hypotheses to some degree, since migrating pastoralists have been known to maintain ties with their brethren who stayed behind, and to back-migrate when political conflicts in their new lands or political opportunities in their ancestral lands called for it. Most famously, the Kalmyks who grew disillusioned with life on the Caspian Steppe and saw opportunities back in Dzhungar mass-migrated to their former lands in 1771, a century and a half (and many intermarriages) later after their original thrust from Mongolia into the Volga Steppes. In these 150 years the Kalmyks allied themselves with Russia, traded with it and fought in its wars with the Muslims but never ceased following the events back in the old home and communicating with the tribal leaders there, apparently eager to back-migrate when an opportunity calls. Could the Yamnaya leaders have been just as attuned to the events South of the Caucasus, ready to give a burst of back-migration across the terrain which one would consider hard to traverse, but across which they were culture-bound with their distant kin tribes?

    “it might even reconcile the “Steppe” and the “Anatolian” hypotheses to some degree” — not sure how these two hypotheses can be reconciled at all, as there is about 3000 years between them, so regardless of WHERE PIE was spoken, it certainly wasn’t spoken in one place and then 3000 years later in another.

    Read More
    • Replies: @Dmitry Pruss
    Asya, of course having PIE radiate from Anatolia soon after Neolithic revolution can't be reconciled with anything "back-migration from the Steppe". Because it's too early. "Some degree" of vindicating Anatlolian theory / modified Gamkrelidze-Ivanov would have been like this: pre-PIE possibly existing South of the Caucasus but Anatolian languages emerging in a migration South, across the mountains. "Gamkrelidze-like (also linking Maikop and Azerbaijani Chalcolite sites or their immediate predecessor cultures) but .... flowing in the opposite direction".
  20. An idea I had on early Indo-European society was a social model very concerned with passing on patrilines (patriarchy, father-son relationships, etc), so maybe did not put as much in the way of resources into behind marriage, etc. for daughters.

    Daughters from outside the culture, from other more matriline and female fertility concerned cultures, who might have been more able to inherit resources from their family (in terms of assistance and material resources), might then have made more attractive partners for men within the IE culture. Equally males within the culture could draw on a strong base of resources and support less available to males outside the culture, helping their prospects of reproducing and spreading their y dna disproportionately.

    Thus turnover of female lineages from outside the culture being drawn into the culture (although perhaps with the Basques and their progenitor culture it may have gone the other way, with males from outside the culture introgressing into the culture). After a certain amount of time, this complex would have to stabilise as it reached its limits.

    May all very well be a fantasy though, and perhaps it was more violent.

    Read More
    • Replies: @anowow
    Yeah, but Kartvelian cultures are very patriarchal.

    It is possible for a culture to be patriarchal and militarily weak and thus exploitable, including having their females taken. Think Mongols and Persian Muslims.


    Mingrelians and Imeretians are interesting groups; if you go there you see people with Anatolian or Armenian features (although not as many as in Georgia east of the Surami range) alongside people with Central European, Italian or Balkan features. I saw a lot of dead ringers for Pres. Zachary Taylor, James Coburn, Jake Gyllenhall and David Arquette. Unz's Phil Girardi has a lot of doppelgangers in the markets of Zugdidi or Kutaisi. And lots (still a minority, but a sizable one) of Auburn haired people with dark eyes. Iranians and Azeris stand out. Iranians particularly, as do RUssians and Poles.

    That and a local breed of cow with tiger stripes.

    , @notanon
    I tend to think it was mostly captives but that may just be my personality talking.

    Although a raider population over time gradually turning into the people they raid (autosomally) seems like a straightforward logical consequence.

    Another more peaceful possibility I wondered about might be a variation of Clarkian effect where the steppe chiefs had arranged dynastic marriages with caucasus chiefs and their offspring had a reproductive advantage over time hence a top down spread of their particular dna mix.
  21. @Asya Pereltsvaig
    Nor does it make sense that PIE speakers were pre-agricultural and pre-horse peoples. Read our book, we talk about this in detail...

    everyone was at some point.

    Read More
    • Replies: @Asya Pereltsvaig
    Not at the time they were speakers of PIE. What you say is equivalent to saying that, for example, French speakers are/were nomadic or hunter-gatherers. Their ancestors millennia before there was such a thing as "the French language" might have been, but it's entirely irrelevant and therefore inappropriate to call such earlier groups "the French".

    http://www.languagesoftheworld.info/historical-linguistics/homeland-problem.html
  22. @Razib Khan
    everyone was at some point.

    Not at the time they were speakers of PIE. What you say is equivalent to saying that, for example, French speakers are/were nomadic or hunter-gatherers. Their ancestors millennia before there was such a thing as “the French language” might have been, but it’s entirely irrelevant and therefore inappropriate to call such earlier groups “the French”.

    http://www.languagesoftheworld.info/historical-linguistics/homeland-problem.html

    Read More
    • Replies: @Razib Khan
    i'm saying that EHG may have spoken the ancestor of PIE.
  23. @Razib Khan
    this is relevant https://logarithmichistory.wordpress.com/2015/09/20/uruk-and-the-empires-before-history/

    Razib, you mean the unidentified massive fighting force which sacked Tell Hamoukar in Northern Mesopotamia about 3500 BCE? Usually the attackers are thought to be Uruk invaders from the South…

    http://archaeology.about.com/od/mesopotamiaarchaeology/a/hamoukar.htm

    Read More
  24. @Asya Pereltsvaig
    Not at the time they were speakers of PIE. What you say is equivalent to saying that, for example, French speakers are/were nomadic or hunter-gatherers. Their ancestors millennia before there was such a thing as "the French language" might have been, but it's entirely irrelevant and therefore inappropriate to call such earlier groups "the French".

    http://www.languagesoftheworld.info/historical-linguistics/homeland-problem.html

    i’m saying that EHG may have spoken the ancestor of PIE.

    Read More
  25. @Asya Pereltsvaig
    "it might even reconcile the “Steppe” and the “Anatolian” hypotheses to some degree" -- not sure how these two hypotheses can be reconciled at all, as there is about 3000 years between them, so regardless of WHERE PIE was spoken, it certainly wasn't spoken in one place and then 3000 years later in another.

    Asya, of course having PIE radiate from Anatolia soon after Neolithic revolution can’t be reconciled with anything “back-migration from the Steppe”. Because it’s too early. “Some degree” of vindicating Anatlolian theory / modified Gamkrelidze-Ivanov would have been like this: pre-PIE possibly existing South of the Caucasus but Anatolian languages emerging in a migration South, across the mountains. “Gamkrelidze-like (also linking Maikop and Azerbaijani Chalcolite sites or their immediate predecessor cultures) but …. flowing in the opposite direction”.

    Read More
    • Replies: @Asya Pereltsvaig
    Pre-PIE was in Africa, like pre-Any-Other-Language. It's not the problem. And using terms like PIE for people who came generations earlier or later is simply meaningless and misleading.
  26. @Dmitry Pruss
    Asya, of course having PIE radiate from Anatolia soon after Neolithic revolution can't be reconciled with anything "back-migration from the Steppe". Because it's too early. "Some degree" of vindicating Anatlolian theory / modified Gamkrelidze-Ivanov would have been like this: pre-PIE possibly existing South of the Caucasus but Anatolian languages emerging in a migration South, across the mountains. "Gamkrelidze-like (also linking Maikop and Azerbaijani Chalcolite sites or their immediate predecessor cultures) but .... flowing in the opposite direction".

    Pre-PIE was in Africa, like pre-Any-Other-Language. It’s not the problem. And using terms like PIE for people who came generations earlier or later is simply meaningless and misleading.

    Read More
    • Replies: @Dmitry Pruss
    Asya, you have a fascinating style of discussion, where very important factual observations on grammar, vocabulary, phonetics etc. are intermixed with the bits of general wisdom like "pre-Any-Other-Language was in Africa". I just assume that you were irritated by my or Razib's cautious approach to assigning names to the various proto-states of early splits of the IE languages. But isn't it a widespread issue, that there is no universal agreement yet, how to call these descendants or predecessors of PIE? In your own articles, you've variously proposed PNIE or perhaps PSIE for the Yamnaya culture (either a predecessor of IE languages other than Anatolian, or perhaps a predecessor of the extant IE languages, that is other than Anatolian and Tokharian), or Proto-Indo-Hittite which would have been the root of Anatolian and PNIE if we decided to apply the label PIE to the PNIE? These are all very interesting suggestions, but they also make it clear that the terminology of these proto-languages is in flux. I don't think I deserve the blame for trying to navigate the maze of these proto-language names with caution :) I certainly meant no disrespect and I remain your loyal fan.

    I understand your well-articulated position that the Anatolian languages split off before Yamnaya, and originated outside of Anatolia, moving in likely from the North-East (the Caucasus) or North-West (the Balkans). I also know that you wrote an extensive critique of Gamkrelidze and Ivanov's South Caucasus Urheimat theory, which they built upon the observations (perhaps faulty) of the Proto-Kartvelian influences early in IE history. The recent discovery that the extant Kartvelians are similar to the agriculturalist ancestors of the Yamnaya thrusts the Gamkrelidze-Ivanov theory to the forefront ... the geography and the genetics are kind of right, and they claim that the linguistics is right too. I would really appreciate hearing your detailed critique of what's wrong with Gamkrelidze-Ivanov.
  27. @Matt_
    An idea I had on early Indo-European society was a social model very concerned with passing on patrilines (patriarchy, father-son relationships, etc), so maybe did not put as much in the way of resources into behind marriage, etc. for daughters.

    Daughters from outside the culture, from other more matriline and female fertility concerned cultures, who might have been more able to inherit resources from their family (in terms of assistance and material resources), might then have made more attractive partners for men within the IE culture. Equally males within the culture could draw on a strong base of resources and support less available to males outside the culture, helping their prospects of reproducing and spreading their y dna disproportionately.

    Thus turnover of female lineages from outside the culture being drawn into the culture (although perhaps with the Basques and their progenitor culture it may have gone the other way, with males from outside the culture introgressing into the culture). After a certain amount of time, this complex would have to stabilise as it reached its limits.

    May all very well be a fantasy though, and perhaps it was more violent.

    Yeah, but Kartvelian cultures are very patriarchal.

    It is possible for a culture to be patriarchal and militarily weak and thus exploitable, including having their females taken. Think Mongols and Persian Muslims.

    Mingrelians and Imeretians are interesting groups; if you go there you see people with Anatolian or Armenian features (although not as many as in Georgia east of the Surami range) alongside people with Central European, Italian or Balkan features. I saw a lot of dead ringers for Pres. Zachary Taylor, James Coburn, Jake Gyllenhall and David Arquette. Unz’s Phil Girardi has a lot of doppelgangers in the markets of Zugdidi or Kutaisi. And lots (still a minority, but a sizable one) of Auburn haired people with dark eyes. Iranians and Azeris stand out. Iranians particularly, as do RUssians and Poles.

    That and a local breed of cow with tiger stripes.

    Read More
    • Replies: @Matt_
    Yes, idea was more about more vs less than not at all, plus present day Kartvelians may be different from whichever related group may have been involved as isolated pioneers in an initial founding of a Yamnaya population.

    I can't rule out any military, men take women sort of thing. I am noting that the expansion of Indo-European peoples into Europe (not so much the Caucasus really) combined with the dwindling of their mtdna and to a large extent autosome (55-40% European Indo-European language speakers, in Northwest and Northeast Europe, who are the maximum) seem like it could be explained via different offspring / reproductive / family preferences that compounded over time between IE and non-IE groups. Take some of those groups with strong son preferences today (sex selective female abortion etc.) and strong preferences for continuing the male family name - if that's culturally communicated over 1000 years, while they live alongside and marry people who aren't as much like that, even with primitive technology, eventually all the male part of their ancestry will be from one group, even though the whole group is mixed.

    Unless there are strong signs in the archaeology, we're not gonna know.
  28. I’m thinking there’s something special about Georgian genetic structure. It’s Near Eastern-related, but probably not recently derived from the Near East.

    Pinhasi has some samples from Satsurblia and Eneolithic sites in western Georgia. I’ve got a feeling that when we see these analyzed they’ll show a close relationship to the Georgian Mingrelians that I ran in the TreeMix analysis.

    This is a paper by Pinhasi et al. on some Satsurblia sites.

    http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0111271

    Read More
  29. A plausible model is that PIE and Uralic were similar languages spoken around the central steppes; then the PIE move west, conquer some Caucasus language speaking farmers, and the mixed language becomes IE.

    IE languages have both commonalities with Uralic and Caucasian languages. And the structure of nominal declension strongly hints at it being a mongrel language; the very idea of different declensions is very bad design. Most languages have a single set of functional morphemes which are quite transparent, IE languages though have messed the whole thing into an irregular system of sound roots which makes no sense; unless it was the result of the mixing of two different language families.

    Read More
    • Replies: @Megalophias
    I would say quite the opposite; absurd irregularities arise from the natural development of language, when mixing with another language if anything it is likely to be simplified. Look at English, heavily influenced by Norse and French; the result is a whole bunch of extraneous vocabulary, but with *less* inflection.
    , @Asya Pereltsvaig
    Correction 1: PIE and Proto-Uralic were not at all linguistically similar, as far as we can reconstruct them.

    Correction 2: Every language (or almost every language) is a "mongrel language" to a degree, but we can diagnose such "mongrel"-ness by a particular linguistic feature, such as fusional morphology, which is what you are talking about here.

    Correction 3: Fusional morphology (including "different declensions") isn't "very bad design" any more than agglutinative morphology or isolating morphology. This issue is discussed in detail here:
    http://www.languagesoftheworld.info/historical-linguistics/more-on-word-order-morphological-types-and-historical-change.html

    Correction 4: "Most languages have a single set of functional morphemes which are quite transparent" -- that's factually false. A great many fusional and polysynthetic languages do not.

    Correction 5: "IE languages though have messed the whole thing into an irregular system of sound roots which makes no sense" -- maybe it makes no sense to you, but it makes perfect sense in terms of the organization of morphological systems (see post linked above). Moreover, non-Pama-Nyungan languages in Australia and many Finno-Ugric languages likewise developed fusional morphology out of the agglutinative morphology (which according to you makes more sense)
  30. @spandrell
    A plausible model is that PIE and Uralic were similar languages spoken around the central steppes; then the PIE move west, conquer some Caucasus language speaking farmers, and the mixed language becomes IE.

    IE languages have both commonalities with Uralic and Caucasian languages. And the structure of nominal declension strongly hints at it being a mongrel language; the very idea of different declensions is very bad design. Most languages have a single set of functional morphemes which are quite transparent, IE languages though have messed the whole thing into an irregular system of sound roots which makes no sense; unless it was the result of the mixing of two different language families.

    I would say quite the opposite; absurd irregularities arise from the natural development of language, when mixing with another language if anything it is likely to be simplified. Look at English, heavily influenced by Norse and French; the result is a whole bunch of extraneous vocabulary, but with *less* inflection.

    Read More
    • Replies: @spandrell
    English is actually very irregular. English has hundreds of irregular verbs, even irregular plurals!

    English grammar might be simple; having shed most of its morphology, but it's more irregular than Saxon or French. And inflection needn't be hard; Finnish has way more inflection than any IE language, but it's very regular and not hard at all.
  31. @Asya Pereltsvaig
    With all due respect, there's no such thing as "a set of PIE-like languages", by definition: it's either PIE or some other language. Just as "a set of English-like languages" is so vague as to be meaningless. Nor could PIE be spoken by hunter-gatherers and then centuries (millennia?) later by agriculturalists, then centuries (millennia?) later by pastoralists... It's complete nonsense.

    Hmm. “A set of Algonquian languages” or “a set of Iroquoian languages” certainly makes sense. And in some cases they recoalesced in the aftermath of colonial contact, as elements of various tribes were thrown together. If, say, modern Ojibwa could grow out of an old Ojibwan core along with multifarious effects from various sister languages, I’m not sure why one couldn’t refer to the Algonquian languages then extant as “a set of Proto-Ojibwan languages” in, say, 1492, . “A set of PIE-like languages” seems pretty plausible – related bands of hunter gatherers, speaking similar languages with varying degrees of mutual comprehension, coalescing into whatever tongue was spoken by the Yamnaya.

    The early divergence of the Anatolian branch of IE remains a bit of a quandary, no? Yamnaya develops expansive tendencies, and the first thing they do is head south to claim a patchwork of territories in Anatolia, in a lingering sea of non-IE. And only much later send their wagon trains out to exploit the softer targets scattered west and east of them. They bring with them to Anatolia their words for draft technology and cattle, technologies which have a long history in Anatolia prior to the new IE speakers. But they fail to establish their words for horse, wheel and cart, which are the very new technologies that give them their superiority?

    Linguistically, it still seems more convincing to highlight the IE roots for cattle and for draft tools shared across Anatolian and the rest of IE, unlike the horse/wheel terminology shared only post-Anatolian. This seems to locate the earliest PIE in one of the dairy cultures of Anatolia. Kartvelian genetics might be our only proxy today for the genetics of a dairying people who crossed the Caucasus or circled the Black Sea, carrying their language into the Pontic steppe, where their skill with cattle gave them the advantage in domesticating a more unruly species, the horse. Something which steppe peoples hadn’t accomplished in the previous millennia.

    I recognize how one of your points, Razib, cuts against what I’m writing. If all the technologies came with the southern peoples, what advantage accrued to the HG that allowed them to contribute so much of the genetics? Is it not the affinity of HG for a pastoral lifestyle that allows them to steal cattle, tame the horse, and take the upper hand in the relationship with whatever other peoples contributed to the genetic pool where PIE was bubbling up? You may be right.

    I’m just trying to play devil’s advocate, while highlighting one weakness with the linguistic argument that has been advanced, often naively, based on an assumption that the only possible counter-narrative would be an expansion with the first farmers. There are still possibilities for later out-of-Anatolia expansion. .

    Is there Hittite DNA coming our way anytime soon?

    Read More
    • Replies: @ryanwc
    I would add that the presumed history of the Anatolian branch under a Pontic origin hypothesis is very similar to the history of its cousins - it was grafted onto various substrate languages of first farmer cultures. Mere passage of time shouldn't have produced so many more structural changes in Anatolian. After all, from PIE to Hittite is only about 1500 years. The Pontic origin theory suggests at least as much time between, say, Latin and Sanskrit, which must have been overlaid onto substrates that were far more different.

    It seems more likely that non-Anatolian IE diverged from Anatolian when the original stock reached the melting pot of Yamnaya.
    , @Asya Pereltsvaig
    Linguistically, talking about "a set of PIE-like languages" doesn't make sense for many reasons. For one thing, "PIE-like" can mean a lot of different things. But even if we restrict the term to "languages/dialects related to PIE through common descent", it's still not very illuminating. Let's say we are discussing a certain Y-DNA haplogroup, i.e. men who have a certain mutation in common (or certain mutations, plural), and we are interested in finding out where and when their MRCA lived. How much sense would it make to talk about "a set of MRCA-like men" who lived maybe at this particular time or maybe 30% earlier, maybe here or maybe there? Not much. We can hypothesize (with 100% certainty!) that the ancestor of Indo-European languages (PIE) was spoken on planet Earth some time in the last 200KY. Great theory and definitely right, but not very scientific (falsifiable), is it?

    Regarding your idea that there were related (and very similar) dialects, that's in all likelihood true, but they can be agglomerated under one language (just as English really means a bunch of different varieties). But still that entire bunch of PIE dialects could not have been spoken over a too-large area for any significant length of time. Because if it was, it would have developed into distinct languages. And we are interested in find out the time and place where that diversification happened. See more in the post that I linked to in a comment above.

    Re: the PIE/PNIE/PSIE terminology for horses, wheels, etc., I will have to refer you to our book, where almost an entire chapter deal with this issue in detail (including in linguistic detail), so I couldn't possibly give it justice in a post here:
    http://www.amazon.com/Indo-European-Controversy-Fallacies-Historical-Linguistics/dp/1107054532/ref=sr_1_1?s=books&ie=UTF8&qid=1419572204&sr=1-1&keywords=indo-european+controversy
    , @Dmitry Pruss
    Asya's blog discusses a potential wheel-word cognate in Hittite (hurki–, per German Dziebel), but also insists that even the absence of such a word in Anatolian wouldn't have been a problem because the Anatolian languages split off too early.

    There is also a fascinating discussion of South-of-the Caucasus origins of PIE in "The Indo-European Controversy". Asya Pereltsvaig & Martin W. Lewis start from asserting that Gamkrelidze and Ivanov's 1995 PIE reconstructions for the "Southern" words such as panther, lion, elephant may have been wrong. The words for "wine", they write, although clearly shared in PIE (including Anatolian but not Tokharian), Proto-Kartvelian, and Proto-Semitic, may have been borrowed by PIE rather than from PIE.

    The most prominent critique in the book is reserved for the Glottalic theory of evolution of phonemes in the vicinity of ejective consonants (which, according to Gamkrelidze-Ivanov, have been borrowed from Proto-Kartvelian into PIE). The ejectives have been widely lost in the extant IE languages, although they've been re-acquired by languages which came in contact with the Kartvelians later (such as Armenian and Ossetian). Pereltsvaig & Lewis argue that numerous North Caucasus languages use ejectives as well, and so PIE may have acquired its ejectives North of the mountains too - in fact even Kartvelian languages could have acquired them from the North, opening a possibility that Proto-Kartvelian didn't even have ejectives yet.

    North-West Caucasus, where Maikop Culture flourished immediately before Yamnaya, is of particular interest to Pereltsvaig & Lewis . They note that Proto-NW Caucasian shared a number of unusual features with PIE: consonant-rich sound systems including labio-velarization of consonants, tonal accent, gender, number suppletion of pronouns etc. But no borrowings between Proto-NWC and PIE have been documented. Of course Davidski tried using contemporary NW Caucasus DNA in his Yamnaya quest, but there is too much latter-age Steppe introgression there to make anything easy.

    Other borrowings, pre-dating Anatolian-PNIE split, seem to point even further South, to Proto-Semitic. For example, PIE roots for "seven" and "star" are from the Proto-Semitic words for "seven" and "Venus", resp. (a review e.g. here http://www.eupedia.com/forum/threads/27182-Proto-Semitic-and-Proto-Indo-European ). Semitic languages weren't spoken North of Upper Mesopotamia, and even there, not attested until 2800 BC, although Bayesian modeling suggests that Proto-Semitic is about a millennium older than that (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2839953/ )
    , @Davidski
    You're ignoring the ancient DNA evidence.

    The steppe wasn't settled by anyone native to Anatolia. Neolithic samples from western and central Anatolia have been sequenced and they don't look ancestral to Yamnaya, not even in part.

    On the other hand, Anatolia saw major population turnovers after the Neolithic, including the arrival at some point of steppe-related ancestry.

    The important thing to establish now is when the steppe-related ancestry first arrived in Anatolia.
  32. @ryanwc
    Hmm. "A set of Algonquian languages" or "a set of Iroquoian languages" certainly makes sense. And in some cases they recoalesced in the aftermath of colonial contact, as elements of various tribes were thrown together. If, say, modern Ojibwa could grow out of an old Ojibwan core along with multifarious effects from various sister languages, I'm not sure why one couldn't refer to the Algonquian languages then extant as "a set of Proto-Ojibwan languages" in, say, 1492, . "A set of PIE-like languages" seems pretty plausible - related bands of hunter gatherers, speaking similar languages with varying degrees of mutual comprehension, coalescing into whatever tongue was spoken by the Yamnaya.

    The early divergence of the Anatolian branch of IE remains a bit of a quandary, no? Yamnaya develops expansive tendencies, and the first thing they do is head south to claim a patchwork of territories in Anatolia, in a lingering sea of non-IE. And only much later send their wagon trains out to exploit the softer targets scattered west and east of them. They bring with them to Anatolia their words for draft technology and cattle, technologies which have a long history in Anatolia prior to the new IE speakers. But they fail to establish their words for horse, wheel and cart, which are the very new technologies that give them their superiority?

    Linguistically, it still seems more convincing to highlight the IE roots for cattle and for draft tools shared across Anatolian and the rest of IE, unlike the horse/wheel terminology shared only post-Anatolian. This seems to locate the earliest PIE in one of the dairy cultures of Anatolia. Kartvelian genetics might be our only proxy today for the genetics of a dairying people who crossed the Caucasus or circled the Black Sea, carrying their language into the Pontic steppe, where their skill with cattle gave them the advantage in domesticating a more unruly species, the horse. Something which steppe peoples hadn't accomplished in the previous millennia.

    I recognize how one of your points, Razib, cuts against what I'm writing. If all the technologies came with the southern peoples, what advantage accrued to the HG that allowed them to contribute so much of the genetics? Is it not the affinity of HG for a pastoral lifestyle that allows them to steal cattle, tame the horse, and take the upper hand in the relationship with whatever other peoples contributed to the genetic pool where PIE was bubbling up? You may be right.

    I'm just trying to play devil's advocate, while highlighting one weakness with the linguistic argument that has been advanced, often naively, based on an assumption that the only possible counter-narrative would be an expansion with the first farmers. There are still possibilities for later out-of-Anatolia expansion. .

    Is there Hittite DNA coming our way anytime soon?

    I would add that the presumed history of the Anatolian branch under a Pontic origin hypothesis is very similar to the history of its cousins – it was grafted onto various substrate languages of first farmer cultures. Mere passage of time shouldn’t have produced so many more structural changes in Anatolian. After all, from PIE to Hittite is only about 1500 years. The Pontic origin theory suggests at least as much time between, say, Latin and Sanskrit, which must have been overlaid onto substrates that were far more different.

    It seems more likely that non-Anatolian IE diverged from Anatolian when the original stock reached the melting pot of Yamnaya.

    Read More
  33. @anowow
    Yeah, but Kartvelian cultures are very patriarchal.

    It is possible for a culture to be patriarchal and militarily weak and thus exploitable, including having their females taken. Think Mongols and Persian Muslims.


    Mingrelians and Imeretians are interesting groups; if you go there you see people with Anatolian or Armenian features (although not as many as in Georgia east of the Surami range) alongside people with Central European, Italian or Balkan features. I saw a lot of dead ringers for Pres. Zachary Taylor, James Coburn, Jake Gyllenhall and David Arquette. Unz's Phil Girardi has a lot of doppelgangers in the markets of Zugdidi or Kutaisi. And lots (still a minority, but a sizable one) of Auburn haired people with dark eyes. Iranians and Azeris stand out. Iranians particularly, as do RUssians and Poles.

    That and a local breed of cow with tiger stripes.

    Yes, idea was more about more vs less than not at all, plus present day Kartvelians may be different from whichever related group may have been involved as isolated pioneers in an initial founding of a Yamnaya population.

    I can’t rule out any military, men take women sort of thing. I am noting that the expansion of Indo-European peoples into Europe (not so much the Caucasus really) combined with the dwindling of their mtdna and to a large extent autosome (55-40% European Indo-European language speakers, in Northwest and Northeast Europe, who are the maximum) seem like it could be explained via different offspring / reproductive / family preferences that compounded over time between IE and non-IE groups. Take some of those groups with strong son preferences today (sex selective female abortion etc.) and strong preferences for continuing the male family name – if that’s culturally communicated over 1000 years, while they live alongside and marry people who aren’t as much like that, even with primitive technology, eventually all the male part of their ancestry will be from one group, even though the whole group is mixed.

    Unless there are strong signs in the archaeology, we’re not gonna know.

    Read More
  34. @Megalophias
    I would say quite the opposite; absurd irregularities arise from the natural development of language, when mixing with another language if anything it is likely to be simplified. Look at English, heavily influenced by Norse and French; the result is a whole bunch of extraneous vocabulary, but with *less* inflection.

    English is actually very irregular. English has hundreds of irregular verbs, even irregular plurals!

    English grammar might be simple; having shed most of its morphology, but it’s more irregular than Saxon or French. And inflection needn’t be hard; Finnish has way more inflection than any IE language, but it’s very regular and not hard at all.

    Read More
    • Replies: @Asya Pereltsvaig
    Megalophias is right here, English has become much more regular over time. Old English had swealg and healp, now we have swallowed and helped. Old English had lambru and bēc, now we have lambs and books. Modern English has relatively few irregular forms, compared to its earlier forms, including Saxon. And although Finnish inflection is more regular, there is a lot more different morphemes. What is "hard" is a subjective notion, but when talking about morphological complexity, we talk about the number of distinctions made (fewer in Modern English than in Old English), the number of regular patterns, the number of exceptions etc.
    , @Megalophias
    English has lots of irregular verbs, but they are all inherited from Germanic. Verbs borrowed from French or Latin are nice and regular. (Which is why if you are an English speaker studying French, the irregular verbs are a pain in the ass, whereas in German half the time you don't even notice they are irregular - singen sang gesungen, bringen bracht gebracht).

    Most irregular plurals are inherited too (e.g. mouse-mice, ox-oxen), except when both the singular and the plural are borrowed together (usually from Greek or Latin). But those borrowed plurals are not productive except in wordplay or mistaken attribution (e.g. octopi).

  35. @Asya Pereltsvaig
    Pre-PIE was in Africa, like pre-Any-Other-Language. It's not the problem. And using terms like PIE for people who came generations earlier or later is simply meaningless and misleading.

    Asya, you have a fascinating style of discussion, where very important factual observations on grammar, vocabulary, phonetics etc. are intermixed with the bits of general wisdom like “pre-Any-Other-Language was in Africa”. I just assume that you were irritated by my or Razib’s cautious approach to assigning names to the various proto-states of early splits of the IE languages. But isn’t it a widespread issue, that there is no universal agreement yet, how to call these descendants or predecessors of PIE? In your own articles, you’ve variously proposed PNIE or perhaps PSIE for the Yamnaya culture (either a predecessor of IE languages other than Anatolian, or perhaps a predecessor of the extant IE languages, that is other than Anatolian and Tokharian), or Proto-Indo-Hittite which would have been the root of Anatolian and PNIE if we decided to apply the label PIE to the PNIE? These are all very interesting suggestions, but they also make it clear that the terminology of these proto-languages is in flux. I don’t think I deserve the blame for trying to navigate the maze of these proto-language names with caution :) I certainly meant no disrespect and I remain your loyal fan.

    I understand your well-articulated position that the Anatolian languages split off before Yamnaya, and originated outside of Anatolia, moving in likely from the North-East (the Caucasus) or North-West (the Balkans). I also know that you wrote an extensive critique of Gamkrelidze and Ivanov’s South Caucasus Urheimat theory, which they built upon the observations (perhaps faulty) of the Proto-Kartvelian influences early in IE history. The recent discovery that the extant Kartvelians are similar to the agriculturalist ancestors of the Yamnaya thrusts the Gamkrelidze-Ivanov theory to the forefront … the geography and the genetics are kind of right, and they claim that the linguistics is right too. I would really appreciate hearing your detailed critique of what’s wrong with Gamkrelidze-Ivanov.

    Read More
    • Replies: @Asya Pereltsvaig
    I am sorry if I sounded annoyed, but I think the precise use of terminology and also understanding what the question is before comparing answers is crucial. Regarding the terminology, I don't think it's "in flux". We certainly didn't invent "PNIE" (for the ancestor of Indo-European languages minus Anatolian), though I did invent "PSIE" (as far as I know there was no label for the ancestor of all Indo-European languages minus Anatolian and Tocharian). But I wouldn't call these "PIE-like languages" and that's certainly not a reading I thought of when I read your comment. Also, I am quite opposed to the use of "Indo-Hittite" or "Proto-Indo-Hittite" as it is muddling the issue and using it means that other terms like "PIE" are understood to mean something other than what they are usually understood to mean.

    As for matching these languages to archaeological cultures, I remain for the most part agnostic about this, although we do discuss it in the book a little (chapter 9). The main argument is that based on our archaeological knowledge, the Steppe theory provides a better story for migrations than the Anatolian theory.

    As for Gamkrelidze-Ivanov theory, I have a detailed discussion on this issue in chapter 9 as well (and our publisher isn't happy if we post large bits of text from the book online, and I don't want even a whiff of trouble with them at the moment). But to give just a brief summary, G-I's theory was based on: (a) lexical reconstructions many of which have since been shown to be wrong, (b) evidence from contact with Kartvelian languages, which also has been challenged more recently, and (c) their version of the Glottalic theory, which (if correct) they use to support point (b) above about contacts with Proto-Kartvelian, but that theory itself is highly problematic, nor does it necessarily indicate contact with Proto-Kartvelian (which, by the way, may or may not be spoken south of the Caucasus, we just don't know). So there are a lot of suppositions their Armenian theory is based on, most of which been shown to be either wrong or quite probably wrong. If it helps, the relevant pages in the book are 192-198.
  36. @ryanwc
    Hmm. "A set of Algonquian languages" or "a set of Iroquoian languages" certainly makes sense. And in some cases they recoalesced in the aftermath of colonial contact, as elements of various tribes were thrown together. If, say, modern Ojibwa could grow out of an old Ojibwan core along with multifarious effects from various sister languages, I'm not sure why one couldn't refer to the Algonquian languages then extant as "a set of Proto-Ojibwan languages" in, say, 1492, . "A set of PIE-like languages" seems pretty plausible - related bands of hunter gatherers, speaking similar languages with varying degrees of mutual comprehension, coalescing into whatever tongue was spoken by the Yamnaya.

    The early divergence of the Anatolian branch of IE remains a bit of a quandary, no? Yamnaya develops expansive tendencies, and the first thing they do is head south to claim a patchwork of territories in Anatolia, in a lingering sea of non-IE. And only much later send their wagon trains out to exploit the softer targets scattered west and east of them. They bring with them to Anatolia their words for draft technology and cattle, technologies which have a long history in Anatolia prior to the new IE speakers. But they fail to establish their words for horse, wheel and cart, which are the very new technologies that give them their superiority?

    Linguistically, it still seems more convincing to highlight the IE roots for cattle and for draft tools shared across Anatolian and the rest of IE, unlike the horse/wheel terminology shared only post-Anatolian. This seems to locate the earliest PIE in one of the dairy cultures of Anatolia. Kartvelian genetics might be our only proxy today for the genetics of a dairying people who crossed the Caucasus or circled the Black Sea, carrying their language into the Pontic steppe, where their skill with cattle gave them the advantage in domesticating a more unruly species, the horse. Something which steppe peoples hadn't accomplished in the previous millennia.

    I recognize how one of your points, Razib, cuts against what I'm writing. If all the technologies came with the southern peoples, what advantage accrued to the HG that allowed them to contribute so much of the genetics? Is it not the affinity of HG for a pastoral lifestyle that allows them to steal cattle, tame the horse, and take the upper hand in the relationship with whatever other peoples contributed to the genetic pool where PIE was bubbling up? You may be right.

    I'm just trying to play devil's advocate, while highlighting one weakness with the linguistic argument that has been advanced, often naively, based on an assumption that the only possible counter-narrative would be an expansion with the first farmers. There are still possibilities for later out-of-Anatolia expansion. .

    Is there Hittite DNA coming our way anytime soon?

    Linguistically, talking about “a set of PIE-like languages” doesn’t make sense for many reasons. For one thing, “PIE-like” can mean a lot of different things. But even if we restrict the term to “languages/dialects related to PIE through common descent”, it’s still not very illuminating. Let’s say we are discussing a certain Y-DNA haplogroup, i.e. men who have a certain mutation in common (or certain mutations, plural), and we are interested in finding out where and when their MRCA lived. How much sense would it make to talk about “a set of MRCA-like men” who lived maybe at this particular time or maybe 30% earlier, maybe here or maybe there? Not much. We can hypothesize (with 100% certainty!) that the ancestor of Indo-European languages (PIE) was spoken on planet Earth some time in the last 200KY. Great theory and definitely right, but not very scientific (falsifiable), is it?

    Regarding your idea that there were related (and very similar) dialects, that’s in all likelihood true, but they can be agglomerated under one language (just as English really means a bunch of different varieties). But still that entire bunch of PIE dialects could not have been spoken over a too-large area for any significant length of time. Because if it was, it would have developed into distinct languages. And we are interested in find out the time and place where that diversification happened. See more in the post that I linked to in a comment above.

    Re: the PIE/PNIE/PSIE terminology for horses, wheels, etc., I will have to refer you to our book, where almost an entire chapter deal with this issue in detail (including in linguistic detail), so I couldn’t possibly give it justice in a post here:

    http://www.amazon.com/Indo-European-Controversy-Fallacies-Historical-Linguistics/dp/1107054532/ref=sr_1_1?s=books&ie=UTF8&qid=1419572204&sr=1-1&keywords=indo-european+controversy

    Read More
  37. @spandrell
    A plausible model is that PIE and Uralic were similar languages spoken around the central steppes; then the PIE move west, conquer some Caucasus language speaking farmers, and the mixed language becomes IE.

    IE languages have both commonalities with Uralic and Caucasian languages. And the structure of nominal declension strongly hints at it being a mongrel language; the very idea of different declensions is very bad design. Most languages have a single set of functional morphemes which are quite transparent, IE languages though have messed the whole thing into an irregular system of sound roots which makes no sense; unless it was the result of the mixing of two different language families.

    Correction 1: PIE and Proto-Uralic were not at all linguistically similar, as far as we can reconstruct them.

    Correction 2: Every language (or almost every language) is a “mongrel language” to a degree, but we can diagnose such “mongrel”-ness by a particular linguistic feature, such as fusional morphology, which is what you are talking about here.

    Correction 3: Fusional morphology (including “different declensions”) isn’t “very bad design” any more than agglutinative morphology or isolating morphology. This issue is discussed in detail here:

    http://www.languagesoftheworld.info/historical-linguistics/more-on-word-order-morphological-types-and-historical-change.html

    Correction 4: “Most languages have a single set of functional morphemes which are quite transparent” — that’s factually false. A great many fusional and polysynthetic languages do not.

    Correction 5: “IE languages though have messed the whole thing into an irregular system of sound roots which makes no sense” — maybe it makes no sense to you, but it makes perfect sense in terms of the organization of morphological systems (see post linked above). Moreover, non-Pama-Nyungan languages in Australia and many Finno-Ugric languages likewise developed fusional morphology out of the agglutinative morphology (which according to you makes more sense)

    Read More
  38. @spandrell
    English is actually very irregular. English has hundreds of irregular verbs, even irregular plurals!

    English grammar might be simple; having shed most of its morphology, but it's more irregular than Saxon or French. And inflection needn't be hard; Finnish has way more inflection than any IE language, but it's very regular and not hard at all.

    Megalophias is right here, English has become much more regular over time. Old English had swealg and healp, now we have swallowed and helped. Old English had lambru and bēc, now we have lambs and books. Modern English has relatively few irregular forms, compared to its earlier forms, including Saxon. And although Finnish inflection is more regular, there is a lot more different morphemes. What is “hard” is a subjective notion, but when talking about morphological complexity, we talk about the number of distinctions made (fewer in Modern English than in Old English), the number of regular patterns, the number of exceptions etc.

    Read More
  39. @ryanwc
    Hmm. "A set of Algonquian languages" or "a set of Iroquoian languages" certainly makes sense. And in some cases they recoalesced in the aftermath of colonial contact, as elements of various tribes were thrown together. If, say, modern Ojibwa could grow out of an old Ojibwan core along with multifarious effects from various sister languages, I'm not sure why one couldn't refer to the Algonquian languages then extant as "a set of Proto-Ojibwan languages" in, say, 1492, . "A set of PIE-like languages" seems pretty plausible - related bands of hunter gatherers, speaking similar languages with varying degrees of mutual comprehension, coalescing into whatever tongue was spoken by the Yamnaya.

    The early divergence of the Anatolian branch of IE remains a bit of a quandary, no? Yamnaya develops expansive tendencies, and the first thing they do is head south to claim a patchwork of territories in Anatolia, in a lingering sea of non-IE. And only much later send their wagon trains out to exploit the softer targets scattered west and east of them. They bring with them to Anatolia their words for draft technology and cattle, technologies which have a long history in Anatolia prior to the new IE speakers. But they fail to establish their words for horse, wheel and cart, which are the very new technologies that give them their superiority?

    Linguistically, it still seems more convincing to highlight the IE roots for cattle and for draft tools shared across Anatolian and the rest of IE, unlike the horse/wheel terminology shared only post-Anatolian. This seems to locate the earliest PIE in one of the dairy cultures of Anatolia. Kartvelian genetics might be our only proxy today for the genetics of a dairying people who crossed the Caucasus or circled the Black Sea, carrying their language into the Pontic steppe, where their skill with cattle gave them the advantage in domesticating a more unruly species, the horse. Something which steppe peoples hadn't accomplished in the previous millennia.

    I recognize how one of your points, Razib, cuts against what I'm writing. If all the technologies came with the southern peoples, what advantage accrued to the HG that allowed them to contribute so much of the genetics? Is it not the affinity of HG for a pastoral lifestyle that allows them to steal cattle, tame the horse, and take the upper hand in the relationship with whatever other peoples contributed to the genetic pool where PIE was bubbling up? You may be right.

    I'm just trying to play devil's advocate, while highlighting one weakness with the linguistic argument that has been advanced, often naively, based on an assumption that the only possible counter-narrative would be an expansion with the first farmers. There are still possibilities for later out-of-Anatolia expansion. .

    Is there Hittite DNA coming our way anytime soon?

    Asya’s blog discusses a potential wheel-word cognate in Hittite (hurki–, per German Dziebel), but also insists that even the absence of such a word in Anatolian wouldn’t have been a problem because the Anatolian languages split off too early.

    There is also a fascinating discussion of South-of-the Caucasus origins of PIE in “The Indo-European Controversy”. Asya Pereltsvaig & Martin W. Lewis start from asserting that Gamkrelidze and Ivanov’s 1995 PIE reconstructions for the “Southern” words such as panther, lion, elephant may have been wrong. The words for “wine”, they write, although clearly shared in PIE (including Anatolian but not Tokharian), Proto-Kartvelian, and Proto-Semitic, may have been borrowed by PIE rather than from PIE.

    The most prominent critique in the book is reserved for the Glottalic theory of evolution of phonemes in the vicinity of ejective consonants (which, according to Gamkrelidze-Ivanov, have been borrowed from Proto-Kartvelian into PIE). The ejectives have been widely lost in the extant IE languages, although they’ve been re-acquired by languages which came in contact with the Kartvelians later (such as Armenian and Ossetian). Pereltsvaig & Lewis argue that numerous North Caucasus languages use ejectives as well, and so PIE may have acquired its ejectives North of the mountains too – in fact even Kartvelian languages could have acquired them from the North, opening a possibility that Proto-Kartvelian didn’t even have ejectives yet.

    North-West Caucasus, where Maikop Culture flourished immediately before Yamnaya, is of particular interest to Pereltsvaig & Lewis . They note that Proto-NW Caucasian shared a number of unusual features with PIE: consonant-rich sound systems including labio-velarization of consonants, tonal accent, gender, number suppletion of pronouns etc. But no borrowings between Proto-NWC and PIE have been documented. Of course Davidski tried using contemporary NW Caucasus DNA in his Yamnaya quest, but there is too much latter-age Steppe introgression there to make anything easy.

    Other borrowings, pre-dating Anatolian-PNIE split, seem to point even further South, to Proto-Semitic. For example, PIE roots for “seven” and “star” are from the Proto-Semitic words for “seven” and “Venus”, resp. (a review e.g. here http://www.eupedia.com/forum/threads/27182-Proto-Semitic-and-Proto-Indo-European ). Semitic languages weren’t spoken North of Upper Mesopotamia, and even there, not attested until 2800 BC, although Bayesian modeling suggests that Proto-Semitic is about a millennium older than that (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2839953/ )

    Read More
    • Replies: @Asya Pereltsvaig
    Thanks for the summary of relevant bits of our book. I guess it preempts the comment I've just written, but I only noticed your comment now...
  40. @ryanwc
    Hmm. "A set of Algonquian languages" or "a set of Iroquoian languages" certainly makes sense. And in some cases they recoalesced in the aftermath of colonial contact, as elements of various tribes were thrown together. If, say, modern Ojibwa could grow out of an old Ojibwan core along with multifarious effects from various sister languages, I'm not sure why one couldn't refer to the Algonquian languages then extant as "a set of Proto-Ojibwan languages" in, say, 1492, . "A set of PIE-like languages" seems pretty plausible - related bands of hunter gatherers, speaking similar languages with varying degrees of mutual comprehension, coalescing into whatever tongue was spoken by the Yamnaya.

    The early divergence of the Anatolian branch of IE remains a bit of a quandary, no? Yamnaya develops expansive tendencies, and the first thing they do is head south to claim a patchwork of territories in Anatolia, in a lingering sea of non-IE. And only much later send their wagon trains out to exploit the softer targets scattered west and east of them. They bring with them to Anatolia their words for draft technology and cattle, technologies which have a long history in Anatolia prior to the new IE speakers. But they fail to establish their words for horse, wheel and cart, which are the very new technologies that give them their superiority?

    Linguistically, it still seems more convincing to highlight the IE roots for cattle and for draft tools shared across Anatolian and the rest of IE, unlike the horse/wheel terminology shared only post-Anatolian. This seems to locate the earliest PIE in one of the dairy cultures of Anatolia. Kartvelian genetics might be our only proxy today for the genetics of a dairying people who crossed the Caucasus or circled the Black Sea, carrying their language into the Pontic steppe, where their skill with cattle gave them the advantage in domesticating a more unruly species, the horse. Something which steppe peoples hadn't accomplished in the previous millennia.

    I recognize how one of your points, Razib, cuts against what I'm writing. If all the technologies came with the southern peoples, what advantage accrued to the HG that allowed them to contribute so much of the genetics? Is it not the affinity of HG for a pastoral lifestyle that allows them to steal cattle, tame the horse, and take the upper hand in the relationship with whatever other peoples contributed to the genetic pool where PIE was bubbling up? You may be right.

    I'm just trying to play devil's advocate, while highlighting one weakness with the linguistic argument that has been advanced, often naively, based on an assumption that the only possible counter-narrative would be an expansion with the first farmers. There are still possibilities for later out-of-Anatolia expansion. .

    Is there Hittite DNA coming our way anytime soon?

    You’re ignoring the ancient DNA evidence.

    The steppe wasn’t settled by anyone native to Anatolia. Neolithic samples from western and central Anatolia have been sequenced and they don’t look ancestral to Yamnaya, not even in part.

    On the other hand, Anatolia saw major population turnovers after the Neolithic, including the arrival at some point of steppe-related ancestry.

    The important thing to establish now is when the steppe-related ancestry first arrived in Anatolia.

    Read More
    • Replies: @ryanwc
    Davidski,

    You're likely right. As are most of the ideas I've advanced arguments against here.

    Part of what I'm trying to do is game out to what degree alternatives remain not quite disproven.

    Razib outlines a setting in which a number of (relatively unrelated) tribes or cultures in close proximity to the Neolithic explosion may all have expanded outward while initially maintaining their separate languages, and while language is not coterminous with genetics, in the middle run, it's often pretty closely related, because aside from tiny HG groups practicing intentional exogamy to avoid inbreeding, most language groups are fairly endogamous.

    During the initial stages of the development of this toolkit it does not seem that there was any particular advantage to single group in a narrow delimited zone. To be more concrete, the linguistic diversity of the Caucasus region, or what we see and know from the edges of history in the Near East, may be close to the reality of the period of the early Neolithic in the Near East. Different polities with radically different languages, and likely divergent genetics, were all crystallizing the new lifestyle in a cheek by jowl fashion in the hills between the Tarsus and Zagros mountains.
     
    In that context, I would ask how many western and central Anatolian samples would be sufficient to be certain that we knew the genetic make-up of all the ethnic groups of Anatolia, and could strike them all as possible forebears to Yamnaya?

    How many Cucuteni genomes do we have? The Ryan Schmidt (a different Ryan, FWIW) poster available on Academia suggests Cucuteni mtDNA isn't a bad match for Yamna - it clusters in H and T2, so only U5 is missing from what I understand to be the primary Yamna haplogroups. The poster seems to be from this year, and it cites only one other work on Cucuteni.

    Your project on Kartvelians is very interesting, but when the options you had to hand were modern populations and ancient Armenians, I'm not convinced that it is final and definitive.

    Are we even certain of the level of the Black Sea at 5,000-4,500 BC? Might there be a contributing population whose traces are hiding under the alluvium of 6 millennia?

    Again, I do concede that yours, Asya's, Razib's are the most likely answers. To some degree, pushing alternatives just helps me digest the mainstream argument. I don't have the range of reading and knowledge that many others here do. So thanks for indulging me enough to answer my previous post.
  41. Asya Pereltsvaig & Martin W. Lewis start from asserting that Gamkrelidze and Ivanov’s 1995 PIE reconstructions for the “Southern” words such as panther, lion, elephant may have been wrong.

    their reference there is to a 1995 book by R. Beekes, “Comparative Indo-European Linguistics: An Introduction”, which is partly available online

    https://books.google.com/books?id=i_JwBsKzgeAC&pg=PA35&source=gbs_toc_r&cad=3#v=onepage&q&f=false

    As it turns out, Beekes didn’t really disprove Gamkrelidze & Ivanov’s reconstructions, but just pointed out that the quality of reconstructions of the Proto-Semitic itself, in 1995, wasn’t sufficient to link the PIE reconstructions to Semitic borrowings. Beeks also pointed out that PIE had words for “mountain” and for “snow” (possibly mountain snow) consistent with either slope of the Caucasus. But he sided with Gimbutas that the presence of a word for horse was indicative of the North side of the Caucasus, since any use of horses further South post-dates the Kurgan cultures

    Read More
  42. @spandrell
    English is actually very irregular. English has hundreds of irregular verbs, even irregular plurals!

    English grammar might be simple; having shed most of its morphology, but it's more irregular than Saxon or French. And inflection needn't be hard; Finnish has way more inflection than any IE language, but it's very regular and not hard at all.

    English has lots of irregular verbs, but they are all inherited from Germanic. Verbs borrowed from French or Latin are nice and regular. (Which is why if you are an English speaker studying French, the irregular verbs are a pain in the ass, whereas in German half the time you don’t even notice they are irregular – singen sang gesungen, bringen bracht gebracht).

    Most irregular plurals are inherited too (e.g. mouse-mice, ox-oxen), except when both the singular and the plural are borrowed together (usually from Greek or Latin). But those borrowed plurals are not productive except in wordplay or mistaken attribution (e.g. octopi).

    Read More
    • Replies: @Reg Cæsar

    …But those borrowed plurals are not productive except in wordplay or mistaken attribution (e.g. octopi)
     
    …or octopodēs, which may be just as wrong, but more fun to pronounce.
  43. Gamkrelidze and Ivanov’s story of Hittite leopards is a good read by itself – the sacred big cat is depicted in frescoes and in ritual poetry, it dances (probably leopard skin-clad humans perform the dance?), it denotes feminine fertility. Hittite “parsana” ~ Farsi pars and palang, borrowed into Turkic and Mongolian bars, Greek pardos, Sanskr. prdaku.

    Hittite word for lion, ‘walwa” is interpreted as a partial reduplication of PIE *leu.

    As to proto-Kartvelian / proto-Semitic relations: Johanna Nichols, a Kartvelian specialist, argued that extant Kartvelian languages have a comparatively too recent ancestor, and that other, presently extinct branches of the family should existed in the past. In her opinion, neither Kartvelian nor Semitic borrowings into PIE tended to be direct, because there is so much phonological change. Some intermediary, perhaps an extinct branch of today’s Kartvelian family, may have transmitted Semitic borrowings to PIE.

    The talk inevitable turns to Kura-Araxes culture, the regional powerhouse of metallurgy with its cultural affinities both to Maikop (ceramics styles, kurgan burials, wheeled carts) and to the far reaches of the late-period Uruk Mesopotamian expansion?

    Read More
    • Replies: @Asya Pereltsvaig
    Good point about Proto-Kartvelian. I will ask Johanna Nichols a bit more about this when she does a guest lecture in my course on 11/9...
  44. @terryt
    "I believe that Indo-European and Afro-Asiatic may both be language families of hunter-gatherer populations who absorbed migrating farmers into a radically different ecology where the competitive playing field was much more level than in the initial zone of expansion".

    Interesting idea. I think you are onto something there.

    I have my own speculation about ethnogenesis of Armenians and Kurds, which happened at the end of the Neolithic period.

    Armenians are descendants of the eastward migrating Indo-European commoners, who end-up being ruled by the Urartian (Caucasian) nobility.

    Kurds are to the contrary descendants of the native Hurrian (Caucasian) commoners who were conquered by Indo-Aryan Mitanni and later by Median nobility.

    “Founded by an Indo-Aryan ruling class governing a predominantly Hurrian population” [1]

    Interestingly Mitanni in Hittite written with the cuneiform KUR [2] (cognates with Kurd) which usually referred to the Zagros mountains.

    [1] https://en.wikipedia.org/wiki/Mitanni

    [2] https://en.wikipedia.org/wiki/Kur

    Read More
  45. I believe that…

    Makes sense – farmers spreading east and west relatively easily but going north and south they reach the edge of viability and by definition the edge of viability would be where HGs could potentially get the upper hand.

    Read More
  46. @Matt_
    An idea I had on early Indo-European society was a social model very concerned with passing on patrilines (patriarchy, father-son relationships, etc), so maybe did not put as much in the way of resources into behind marriage, etc. for daughters.

    Daughters from outside the culture, from other more matriline and female fertility concerned cultures, who might have been more able to inherit resources from their family (in terms of assistance and material resources), might then have made more attractive partners for men within the IE culture. Equally males within the culture could draw on a strong base of resources and support less available to males outside the culture, helping their prospects of reproducing and spreading their y dna disproportionately.

    Thus turnover of female lineages from outside the culture being drawn into the culture (although perhaps with the Basques and their progenitor culture it may have gone the other way, with males from outside the culture introgressing into the culture). After a certain amount of time, this complex would have to stabilise as it reached its limits.

    May all very well be a fantasy though, and perhaps it was more violent.

    I tend to think it was mostly captives but that may just be my personality talking.

    Although a raider population over time gradually turning into the people they raid (autosomally) seems like a straightforward logical consequence.

    Another more peaceful possibility I wondered about might be a variation of Clarkian effect where the steppe chiefs had arranged dynastic marriages with caucasus chiefs and their offspring had a reproductive advantage over time hence a top down spread of their particular dna mix.

    Read More
  47. @Asya Pereltsvaig
    Nor does it make sense that PIE speakers were pre-agricultural and pre-horse peoples. Read our book, we talk about this in detail...

    Isn’t PIE considered to be a point of transition – if so wouldn’t it have to be both pre and post agricultural/horse?

    Read More
  48. @Dmitry Pruss
    Asya, you have a fascinating style of discussion, where very important factual observations on grammar, vocabulary, phonetics etc. are intermixed with the bits of general wisdom like "pre-Any-Other-Language was in Africa". I just assume that you were irritated by my or Razib's cautious approach to assigning names to the various proto-states of early splits of the IE languages. But isn't it a widespread issue, that there is no universal agreement yet, how to call these descendants or predecessors of PIE? In your own articles, you've variously proposed PNIE or perhaps PSIE for the Yamnaya culture (either a predecessor of IE languages other than Anatolian, or perhaps a predecessor of the extant IE languages, that is other than Anatolian and Tokharian), or Proto-Indo-Hittite which would have been the root of Anatolian and PNIE if we decided to apply the label PIE to the PNIE? These are all very interesting suggestions, but they also make it clear that the terminology of these proto-languages is in flux. I don't think I deserve the blame for trying to navigate the maze of these proto-language names with caution :) I certainly meant no disrespect and I remain your loyal fan.

    I understand your well-articulated position that the Anatolian languages split off before Yamnaya, and originated outside of Anatolia, moving in likely from the North-East (the Caucasus) or North-West (the Balkans). I also know that you wrote an extensive critique of Gamkrelidze and Ivanov's South Caucasus Urheimat theory, which they built upon the observations (perhaps faulty) of the Proto-Kartvelian influences early in IE history. The recent discovery that the extant Kartvelians are similar to the agriculturalist ancestors of the Yamnaya thrusts the Gamkrelidze-Ivanov theory to the forefront ... the geography and the genetics are kind of right, and they claim that the linguistics is right too. I would really appreciate hearing your detailed critique of what's wrong with Gamkrelidze-Ivanov.

    I am sorry if I sounded annoyed, but I think the precise use of terminology and also understanding what the question is before comparing answers is crucial. Regarding the terminology, I don’t think it’s “in flux”. We certainly didn’t invent “PNIE” (for the ancestor of Indo-European languages minus Anatolian), though I did invent “PSIE” (as far as I know there was no label for the ancestor of all Indo-European languages minus Anatolian and Tocharian). But I wouldn’t call these “PIE-like languages” and that’s certainly not a reading I thought of when I read your comment. Also, I am quite opposed to the use of “Indo-Hittite” or “Proto-Indo-Hittite” as it is muddling the issue and using it means that other terms like “PIE” are understood to mean something other than what they are usually understood to mean.

    As for matching these languages to archaeological cultures, I remain for the most part agnostic about this, although we do discuss it in the book a little (chapter 9). The main argument is that based on our archaeological knowledge, the Steppe theory provides a better story for migrations than the Anatolian theory.

    As for Gamkrelidze-Ivanov theory, I have a detailed discussion on this issue in chapter 9 as well (and our publisher isn’t happy if we post large bits of text from the book online, and I don’t want even a whiff of trouble with them at the moment). But to give just a brief summary, G-I’s theory was based on: (a) lexical reconstructions many of which have since been shown to be wrong, (b) evidence from contact with Kartvelian languages, which also has been challenged more recently, and (c) their version of the Glottalic theory, which (if correct) they use to support point (b) above about contacts with Proto-Kartvelian, but that theory itself is highly problematic, nor does it necessarily indicate contact with Proto-Kartvelian (which, by the way, may or may not be spoken south of the Caucasus, we just don’t know). So there are a lot of suppositions their Armenian theory is based on, most of which been shown to be either wrong or quite probably wrong. If it helps, the relevant pages in the book are 192-198.

    Read More
  49. @Dmitry Pruss
    Gamkrelidze and Ivanov's story of Hittite leopards is a good read by itself - the sacred big cat is depicted in frescoes and in ritual poetry, it dances (probably leopard skin-clad humans perform the dance?), it denotes feminine fertility. Hittite "parsana" ~ Farsi pars and palang, borrowed into Turkic and Mongolian bars, Greek pardos, Sanskr. prdaku.

    Hittite word for lion, 'walwa" is interpreted as a partial reduplication of PIE *leu.

    As to proto-Kartvelian / proto-Semitic relations: Johanna Nichols, a Kartvelian specialist, argued that extant Kartvelian languages have a comparatively too recent ancestor, and that other, presently extinct branches of the family should existed in the past. In her opinion, neither Kartvelian nor Semitic borrowings into PIE tended to be direct, because there is so much phonological change. Some intermediary, perhaps an extinct branch of today's Kartvelian family, may have transmitted Semitic borrowings to PIE.

    The talk inevitable turns to Kura-Araxes culture, the regional powerhouse of metallurgy with its cultural affinities both to Maikop (ceramics styles, kurgan burials, wheeled carts) and to the far reaches of the late-period Uruk Mesopotamian expansion?

    Good point about Proto-Kartvelian. I will ask Johanna Nichols a bit more about this when she does a guest lecture in my course on 11/9…

    Read More
    • Replies: @Dmitry Pruss

    Good point about Proto-Kartvelian. I will ask Johanna Nichols a bit more
     
    As I understand, she noted that in historic times, languages kept spreading West across Central Asia, and only very rarely in different directions. She didn't think that the conditions for the earlier migrations differed all that much [and I'm not going to argue with that, for the sake of brevity]. Ergo, Indo-Europeans, Kartvelians, etc. also must have spread West across Central Asia, in her opinion?
  50. @Dmitry Pruss
    Asya's blog discusses a potential wheel-word cognate in Hittite (hurki–, per German Dziebel), but also insists that even the absence of such a word in Anatolian wouldn't have been a problem because the Anatolian languages split off too early.

    There is also a fascinating discussion of South-of-the Caucasus origins of PIE in "The Indo-European Controversy". Asya Pereltsvaig & Martin W. Lewis start from asserting that Gamkrelidze and Ivanov's 1995 PIE reconstructions for the "Southern" words such as panther, lion, elephant may have been wrong. The words for "wine", they write, although clearly shared in PIE (including Anatolian but not Tokharian), Proto-Kartvelian, and Proto-Semitic, may have been borrowed by PIE rather than from PIE.

    The most prominent critique in the book is reserved for the Glottalic theory of evolution of phonemes in the vicinity of ejective consonants (which, according to Gamkrelidze-Ivanov, have been borrowed from Proto-Kartvelian into PIE). The ejectives have been widely lost in the extant IE languages, although they've been re-acquired by languages which came in contact with the Kartvelians later (such as Armenian and Ossetian). Pereltsvaig & Lewis argue that numerous North Caucasus languages use ejectives as well, and so PIE may have acquired its ejectives North of the mountains too - in fact even Kartvelian languages could have acquired them from the North, opening a possibility that Proto-Kartvelian didn't even have ejectives yet.

    North-West Caucasus, where Maikop Culture flourished immediately before Yamnaya, is of particular interest to Pereltsvaig & Lewis . They note that Proto-NW Caucasian shared a number of unusual features with PIE: consonant-rich sound systems including labio-velarization of consonants, tonal accent, gender, number suppletion of pronouns etc. But no borrowings between Proto-NWC and PIE have been documented. Of course Davidski tried using contemporary NW Caucasus DNA in his Yamnaya quest, but there is too much latter-age Steppe introgression there to make anything easy.

    Other borrowings, pre-dating Anatolian-PNIE split, seem to point even further South, to Proto-Semitic. For example, PIE roots for "seven" and "star" are from the Proto-Semitic words for "seven" and "Venus", resp. (a review e.g. here http://www.eupedia.com/forum/threads/27182-Proto-Semitic-and-Proto-Indo-European ). Semitic languages weren't spoken North of Upper Mesopotamia, and even there, not attested until 2800 BC, although Bayesian modeling suggests that Proto-Semitic is about a millennium older than that (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2839953/ )

    Thanks for the summary of relevant bits of our book. I guess it preempts the comment I’ve just written, but I only noticed your comment now…

    Read More
    • Replies: @Dmitry Pruss
    Asya, I tried summarizing the key objections to Gamkrelidze & Ivanov's conclusions, but some parts give me more questions then answers. Like you quoted R. Beekes, “Comparative Indo-European Linguistics: An Introduction” (1995) to make the point that Gamkrelidze's PIE reconstruction of "panther" is wrong. But when I check Beekes, he doesn't disagree with Gamkrelidze & Ivanov’s reconstruction ... Beekes just notes that it may not be conclusively attributed to borrowing from Proto-Semitic.

    So I look what Gamkrelidze and Ivanov had to say, and, lo and behold, they rather unequivocally say that the PIE word for "panther" was NOT from Proto-Semitic *nimr (Fronzarolli 1968). Rather, they show a continuity of the panther cult traditions between the pre-Hittite Anatolia and the Hittite texts (which is not surprising by itself) and non-Anatolian IE languages, to hypothesize that the word *pars/*pard entered IE languages South of the Caucasus with these pre-IE religious traditions. They also hypothesize that non-IE borrowed word ended with a dental fricative in the original source language, which is why some IE languages got is as "s" and others as "d".

    Now it wouldn't be unreasonable to argue with either of these finer points (spread of Anatolian pre-IE cults / dental fricative sound), but instead Beekes argues that the word isn't from Proto-Semitic (which G&I didn't claim in the first place) and you argue that the PIE reconstruction must be wrong (which Beeker never claimed, in turn). It almost makes me guess that Gamkrelidze's theories were so annoying and long-winded that the critics never got patience to actually read the stuff?

  51. A little off topic, but is there any genetic evidence for Veblen’s hypothesis that the earliest military conquests in that part of the world were carried out by pastoralists against settled agriculturalists? Also, is there any genetic evidence yet for where the Sumerians originated?

    Read More
  52. @Davidski
    You're ignoring the ancient DNA evidence.

    The steppe wasn't settled by anyone native to Anatolia. Neolithic samples from western and central Anatolia have been sequenced and they don't look ancestral to Yamnaya, not even in part.

    On the other hand, Anatolia saw major population turnovers after the Neolithic, including the arrival at some point of steppe-related ancestry.

    The important thing to establish now is when the steppe-related ancestry first arrived in Anatolia.

    Davidski,

    You’re likely right. As are most of the ideas I’ve advanced arguments against here.

    Part of what I’m trying to do is game out to what degree alternatives remain not quite disproven.

    Razib outlines a setting in which a number of (relatively unrelated) tribes or cultures in close proximity to the Neolithic explosion may all have expanded outward while initially maintaining their separate languages, and while language is not coterminous with genetics, in the middle run, it’s often pretty closely related, because aside from tiny HG groups practicing intentional exogamy to avoid inbreeding, most language groups are fairly endogamous.

    During the initial stages of the development of this toolkit it does not seem that there was any particular advantage to single group in a narrow delimited zone. To be more concrete, the linguistic diversity of the Caucasus region, or what we see and know from the edges of history in the Near East, may be close to the reality of the period of the early Neolithic in the Near East. Different polities with radically different languages, and likely divergent genetics, were all crystallizing the new lifestyle in a cheek by jowl fashion in the hills between the Tarsus and Zagros mountains.

    In that context, I would ask how many western and central Anatolian samples would be sufficient to be certain that we knew the genetic make-up of all the ethnic groups of Anatolia, and could strike them all as possible forebears to Yamnaya?

    How many Cucuteni genomes do we have? The Ryan Schmidt (a different Ryan, FWIW) poster available on Academia suggests Cucuteni mtDNA isn’t a bad match for Yamna – it clusters in H and T2, so only U5 is missing from what I understand to be the primary Yamna haplogroups. The poster seems to be from this year, and it cites only one other work on Cucuteni.

    Your project on Kartvelians is very interesting, but when the options you had to hand were modern populations and ancient Armenians, I’m not convinced that it is final and definitive.

    Are we even certain of the level of the Black Sea at 5,000-4,500 BC? Might there be a contributing population whose traces are hiding under the alluvium of 6 millennia?

    Again, I do concede that yours, Asya’s, Razib’s are the most likely answers. To some degree, pushing alternatives just helps me digest the mainstream argument. I don’t have the range of reading and knowledge that many others here do. So thanks for indulging me enough to answer my previous post.

    Read More
    • Replies: @ryanwc
    I did a little more searching for Tripolye Cucuteni genetic data. I might have had more success if my slavic language ability went beyond djiekuje. Interestingly, Schmidt references another mtDNA study from 2010, of some very early TC individuals. On the one hand, they show little similarity with the mtDNA from Yamna published this year. On the other, they don't have much in common with Schmidt's TC samples. Was one set of samples mistakenly attributed? Was there a sea change in mtDNA of Cucuteni over time? Was there genetic structure across the area covered by Cucuteni, structure that may have shifted on the map over a few hundred years? Should results like these suggest at least some caution about our ability to generalize until we have bigger data sets to work with.

    No one answered my earlier question. Do we have any Hittite genomes, or any hints of Hittite genomes to come? Probably because there is none, and maybe because it's a dumb question - until recently, was it thought that preservation was insufficient? Maybe the new work on petrous bone will encourage more sampling from populations thought to be untestable?
  53. @ryanwc
    Davidski,

    You're likely right. As are most of the ideas I've advanced arguments against here.

    Part of what I'm trying to do is game out to what degree alternatives remain not quite disproven.

    Razib outlines a setting in which a number of (relatively unrelated) tribes or cultures in close proximity to the Neolithic explosion may all have expanded outward while initially maintaining their separate languages, and while language is not coterminous with genetics, in the middle run, it's often pretty closely related, because aside from tiny HG groups practicing intentional exogamy to avoid inbreeding, most language groups are fairly endogamous.

    During the initial stages of the development of this toolkit it does not seem that there was any particular advantage to single group in a narrow delimited zone. To be more concrete, the linguistic diversity of the Caucasus region, or what we see and know from the edges of history in the Near East, may be close to the reality of the period of the early Neolithic in the Near East. Different polities with radically different languages, and likely divergent genetics, were all crystallizing the new lifestyle in a cheek by jowl fashion in the hills between the Tarsus and Zagros mountains.
     
    In that context, I would ask how many western and central Anatolian samples would be sufficient to be certain that we knew the genetic make-up of all the ethnic groups of Anatolia, and could strike them all as possible forebears to Yamnaya?

    How many Cucuteni genomes do we have? The Ryan Schmidt (a different Ryan, FWIW) poster available on Academia suggests Cucuteni mtDNA isn't a bad match for Yamna - it clusters in H and T2, so only U5 is missing from what I understand to be the primary Yamna haplogroups. The poster seems to be from this year, and it cites only one other work on Cucuteni.

    Your project on Kartvelians is very interesting, but when the options you had to hand were modern populations and ancient Armenians, I'm not convinced that it is final and definitive.

    Are we even certain of the level of the Black Sea at 5,000-4,500 BC? Might there be a contributing population whose traces are hiding under the alluvium of 6 millennia?

    Again, I do concede that yours, Asya's, Razib's are the most likely answers. To some degree, pushing alternatives just helps me digest the mainstream argument. I don't have the range of reading and knowledge that many others here do. So thanks for indulging me enough to answer my previous post.

    I did a little more searching for Tripolye Cucuteni genetic data. I might have had more success if my slavic language ability went beyond djiekuje. Interestingly, Schmidt references another mtDNA study from 2010, of some very early TC individuals. On the one hand, they show little similarity with the mtDNA from Yamna published this year. On the other, they don’t have much in common with Schmidt’s TC samples. Was one set of samples mistakenly attributed? Was there a sea change in mtDNA of Cucuteni over time? Was there genetic structure across the area covered by Cucuteni, structure that may have shifted on the map over a few hundred years? Should results like these suggest at least some caution about our ability to generalize until we have bigger data sets to work with.

    No one answered my earlier question. Do we have any Hittite genomes, or any hints of Hittite genomes to come? Probably because there is none, and maybe because it’s a dumb question – until recently, was it thought that preservation was insufficient? Maybe the new work on petrous bone will encourage more sampling from populations thought to be untestable?

    Read More
    • Replies: @Davidski
    We don't yet have any Hittite genomes, but it's likely that at some point they will be available.

    TC genomes might also be available soon. But since TC is partly of Starcevo origin, and we already know that Starcevo people were in large part of Neolithic Anatolian origin, then it's unlikely that they'll fit the bill as ancestral to Yamnaya and related groups on the steppe.
  54. @Megalophias
    English has lots of irregular verbs, but they are all inherited from Germanic. Verbs borrowed from French or Latin are nice and regular. (Which is why if you are an English speaker studying French, the irregular verbs are a pain in the ass, whereas in German half the time you don't even notice they are irregular - singen sang gesungen, bringen bracht gebracht).

    Most irregular plurals are inherited too (e.g. mouse-mice, ox-oxen), except when both the singular and the plural are borrowed together (usually from Greek or Latin). But those borrowed plurals are not productive except in wordplay or mistaken attribution (e.g. octopi).

    …But those borrowed plurals are not productive except in wordplay or mistaken attribution (e.g. octopi)

    …or octopodēs, which may be just as wrong, but more fun to pronounce.

    Read More
  55. @notanon
    Isn't PIE considered to be a point of transition - if so wouldn't it have to be both pre and post agricultural/horse?

    Linguistically, no.

    Read More
  56. @Davidski
    I think they probably did in many cases, or at least at the very beginning. But something like this might have happened too. This is in Bronze Age Germany, but it's probably a population of steppe origin to some degree...

    On the other hand, local population continuity was a prerequisite for the accumulation of wealth, the establishment of enduring social differentiation, and the formation of regional elites.

    ...

    The results indicate both local genetic continuity spanning the cultural transition, and, following the onset of the Early Bronze Age, a major influx of mtDNA types previously not found in this region. Integrating stable isotope data with the genetic data reveals a picture of a patrilocal society with remarkable mobility in women.
     

    Mittnik et al., Ancient DNA reveals patterns of residential continuity and mobility at the onset of the Central European Bronze Age

    http://smbe2015.at/program/program-and-abstracts/

    Shift to the far-distance mobility of the women in Bavaria around 2000 BCE (in Alissa Mittnik’s abstract) is remarkable but still fades in comparison with your proposed wholesale change of all mtDNA earlier in the Steppe. Could you describe in a few more words what you see in Khvalynsk and Yamnaya, please? Is it the mtDNA spectrum completely matching the agricultural South in Khvalynsk already? What HG groups does it contrast to?

    I’m tying to understand if the mtDNA transition may have been an earlier, separate phenomenon from the gradual increase of the agriculturalist autosomal DNA in the 4th and 3rd millennia BCE (which as I understand went from about 1/4 in Khvalynsk to less one half in early Yamnaya to more than 50% in late-period Yamnaya?)

    Read More
    • Replies: @Davidski
    It wasn't a wholesale change in mtDNA, because Yamnaya still show a decent percentage of mtDNA U lineages which are likely to be of hunter-gatherer origin.

    But what appears to have happened is that around 4,000 BC during the Khvalynsk phase, highly patriarchal, patrilocal, and territorial bands of young males of steppe hunter-gatherer ancestry got their women from wherever they could, and probably because more southerly populations with a longer history of agriculture and pastoralism had bigger populations, most of the women came from these societies, one way or another.

    This process continued until the genome-wide makeup of the steppe groups was ~50% Georgian-like, and it seems that at this point the population density on the steppe was high enough so that it was no longer necessary to import women from the edges of the steppe.

    Then the steppe started to dry up rapidly from 2,600 BC and a lot of the bands moved west into Europe, and took their languages with them.
  57. @ryanwc
    I did a little more searching for Tripolye Cucuteni genetic data. I might have had more success if my slavic language ability went beyond djiekuje. Interestingly, Schmidt references another mtDNA study from 2010, of some very early TC individuals. On the one hand, they show little similarity with the mtDNA from Yamna published this year. On the other, they don't have much in common with Schmidt's TC samples. Was one set of samples mistakenly attributed? Was there a sea change in mtDNA of Cucuteni over time? Was there genetic structure across the area covered by Cucuteni, structure that may have shifted on the map over a few hundred years? Should results like these suggest at least some caution about our ability to generalize until we have bigger data sets to work with.

    No one answered my earlier question. Do we have any Hittite genomes, or any hints of Hittite genomes to come? Probably because there is none, and maybe because it's a dumb question - until recently, was it thought that preservation was insufficient? Maybe the new work on petrous bone will encourage more sampling from populations thought to be untestable?

    We don’t yet have any Hittite genomes, but it’s likely that at some point they will be available.

    TC genomes might also be available soon. But since TC is partly of Starcevo origin, and we already know that Starcevo people were in large part of Neolithic Anatolian origin, then it’s unlikely that they’ll fit the bill as ancestral to Yamnaya and related groups on the steppe.

    Read More
    • Replies: @Dmitry Pruss
    Still fascinated by the R1b transition to R1a-Z93 in Poltavka and Srubnaya which coincides with the emergence of European Neolithic ancestry in the East of the region (and obviously reinforces the opinion that the Indo-Iranians descend from these Steppe cultures).

    Davidski - what do you think of the earliest R1a-Z93, Poltavka male 10432 / SVP42? He was buried at Potapovka I, a cemetery which continued to be used in later Bronze Age and in a grave which intersected with a later burial, in a way which removed 60% of the bones, but apparently was directly carbon-dated to 2900-2500 BCE. Note that this Poltavka sample was an outlier similar to Srubnaya in autosomal composition, while the other 3 mapped with Yamnaya. Could it be a misplaced cranium from an overlapping later-age burial? Do we know what material was used for carbon-dating?

    Archaeologically, Srubnaya burials has long been likened to the enclosed ossiaries in Central-East Europe's Globular Amphora Culture, of 3400-2800 BC (and perhaps both explained by a proto-Avestan belief against contamination of sacred Earth). Are there any Globe Amphora DNA data?

    Of course another theory (Sarianidi etc.) links the emergence of Avestan beliefs to Bactria-Margiana and the city culture of Gonur-Depe (2500 BC and on) (purity of flame and cinders, human bones stripped of flesh for burial, special burials of "impure" humans, ritual sacrifice of colts with cut heads and tails...) Soghd and Margiana appear to be mentioned right after the mystic urheimat of Avestans in the myth. Andronovo-type culture existed in the surrounding lands; Andronovo ceramics becomes frequent in town.The non-IE language of Margiana has been hypothesized by Lubotsky to be the source of numerous borrowings shared in Proto-Indic and Iranian languages but not elsewhere in the IE family, which refer to construction, irrigation, clothes, utensils, and cult, as well as to animals and birds many of which - like donkey and camel - were domesticated in Margiana-Bactria. The group origin of the Margiana language can't be reconstructed but its features are similar to the borrowings specific to Proto-Indic but not Iranian, indicating longer or more extensive contact of the Proto-Indians with this unknown language. But I'm sure there is no aDNA data from anywhere in Margiana...

  58. I saw this paper just came out. I’m sure Razib will have a more detailed post regarding it soon. Since it deals with Africa rather than Europe, I won’t discuss it here in detail.

    I can’t read the paper due to a paywall, but the interesting thing the supplementary materials note is that the West Eurasian admixture into Sub-Saharan Africans (which is shared to a limited degree by all African populations, even Pygmies and West Africans) is a better fit if modeled as Sardinian or even LBK rather than as the (non-SSA admixed) portion of the Druze genome.

    It seems that either the first “push” of farmers into Africa were not Afro-Asiatic (instead being part of the same group as “Old Europe” whose languages were later erased) or that the genomes of contemporary Semites are not similar to those who spoke Proto-Afro Asiatic at all. The latter is most intriguing – what the heck are modern Semites (particularly Arabs) in this case?

    Read More
    • Replies: @Josh Lipson
    R1b-V88? Doesn't fit the Afro-Asiatic bill. Were any Chadic-speaking populations tested here?
  59. @Dmitry Pruss
    Shift to the far-distance mobility of the women in Bavaria around 2000 BCE (in Alissa Mittnik's abstract) is remarkable but still fades in comparison with your proposed wholesale change of all mtDNA earlier in the Steppe. Could you describe in a few more words what you see in Khvalynsk and Yamnaya, please? Is it the mtDNA spectrum completely matching the agricultural South in Khvalynsk already? What HG groups does it contrast to?

    I'm tying to understand if the mtDNA transition may have been an earlier, separate phenomenon from the gradual increase of the agriculturalist autosomal DNA in the 4th and 3rd millennia BCE (which as I understand went from about 1/4 in Khvalynsk to less one half in early Yamnaya to more than 50% in late-period Yamnaya?)

    It wasn’t a wholesale change in mtDNA, because Yamnaya still show a decent percentage of mtDNA U lineages which are likely to be of hunter-gatherer origin.

    But what appears to have happened is that around 4,000 BC during the Khvalynsk phase, highly patriarchal, patrilocal, and territorial bands of young males of steppe hunter-gatherer ancestry got their women from wherever they could, and probably because more southerly populations with a longer history of agriculture and pastoralism had bigger populations, most of the women came from these societies, one way or another.

    This process continued until the genome-wide makeup of the steppe groups was ~50% Georgian-like, and it seems that at this point the population density on the steppe was high enough so that it was no longer necessary to import women from the edges of the steppe.

    Then the steppe started to dry up rapidly from 2,600 BC and a lot of the bands moved west into Europe, and took their languages with them.

    Read More
    • Replies: @Dmitry Pruss
    Thanks, cool!

    around 4,000 BC during the Khvalynsk phase, highly patriarchal, patrilocal, and territorial bands of young males of steppe hunter-gatherer ancestry got their women from wherever they could, and probably because more southerly populations with a longer history of agriculture and pastoralism had bigger populations, most of the women came from these societies, one way or another.

    This process continued until the genome-wide makeup of the steppe groups was ~50% Georgian-like, and it seems that at this point the population density on the steppe was high enough so that it was no longer necessary to import women from the edges of the steppe.
     
    if the process was ongoing and recurrent, then we should expect high prevalence of very long blocks of "Georgian" DNA throughout the epoch, while HG haploblocks will get progressively shorter and shorter over time. Perhaps it could be tested even in low-coverage sequencing results?

    (And perhaps compared with the Turkmen tribal patterns where wive-grabbing from the high-population density was the norm, with the Westernmost tribes reportedly raiding Farsi oases and the Easternmost tribes raiding their Turkmen neighbors, until a strong East-West phenotypical cline has become established?)

    (But I wouldn't be surprised if Khvalynsk was simply too far from any high-density agricultural settlements for the continuous bride-raiding to take place ... it could have been a discontinuous, now and then again, affair...)
  60. @Karl Zimmerman
    I saw this paper just came out. I'm sure Razib will have a more detailed post regarding it soon. Since it deals with Africa rather than Europe, I won't discuss it here in detail.

    I can't read the paper due to a paywall, but the interesting thing the supplementary materials note is that the West Eurasian admixture into Sub-Saharan Africans (which is shared to a limited degree by all African populations, even Pygmies and West Africans) is a better fit if modeled as Sardinian or even LBK rather than as the (non-SSA admixed) portion of the Druze genome.

    It seems that either the first "push" of farmers into Africa were not Afro-Asiatic (instead being part of the same group as "Old Europe" whose languages were later erased) or that the genomes of contemporary Semites are not similar to those who spoke Proto-Afro Asiatic at all. The latter is most intriguing - what the heck are modern Semites (particularly Arabs) in this case?

    R1b-V88? Doesn’t fit the Afro-Asiatic bill. Were any Chadic-speaking populations tested here?

    Read More
    • Replies: @Karl Zimmerman
    I can't read the article because it's behind a paywall, but judging from the supplementary material, they aren't really claiming the 4,500 year old remains are of an Afro-Asiatic group, only that they show a close genetic linkage to modern Omotic speakers. They actually seem to be claiming that the remains are of a hunter gatherer who predated the expansion of Neolithic farmers into Sub-Saharan Africa (since he shows no West Eurasian admixture). They note the next closest relations are the Nilo-Saharan Gumuz (who they wrongly claim are Afro-Asiatic) and the Sandawe.

    I think it's fair to presume the remains are of a non Afro-Asiatic population, some sort of "Basal East African" group. I think they either got the dating wrong or this is a relic, pygmy-like population though. Even if you throw out the linguistic evidence suggesting Afro-Asiatic has far deeper roots than that within Africa, my understanding is there is plentiful archaeological evidence of agriculture within Sub-Saharan Africa prior to 2,500 BC.

    Oh, and they did not sample any Chadic-speaking populations.
  61. I believe that Indo-European and Afro-Asiatic may both be language families of hunter-gatherer populations who absorbed migrating farmers into a radically different ecology where the competitive playing field was much more level than in the initial zone of expansion.

    I’ve been thinking about this for some time and am struck by what seem to be some interesting parallels in the origins and spread of the Indo-Europeans and Afro-Asiatic language families.

    Assembling the information that I’ve been able to find, it seems that Northern Afro-Asiatic languages have their origins in Southern Egypt / Northern Sudan where a farmer population with its main origins in the Levant encountered a hunter gatherer Afro-Asiatic speaking population in an environment that was not suitable for settled farming.

    For this scenario to work the forbears of the Nubian speakers, who later occupied this region, would have had to be living elsewhere at the time, maybe around the Yellow Nile area to the west.

    Like the Proto Indo Europeans with the horse, the Northern Afro-Asiatics had their own special domesticated animal, the donkey. Donkeys were tremendously useful to the Neolithic farmers but also facilitated a whole new pastoralist way of life in the semi-arid areas adjacent to the lands suitable for permanent farming. With the domesticated browsing animals of the farmer ancestors and the donkey as a beast of burden, pastoralist Afro-Asiatics were able to have their whole community on the move living in tents which were carried around by their donkeys.

    I imagine that the original Northern Afro Asiatic population, like the Proto Indo Europeans was about a half and half mixture between the Neolithic farmers and the sub Saharan hunter gatherers (who give the new population their language) but as this population expanded northwards down the Nile valley, the sub Saharan part of their ancestry got diluted as they mixed in with the pre-existing Neolithic farmers.

    Northern Afro Asiatic was the parent language to the Berber languages, to the Egyptian language and to the Semitic languages. I imagine that Proto Semitic has its origins in a pastoralist population located in the Southern Levant which had stemmed from an Afro-Asiatic blended population from Lower Egypt.

    This scenario is supported by the Admixture analysis that appears in Haak. At K=18, the Bedouin of the Negev desert are assigned their own new purple coloured cluster at almost 100%. At the same time this component shows up in varying degrees in all of the Semitic language speaking populations. In thinking about what the origins of this new population was, the analysis shows that at K=17 the Negev Bedouin are represented by roughly 65% Neolithic Levantine farmer (brownish yellow), 25% Highland Middle East farmer (teal) and 10% by the pink factor which is significant in many populations living close the Nile such as the Dinka and other populations more widely in East Africa such as the Oromo and the Somalis. The Bedouin may be one of those “fossil” populations which even today survive more or less unchanged genetically from what they were at their origin.

    The physical environment of the Fertile Crescent was perfect for the expansion and spread of the Proto Semitic pastoralists as there are vast tracts of semi-arid land which were open for their exploitation. Mobile and living next to an array of settled farming peoples, the Proto Semites were able to become influential as traders and exchangers of information from the Levant to Mesopotamia.

    I can’t read the paper due to a paywall, but the interesting thing the supplementary materials note is that the West Eurasian admixture into Sub-Saharan Africans (which is shared to a limited degree by all African populations, even Pygmies and West Africans) is a better fit if modeled as Sardinian or even LBK rather than as the (non-SSA admixed) portion of the Druze genome.

    It seems to me that this information is consistent with the scenario I have suggested. As cultural innovators, one would expect the Northern Afro Asiatics to expand in all directions, to the south as well as to the north. Moving southwards the population introduced those original Levantine farmer genes to other parts of Africa.

    Read More
  62. @Davidski
    It wasn't a wholesale change in mtDNA, because Yamnaya still show a decent percentage of mtDNA U lineages which are likely to be of hunter-gatherer origin.

    But what appears to have happened is that around 4,000 BC during the Khvalynsk phase, highly patriarchal, patrilocal, and territorial bands of young males of steppe hunter-gatherer ancestry got their women from wherever they could, and probably because more southerly populations with a longer history of agriculture and pastoralism had bigger populations, most of the women came from these societies, one way or another.

    This process continued until the genome-wide makeup of the steppe groups was ~50% Georgian-like, and it seems that at this point the population density on the steppe was high enough so that it was no longer necessary to import women from the edges of the steppe.

    Then the steppe started to dry up rapidly from 2,600 BC and a lot of the bands moved west into Europe, and took their languages with them.

    Thanks, cool!

    around 4,000 BC during the Khvalynsk phase, highly patriarchal, patrilocal, and territorial bands of young males of steppe hunter-gatherer ancestry got their women from wherever they could, and probably because more southerly populations with a longer history of agriculture and pastoralism had bigger populations, most of the women came from these societies, one way or another.

    This process continued until the genome-wide makeup of the steppe groups was ~50% Georgian-like, and it seems that at this point the population density on the steppe was high enough so that it was no longer necessary to import women from the edges of the steppe.

    if the process was ongoing and recurrent, then we should expect high prevalence of very long blocks of “Georgian” DNA throughout the epoch, while HG haploblocks will get progressively shorter and shorter over time. Perhaps it could be tested even in low-coverage sequencing results?

    (And perhaps compared with the Turkmen tribal patterns where wive-grabbing from the high-population density was the norm, with the Westernmost tribes reportedly raiding Farsi oases and the Easternmost tribes raiding their Turkmen neighbors, until a strong East-West phenotypical cline has become established?)

    (But I wouldn’t be surprised if Khvalynsk was simply too far from any high-density agricultural settlements for the continuous bride-raiding to take place … it could have been a discontinuous, now and then again, affair…)

    Read More
  63. I imagine that the original Northern Afro Asiatic population, like the Proto Indo Europeans was about a half and half mixture between the Neolithic farmers and the sub Saharan hunter gatherers…

    Sorry that should read instead “I imagine that the original Northern Afro Asiatic population, like the Proto Indo Europeans was about a half and half mixture between the farmers and the hunter gatherers. For the Northern Afro Asiatics, the mixture was between Neolithic farmers of Levantine origin and sub Saharan hunter gatherers.”

    Read More
  64. @Josh Lipson
    R1b-V88? Doesn't fit the Afro-Asiatic bill. Were any Chadic-speaking populations tested here?

    I can’t read the article because it’s behind a paywall, but judging from the supplementary material, they aren’t really claiming the 4,500 year old remains are of an Afro-Asiatic group, only that they show a close genetic linkage to modern Omotic speakers. They actually seem to be claiming that the remains are of a hunter gatherer who predated the expansion of Neolithic farmers into Sub-Saharan Africa (since he shows no West Eurasian admixture). They note the next closest relations are the Nilo-Saharan Gumuz (who they wrongly claim are Afro-Asiatic) and the Sandawe.

    I think it’s fair to presume the remains are of a non Afro-Asiatic population, some sort of “Basal East African” group. I think they either got the dating wrong or this is a relic, pygmy-like population though. Even if you throw out the linguistic evidence suggesting Afro-Asiatic has far deeper roots than that within Africa, my understanding is there is plentiful archaeological evidence of agriculture within Sub-Saharan Africa prior to 2,500 BC.

    Oh, and they did not sample any Chadic-speaking populations.

    Read More
    • Replies: @ohwilleke
    The key conclusions of this ancient DNA find in Ethiopia from 2500 BCE are that:

    1. This individual did not have Holocene era Eurasian admixture which is present in all modern African populations (even pygmies and Khoi-San) to the tune of at least 6%.

    2. The Holocene era Eurasian admixture in all sub-Saharan populations studies (which do not include Chadic) is basically LBK/Sardinian.

    3. The Holocene era Eurasian admixture in sub-Saharan Africa probably happened mostly after 2500 BCE (which is consistent with prior studies).

    4. The Sandawe and Omotic Ari people form a tight cluster that this individual shares which is midway between other Afro-Asiatic populations (Cushitic and Ethio-Semitic) in one direction, Nilotic populations in another direction, and the Hadza population in another direction. The Omotic Ari people still live in the region, but the Sandawe live in Tanzania are are less closely related by other tests. The modern Omotic Ari people are about 15% Eurasian admixed relative to the ancient DNA sample.

    5. The ancient DNA sample has zero Neanderthal DNA while other Africans have Neanderthal DNA proportionate to their Neolithic Eurasian ancestry.

    6. The individual is mtDNA L3x2a and Y-DNA E1b1 (E-P2) both of which are highly typical of the area where he was buried today. He is almost surely ancestral to the modern local Ari people, although that doesn't allow one to determine his linguistic affiliation reliably. This also tends to disfavor a Eurasian origin for Y-DNA E.

    7. Given the context of his burial (stone blades, no metal, no signs of farming or herding) and the timing, it is likely that he was a hunter-gatherer, as the Sandawe were until recently to the South. But, given his mtDNA and Y-DNA, the population from which he derived was not particular basal within Africa in the way that the Pygmies and Khoisan are based on their uniparental genetics. He was an ordinary sub-Saharan African genetically, not a Paleo-African.

    A 2014 study estimated that Eurasian Khoisan admixture is traceable to migration of an Omotic-like population that admixed with Eurasians ca. 1900 BCE. The Omotic people were some of the first farmers of Ethiopia (before most Cushitic peoples) and used a locally developed traditional method until Ethio-Semitic methods arrived ca. 1500 BCE.

    Cattle reached Egypt in the earliest part of the Neolithic revolution in Africa around 7000 BCE, the donkey was locally domesticated around 6000 BCE, and sheep and goats appear around 5000 BCE. Fertile Crescent crops were mostly unsuited to sub-Saharan climates, so farming came much later to African than herding. The domestication of sorghum has its origins in Ethiopia and surrounding countries, commencing around 4000–3000 BC. Pearl millet cultivation became ca. 3200-2700 BCE in Africa and started in the West with a transfer to the East and to India by 1700 BCE. The Sahel farming package combined West African and East African crops into a functional unit happened sometime between 3200 BCE and 1700 BCE.
  65. @Asya Pereltsvaig
    Thanks for the summary of relevant bits of our book. I guess it preempts the comment I've just written, but I only noticed your comment now...

    Asya, I tried summarizing the key objections to Gamkrelidze & Ivanov’s conclusions, but some parts give me more questions then answers. Like you quoted R. Beekes, “Comparative Indo-European Linguistics: An Introduction” (1995) to make the point that Gamkrelidze’s PIE reconstruction of “panther” is wrong. But when I check Beekes, he doesn’t disagree with Gamkrelidze & Ivanov’s reconstruction … Beekes just notes that it may not be conclusively attributed to borrowing from Proto-Semitic.

    So I look what Gamkrelidze and Ivanov had to say, and, lo and behold, they rather unequivocally say that the PIE word for “panther” was NOT from Proto-Semitic *nimr (Fronzarolli 1968). Rather, they show a continuity of the panther cult traditions between the pre-Hittite Anatolia and the Hittite texts (which is not surprising by itself) and non-Anatolian IE languages, to hypothesize that the word *pars/*pard entered IE languages South of the Caucasus with these pre-IE religious traditions. They also hypothesize that non-IE borrowed word ended with a dental fricative in the original source language, which is why some IE languages got is as “s” and others as “d”.

    Now it wouldn’t be unreasonable to argue with either of these finer points (spread of Anatolian pre-IE cults / dental fricative sound), but instead Beekes argues that the word isn’t from Proto-Semitic (which G&I didn’t claim in the first place) and you argue that the PIE reconstruction must be wrong (which Beeker never claimed, in turn). It almost makes me guess that Gamkrelidze’s theories were so annoying and long-winded that the critics never got patience to actually read the stuff?

    Read More
  66. @Asya Pereltsvaig
    Good point about Proto-Kartvelian. I will ask Johanna Nichols a bit more about this when she does a guest lecture in my course on 11/9...

    Good point about Proto-Kartvelian. I will ask Johanna Nichols a bit more

    As I understand, she noted that in historic times, languages kept spreading West across Central Asia, and only very rarely in different directions. She didn’t think that the conditions for the earlier migrations differed all that much [and I'm not going to argue with that, for the sake of brevity]. Ergo, Indo-Europeans, Kartvelians, etc. also must have spread West across Central Asia, in her opinion?

    Read More
    • Replies: @Asya Pereltsvaig
    I don't live in Prof. Nichols' head, so I can't say what she thinks. If you're interested, I can report what she says in her guest lecture.
  67. @Karl Zimmerman
    I can't read the article because it's behind a paywall, but judging from the supplementary material, they aren't really claiming the 4,500 year old remains are of an Afro-Asiatic group, only that they show a close genetic linkage to modern Omotic speakers. They actually seem to be claiming that the remains are of a hunter gatherer who predated the expansion of Neolithic farmers into Sub-Saharan Africa (since he shows no West Eurasian admixture). They note the next closest relations are the Nilo-Saharan Gumuz (who they wrongly claim are Afro-Asiatic) and the Sandawe.

    I think it's fair to presume the remains are of a non Afro-Asiatic population, some sort of "Basal East African" group. I think they either got the dating wrong or this is a relic, pygmy-like population though. Even if you throw out the linguistic evidence suggesting Afro-Asiatic has far deeper roots than that within Africa, my understanding is there is plentiful archaeological evidence of agriculture within Sub-Saharan Africa prior to 2,500 BC.

    Oh, and they did not sample any Chadic-speaking populations.

    The key conclusions of this ancient DNA find in Ethiopia from 2500 BCE are that:

    1. This individual did not have Holocene era Eurasian admixture which is present in all modern African populations (even pygmies and Khoi-San) to the tune of at least 6%.

    2. The Holocene era Eurasian admixture in all sub-Saharan populations studies (which do not include Chadic) is basically LBK/Sardinian.

    3. The Holocene era Eurasian admixture in sub-Saharan Africa probably happened mostly after 2500 BCE (which is consistent with prior studies).

    4. The Sandawe and Omotic Ari people form a tight cluster that this individual shares which is midway between other Afro-Asiatic populations (Cushitic and Ethio-Semitic) in one direction, Nilotic populations in another direction, and the Hadza population in another direction. The Omotic Ari people still live in the region, but the Sandawe live in Tanzania are are less closely related by other tests. The modern Omotic Ari people are about 15% Eurasian admixed relative to the ancient DNA sample.

    5. The ancient DNA sample has zero Neanderthal DNA while other Africans have Neanderthal DNA proportionate to their Neolithic Eurasian ancestry.

    6. The individual is mtDNA L3x2a and Y-DNA E1b1 (E-P2) both of which are highly typical of the area where he was buried today. He is almost surely ancestral to the modern local Ari people, although that doesn’t allow one to determine his linguistic affiliation reliably. This also tends to disfavor a Eurasian origin for Y-DNA E.

    7. Given the context of his burial (stone blades, no metal, no signs of farming or herding) and the timing, it is likely that he was a hunter-gatherer, as the Sandawe were until recently to the South. But, given his mtDNA and Y-DNA, the population from which he derived was not particular basal within Africa in the way that the Pygmies and Khoisan are based on their uniparental genetics. He was an ordinary sub-Saharan African genetically, not a Paleo-African.

    A 2014 study estimated that Eurasian Khoisan admixture is traceable to migration of an Omotic-like population that admixed with Eurasians ca. 1900 BCE. The Omotic people were some of the first farmers of Ethiopia (before most Cushitic peoples) and used a locally developed traditional method until Ethio-Semitic methods arrived ca. 1500 BCE.

    Cattle reached Egypt in the earliest part of the Neolithic revolution in Africa around 7000 BCE, the donkey was locally domesticated around 6000 BCE, and sheep and goats appear around 5000 BCE. Fertile Crescent crops were mostly unsuited to sub-Saharan climates, so farming came much later to African than herding. The domestication of sorghum has its origins in Ethiopia and surrounding countries, commencing around 4000–3000 BC. Pearl millet cultivation became ca. 3200-2700 BCE in Africa and started in the West with a transfer to the East and to India by 1700 BCE. The Sahel farming package combined West African and East African crops into a functional unit happened sometime between 3200 BCE and 1700 BCE.

    Read More
    • Replies: @Rick
    "5. The ancient DNA sample has zero Neanderthal DNA while other Africans have Neanderthal DNA proportionate to their Neolithic Eurasian ancestry."

    This is only true because the comparison was done using itself as the reference for 0% Neanderthal. You can not actually say that this person had no Neanderthal component. He could have had quite a bit, in the case that you have no actual 0% reference.

    "6. The individual is mtDNA L3x2a and Y-DNA E1b1 (E-P2) both of which are highly typical of the area where he was buried today. This also tends to disfavor a Eurasian origin for Y-DNA E."

    It actually only disfavors a more modern introduction into Africa. It could have an origin in Eurasia and arrived with cattle farmers in 7000 BCE. It could also have arrived much much earlier than this. In that case, then the baseline for Eurasian and Neanderthal admixture in Africa will only be found by looking at genomes several tens of thousands of years old.

    It could actually be that populations of humans in Sub-Saharan-Africa were mostly very divergent from each other, like the Pygmies and Khoisan still mostly are, and what you called "Paleo-African". Several early waves of migration from North-Africa and Eurasia would have made this less true today, and result in most Africans today being what you called, "ordinary sub-Saharan Africans".
    , @Karl Zimmerman
    I think your numbered points are not really controversial at all. However, your last paragraph gets to the crux of the issue for me. We know that animal husbandry and even agriculture predated 2,500 BC in parts of the Ethiopian Highlands and the Sahel by hundreds to thousands of years. It's also generally believed that proto-Afro-Asiatic was spoken by groups which ate wild cereals but didn't farm, as the different branches share cognates for crops, but not for cultivation words, with a suggested origin of around 8,000 BC. So the archaeological and genetic information suggests a deeply rooted Afro-Asiatic, and longer period of agriculture in some parts of Africa, while the genetics suggest a single admixture event which took place at roughly the same time as the Bronze Age in the Near East.

    One other thing to note - given all modern African groups show some Eurasian admixture, it's now feasible to suggest that Niger-Kordofanian and Nilo-Saharan languages may also have their roots in the Near East with the First Farmers. Certainly both are thought to have expanded only with the development of Sub-Saharan African agriculture, and if some infusion of Eurasian genes came along with this across all African groups, they might have brought their languages via elite dominance. To the best of my knowledge, no one but crazy Afrocentrists have tried to find links between the African languages and extinct Eurasian isolates (and modern families), so this may be fertile ground for linguists.

  68. I’ll have to look at this new paper in full but what is the method used to find the Eurasian signal / proportions? F4 estimation?

    It seems strange that, as far as I remember, d(chimpanzee outgroup, african;Sardinian,east Asian) stats are not statistically different from zero, if many Africans have LBK related ancestry in substantial proportion.

    Read More
    • Replies: @Rick
    Yes, the Eurasian signal was by F4 estimation.

    And yes, I think it is true that Dstats of d(chimpanzee outgroup, african;Sardinian,east Asian) are actually significant. This previously led some people to think that the 'basal eurasian' component seen in Sardinians was actually admixture from Africa into early farmers.
    , @Tobus
    Dstats do show a small, but significant, African/Euro affinity:

    Chimp Yoruba French Han -0.0053 -2.808
    Chimp Yoruba French Dai -0.0051 -2.628
    Chimp Yoruba Sardinian Han -0.0054 -2.786
    Chimp Yoruba Sardinian Dai -0.0053 -2.657
    Chimp Mandenka French Han -0.007 -3.656
    Chimp Mandenka French Dai -0.0069 -3.518
    Chimp Mandenka Sardinian Han -0.0074 -3.775
    Chimp Mandenka Sardinian Dai -0.0073 -3.678
  69. @Matt_
    I'll have to look at this new paper in full but what is the method used to find the Eurasian signal / proportions? F4 estimation?

    It seems strange that, as far as I remember, d(chimpanzee outgroup, african;Sardinian,east Asian) stats are not statistically different from zero, if many Africans have LBK related ancestry in substantial proportion.

    Yes, the Eurasian signal was by F4 estimation.

    And yes, I think it is true that Dstats of d(chimpanzee outgroup, african;Sardinian,east Asian) are actually significant. This previously led some people to think that the ‘basal eurasian’ component seen in Sardinians was actually admixture from Africa into early farmers.

    Read More
  70. @ohwilleke
    The key conclusions of this ancient DNA find in Ethiopia from 2500 BCE are that:

    1. This individual did not have Holocene era Eurasian admixture which is present in all modern African populations (even pygmies and Khoi-San) to the tune of at least 6%.

    2. The Holocene era Eurasian admixture in all sub-Saharan populations studies (which do not include Chadic) is basically LBK/Sardinian.

    3. The Holocene era Eurasian admixture in sub-Saharan Africa probably happened mostly after 2500 BCE (which is consistent with prior studies).

    4. The Sandawe and Omotic Ari people form a tight cluster that this individual shares which is midway between other Afro-Asiatic populations (Cushitic and Ethio-Semitic) in one direction, Nilotic populations in another direction, and the Hadza population in another direction. The Omotic Ari people still live in the region, but the Sandawe live in Tanzania are are less closely related by other tests. The modern Omotic Ari people are about 15% Eurasian admixed relative to the ancient DNA sample.

    5. The ancient DNA sample has zero Neanderthal DNA while other Africans have Neanderthal DNA proportionate to their Neolithic Eurasian ancestry.

    6. The individual is mtDNA L3x2a and Y-DNA E1b1 (E-P2) both of which are highly typical of the area where he was buried today. He is almost surely ancestral to the modern local Ari people, although that doesn't allow one to determine his linguistic affiliation reliably. This also tends to disfavor a Eurasian origin for Y-DNA E.

    7. Given the context of his burial (stone blades, no metal, no signs of farming or herding) and the timing, it is likely that he was a hunter-gatherer, as the Sandawe were until recently to the South. But, given his mtDNA and Y-DNA, the population from which he derived was not particular basal within Africa in the way that the Pygmies and Khoisan are based on their uniparental genetics. He was an ordinary sub-Saharan African genetically, not a Paleo-African.

    A 2014 study estimated that Eurasian Khoisan admixture is traceable to migration of an Omotic-like population that admixed with Eurasians ca. 1900 BCE. The Omotic people were some of the first farmers of Ethiopia (before most Cushitic peoples) and used a locally developed traditional method until Ethio-Semitic methods arrived ca. 1500 BCE.

    Cattle reached Egypt in the earliest part of the Neolithic revolution in Africa around 7000 BCE, the donkey was locally domesticated around 6000 BCE, and sheep and goats appear around 5000 BCE. Fertile Crescent crops were mostly unsuited to sub-Saharan climates, so farming came much later to African than herding. The domestication of sorghum has its origins in Ethiopia and surrounding countries, commencing around 4000–3000 BC. Pearl millet cultivation became ca. 3200-2700 BCE in Africa and started in the West with a transfer to the East and to India by 1700 BCE. The Sahel farming package combined West African and East African crops into a functional unit happened sometime between 3200 BCE and 1700 BCE.

    “5. The ancient DNA sample has zero Neanderthal DNA while other Africans have Neanderthal DNA proportionate to their Neolithic Eurasian ancestry.”

    This is only true because the comparison was done using itself as the reference for 0% Neanderthal. You can not actually say that this person had no Neanderthal component. He could have had quite a bit, in the case that you have no actual 0% reference.

    “6. The individual is mtDNA L3x2a and Y-DNA E1b1 (E-P2) both of which are highly typical of the area where he was buried today. This also tends to disfavor a Eurasian origin for Y-DNA E.”

    It actually only disfavors a more modern introduction into Africa. It could have an origin in Eurasia and arrived with cattle farmers in 7000 BCE. It could also have arrived much much earlier than this. In that case, then the baseline for Eurasian and Neanderthal admixture in Africa will only be found by looking at genomes several tens of thousands of years old.

    It could actually be that populations of humans in Sub-Saharan-Africa were mostly very divergent from each other, like the Pygmies and Khoisan still mostly are, and what you called “Paleo-African”. Several early waves of migration from North-Africa and Eurasia would have made this less true today, and result in most Africans today being what you called, “ordinary sub-Saharan Africans”.

    Read More
  71. Liorente et. al. in Science suggest that this spread also reaches into Africa!

    “Characterizing genetic diversity in Africa is a crucial step for most analyses reconstructing the evolutionary history of anatomically modern humans. However, historic migrations from Eurasia into Africa have affected many contemporary populations, confounding inferences. Here, we present a 12.5x coverage ancient genome of an Ethiopian male (‘Mota’) who lived approximately 4,500 years ago. We use this genome to demonstrate that the Eurasian backflow into Africa came from a population closely related to Early Neolithic farmers, who had colonized Europe 4,000 years earlier. The extent of this backflow was much greater than previously reported, reaching all the way to Central, West and Southern Africa, affecting even populations such as Yoruba and Mbuti, previously thought to be relatively unadmixed, who harbor 6-7% Eurasian ancestry.”

    Read More
  72. @ohwilleke
    The key conclusions of this ancient DNA find in Ethiopia from 2500 BCE are that:

    1. This individual did not have Holocene era Eurasian admixture which is present in all modern African populations (even pygmies and Khoi-San) to the tune of at least 6%.

    2. The Holocene era Eurasian admixture in all sub-Saharan populations studies (which do not include Chadic) is basically LBK/Sardinian.

    3. The Holocene era Eurasian admixture in sub-Saharan Africa probably happened mostly after 2500 BCE (which is consistent with prior studies).

    4. The Sandawe and Omotic Ari people form a tight cluster that this individual shares which is midway between other Afro-Asiatic populations (Cushitic and Ethio-Semitic) in one direction, Nilotic populations in another direction, and the Hadza population in another direction. The Omotic Ari people still live in the region, but the Sandawe live in Tanzania are are less closely related by other tests. The modern Omotic Ari people are about 15% Eurasian admixed relative to the ancient DNA sample.

    5. The ancient DNA sample has zero Neanderthal DNA while other Africans have Neanderthal DNA proportionate to their Neolithic Eurasian ancestry.

    6. The individual is mtDNA L3x2a and Y-DNA E1b1 (E-P2) both of which are highly typical of the area where he was buried today. He is almost surely ancestral to the modern local Ari people, although that doesn't allow one to determine his linguistic affiliation reliably. This also tends to disfavor a Eurasian origin for Y-DNA E.

    7. Given the context of his burial (stone blades, no metal, no signs of farming or herding) and the timing, it is likely that he was a hunter-gatherer, as the Sandawe were until recently to the South. But, given his mtDNA and Y-DNA, the population from which he derived was not particular basal within Africa in the way that the Pygmies and Khoisan are based on their uniparental genetics. He was an ordinary sub-Saharan African genetically, not a Paleo-African.

    A 2014 study estimated that Eurasian Khoisan admixture is traceable to migration of an Omotic-like population that admixed with Eurasians ca. 1900 BCE. The Omotic people were some of the first farmers of Ethiopia (before most Cushitic peoples) and used a locally developed traditional method until Ethio-Semitic methods arrived ca. 1500 BCE.

    Cattle reached Egypt in the earliest part of the Neolithic revolution in Africa around 7000 BCE, the donkey was locally domesticated around 6000 BCE, and sheep and goats appear around 5000 BCE. Fertile Crescent crops were mostly unsuited to sub-Saharan climates, so farming came much later to African than herding. The domestication of sorghum has its origins in Ethiopia and surrounding countries, commencing around 4000–3000 BC. Pearl millet cultivation became ca. 3200-2700 BCE in Africa and started in the West with a transfer to the East and to India by 1700 BCE. The Sahel farming package combined West African and East African crops into a functional unit happened sometime between 3200 BCE and 1700 BCE.

    I think your numbered points are not really controversial at all. However, your last paragraph gets to the crux of the issue for me. We know that animal husbandry and even agriculture predated 2,500 BC in parts of the Ethiopian Highlands and the Sahel by hundreds to thousands of years. It’s also generally believed that proto-Afro-Asiatic was spoken by groups which ate wild cereals but didn’t farm, as the different branches share cognates for crops, but not for cultivation words, with a suggested origin of around 8,000 BC. So the archaeological and genetic information suggests a deeply rooted Afro-Asiatic, and longer period of agriculture in some parts of Africa, while the genetics suggest a single admixture event which took place at roughly the same time as the Bronze Age in the Near East.

    One other thing to note – given all modern African groups show some Eurasian admixture, it’s now feasible to suggest that Niger-Kordofanian and Nilo-Saharan languages may also have their roots in the Near East with the First Farmers. Certainly both are thought to have expanded only with the development of Sub-Saharan African agriculture, and if some infusion of Eurasian genes came along with this across all African groups, they might have brought their languages via elite dominance. To the best of my knowledge, no one but crazy Afrocentrists have tried to find links between the African languages and extinct Eurasian isolates (and modern families), so this may be fertile ground for linguists.

    Read More
    • Replies: @ohwilleke
    I wouldn't call myself a "crazy Afrocentrist", but I have explored a couple of links between African languages and extinct Eurasian isolated (and modern families).

    1. There is a significant similarity between Dravidian languages and some of the languages at the fringe of the Niger-Congo language area, coupled with the fact that the crop package that works in Southern India consists mostly of domesticates whose wild types and domestication happened in the Sahel region of India, some interesting cultural coincidences in terms of games, architecture and metaphysical views of the world between the Sahel and South India, a spike in Y-DNA T frequencies right around the most likely homeland within India of proto-Dravidian, the coincidence of time between the South Indian Neolithic and the emergence of a complete Sahel crop package in Africa, and known maritime transportation between the two regions in the relevant time period from Sumerian records.

    2. Maju has stumbled upon some surprisingly strong correspondences between Nubian Nilo-Saharan languages and Basque, with significant PIE lexical borrowing (Vasco-Nubian).

    3. I have long hypothesized that the ergative languages of Western Eurasia (e.g. Basque, Caucasian languages, Sumerian, Elamite) have some deep level linguistic family relationships. There also appears to be an ergative substrate in NW African Berber languages, and a couple of Nilo-Saharan languages that show ergativity.

    4. I have also explored possible connections between languages traceable through glottal consonants, some of which tend to replicate (3).

    5. I am inclined to think that in instances where we can't be sure, that the major language families of African populations that have herders or farmers now have roots in the Neolithic. But, it is important to note that most of Africa received herding much sooner than they received farming. So, the fact that a proto-language lacks words for cultivation, but has words for crops, could indicate a Neolithic herder origin that traded for the products of domesticated crops but didn't grow them themselves, rather than a Mesolithic community of proto-farmers. Indeed, we now know that Western Europe also received the products of domesticated crops through trade a thousand years or so before the Neolithic revolution reached it.

    This said, there has been considerable improvement in our knowledge of absolute dates and relative time sequences of many African language families (e.g. Berber, Chadic, Semitic, Nilo-Saharan, Cushitic).

  73. One thing it’s important not to forget when drawing arrows directly from thousands of years ago to today is that a lot has happened since. Mortality was not neutral or random in its effect on different genomes, and much of it was in the form of disease.

    Europe had close to 50% replacement at the time of the plague, and there may well have been genetic factors in survival rates. It’s likely that the diseases of civilization similarly worked their horrors on sub-Saharan populations soon after admixture. One can imagine Afro-Asiatic populations stretching back 8,000 years, with centuries or millennia of slowing intensifying horticulture and pastoralism. Then a small admixture with middle eastern populations is contemporary with the intensification of horticulture into agriculture, and the achievement of a coherent cultural package of crops that can be successful in these climate zones, leading too much greater, sustained population density. This would be followed by a few centuries where mortality is defined primarily by the diseases of density, during which the African genetic component of the Afro-Asiatic population is hit much harder because of lower genetic immunity. Genetically, this would look like a single, dramatic admixture giving rise to the post-admixture expansion into today’s Afro-Asiatic populations, but it’s really just because much of the pre-existing genetic diversity was erased by disease.

    Read More
    • Replies: @Karl Zimmerman
    I'm hardly an expert in genetics, but even if you presume that "ancestral neoafricans" had disadvantageous immune systems, due to the effects of recombination the SSA and West Eurasian chunks would become smaller and smaller with each generation post-admixture, meaning progressively less and less SSA genes would be "taken along for the ride" with the disadvantageous alleles. To give an example, Razib has noted that after 1,000 years of admixture, Uighurs' phenotype no longer match their genotype. That is to say, West Eurasian and East Eurasian looking Uighurs have similar admixture, it's just that the former happened to chance upon genes in their highly blended genome for a West Eurasian phenotype (light hair/eyes, lack of EDAR, etc), while the latter did not. The genomes for African-Americans and Latinos are a bit chunkier (since the admixture cannot be much more than half the value of the Uighurs) but it's still highly recombined. Unless there were hundreds of genes affecting immune system performance, I'm not sure the whole genome proportion which was SSA would really matter - particularly if it took a few hundred years for the population density to rise enough for plagues to matter.

    More generally though, there are interesting ways this could be tested as a general hypothesis. For example, extant Mayans and Quecha almost always have some European admixture. Has this proportion been selected for? I also wonder if this happened in Madagascar, with the introduction of malaria causing a decline in Southeast Asian ancestry everywhere but the highlands.
  74. It looks like moderation is taking forever, so I’m going to take my time before commenting on the PIE family structure, on Johanna Nichols’s updated migration-routes theory, and more topics relevant for Kartvelian-like female-driven admixture story. Just because my understanding has a chance to improve before the comments are published :)

    So just a short tidbit for now: vowel phonology hints at PIE – Proto-Kartvelian interaction

    http://www.cranberryletters.com/home/2014/4/4/phonology-in-the-caucasus

    Read More
    • Replies: @Asya Pereltsvaig
    Dmitry, I don't know about you, but I take time off work on Shabbat.
  75. @ryanwc
    One thing it's important not to forget when drawing arrows directly from thousands of years ago to today is that a lot has happened since. Mortality was not neutral or random in its effect on different genomes, and much of it was in the form of disease.

    Europe had close to 50% replacement at the time of the plague, and there may well have been genetic factors in survival rates. It's likely that the diseases of civilization similarly worked their horrors on sub-Saharan populations soon after admixture. One can imagine Afro-Asiatic populations stretching back 8,000 years, with centuries or millennia of slowing intensifying horticulture and pastoralism. Then a small admixture with middle eastern populations is contemporary with the intensification of horticulture into agriculture, and the achievement of a coherent cultural package of crops that can be successful in these climate zones, leading too much greater, sustained population density. This would be followed by a few centuries where mortality is defined primarily by the diseases of density, during which the African genetic component of the Afro-Asiatic population is hit much harder because of lower genetic immunity. Genetically, this would look like a single, dramatic admixture giving rise to the post-admixture expansion into today's Afro-Asiatic populations, but it's really just because much of the pre-existing genetic diversity was erased by disease.

    I’m hardly an expert in genetics, but even if you presume that “ancestral neoafricans” had disadvantageous immune systems, due to the effects of recombination the SSA and West Eurasian chunks would become smaller and smaller with each generation post-admixture, meaning progressively less and less SSA genes would be “taken along for the ride” with the disadvantageous alleles. To give an example, Razib has noted that after 1,000 years of admixture, Uighurs’ phenotype no longer match their genotype. That is to say, West Eurasian and East Eurasian looking Uighurs have similar admixture, it’s just that the former happened to chance upon genes in their highly blended genome for a West Eurasian phenotype (light hair/eyes, lack of EDAR, etc), while the latter did not. The genomes for African-Americans and Latinos are a bit chunkier (since the admixture cannot be much more than half the value of the Uighurs) but it’s still highly recombined. Unless there were hundreds of genes affecting immune system performance, I’m not sure the whole genome proportion which was SSA would really matter – particularly if it took a few hundred years for the population density to rise enough for plagues to matter.

    More generally though, there are interesting ways this could be tested as a general hypothesis. For example, extant Mayans and Quecha almost always have some European admixture. Has this proportion been selected for? I also wonder if this happened in Madagascar, with the introduction of malaria causing a decline in Southeast Asian ancestry everywhere but the highlands.

    Read More
    • Replies: @ryanwc
    I was presuming the opposite, that the Levantine population would have had significant genetic immune system advantages, and I certainly get that the ibd chunks would get smaller and smaller, especially by 1,000 years. But the disease effects may have been immediate and dramatic over the first 8-10 generations. It took very little admixture to utterly muddy the genetics of Cherokees, in part because of the toll taken by Old World diseases.

    Native American languages are mostly lost now. Even Cherokee may not survive. But if similar population processes had worked themselves out in the absence of modern transportation and communications networks, with the balance tipped just a tiny bit more in favor of natives, one could easily imagine Cherokee and the 5 Iroquois languages surviving, and one day, someone would find it bizarre that the genetics of Iroquoian populations go back to a moment of Euro-native admixture in the late 1600's/1700's, even though the linguists claim the Iroquoian languages have a much older history.

    I was just trying to explain how the linguistic and archaeological evidence of Afro-Asiatic might say something different from the genetics, without actually contradicting it. Both may be correct.
  76. @Karl Zimmerman
    I think your numbered points are not really controversial at all. However, your last paragraph gets to the crux of the issue for me. We know that animal husbandry and even agriculture predated 2,500 BC in parts of the Ethiopian Highlands and the Sahel by hundreds to thousands of years. It's also generally believed that proto-Afro-Asiatic was spoken by groups which ate wild cereals but didn't farm, as the different branches share cognates for crops, but not for cultivation words, with a suggested origin of around 8,000 BC. So the archaeological and genetic information suggests a deeply rooted Afro-Asiatic, and longer period of agriculture in some parts of Africa, while the genetics suggest a single admixture event which took place at roughly the same time as the Bronze Age in the Near East.

    One other thing to note - given all modern African groups show some Eurasian admixture, it's now feasible to suggest that Niger-Kordofanian and Nilo-Saharan languages may also have their roots in the Near East with the First Farmers. Certainly both are thought to have expanded only with the development of Sub-Saharan African agriculture, and if some infusion of Eurasian genes came along with this across all African groups, they might have brought their languages via elite dominance. To the best of my knowledge, no one but crazy Afrocentrists have tried to find links between the African languages and extinct Eurasian isolates (and modern families), so this may be fertile ground for linguists.

    I wouldn’t call myself a “crazy Afrocentrist”, but I have explored a couple of links between African languages and extinct Eurasian isolated (and modern families).

    1. There is a significant similarity between Dravidian languages and some of the languages at the fringe of the Niger-Congo language area, coupled with the fact that the crop package that works in Southern India consists mostly of domesticates whose wild types and domestication happened in the Sahel region of India, some interesting cultural coincidences in terms of games, architecture and metaphysical views of the world between the Sahel and South India, a spike in Y-DNA T frequencies right around the most likely homeland within India of proto-Dravidian, the coincidence of time between the South Indian Neolithic and the emergence of a complete Sahel crop package in Africa, and known maritime transportation between the two regions in the relevant time period from Sumerian records.

    2. Maju has stumbled upon some surprisingly strong correspondences between Nubian Nilo-Saharan languages and Basque, with significant PIE lexical borrowing (Vasco-Nubian).

    3. I have long hypothesized that the ergative languages of Western Eurasia (e.g. Basque, Caucasian languages, Sumerian, Elamite) have some deep level linguistic family relationships. There also appears to be an ergative substrate in NW African Berber languages, and a couple of Nilo-Saharan languages that show ergativity.

    4. I have also explored possible connections between languages traceable through glottal consonants, some of which tend to replicate (3).

    5. I am inclined to think that in instances where we can’t be sure, that the major language families of African populations that have herders or farmers now have roots in the Neolithic. But, it is important to note that most of Africa received herding much sooner than they received farming. So, the fact that a proto-language lacks words for cultivation, but has words for crops, could indicate a Neolithic herder origin that traded for the products of domesticated crops but didn’t grow them themselves, rather than a Mesolithic community of proto-farmers. Indeed, we now know that Western Europe also received the products of domesticated crops through trade a thousand years or so before the Neolithic revolution reached it.

    This said, there has been considerable improvement in our knowledge of absolute dates and relative time sequences of many African language families (e.g. Berber, Chadic, Semitic, Nilo-Saharan, Cushitic).

    Read More
    • Replies: @greysquirrell
    I have to disagree on the similarities between Dravidian and any African language. Having Tamil as my mother tongue I've been reading the theories of Linguists regarding Dravidian languages ; I've not seen any mention from reputable academic sources that Dravidian was related to any language outside the subcontinent . McAlpin was the only one to suggest (Elamo-Dravidian) but his hypothesis has not been accepted as theory.

    There is also no genetic commonality between Southern India and Africa.
    , @LevantineJew


    3. I have long hypothesized that the ergative languages of Western Eurasia (e.g. Basque, Caucasian languages, Sumerian, Elamite) have some deep level linguistic family relationships. There also appears to be an ergative substrate in NW African Berber languages, and a couple of Nilo-Saharan languages that show ergativity.

     

    you might be onto something here.

    It looks that Hurro-Urartian languages were relexified into some North-Western Iranian Kurdish languages, like Zazaki, Gorani, Sorani, Kurmanji and Elamite into some South-Western Iranian languages like Davani [1]


    [1] https://en.wikipedia.org/wiki/Ergative%E2%80%93absolutive_language#Distribution_of_ergative_languages
  77. I believe that Indo-European and Afro-Asiatic may both be language families of hunter-gatherer populations who absorbed migrating farmers into a radically different ecology where the competitive playing field was much more level than in the initial zone of expansion.

    The above quote and the follow-up comments by posters has me confused.

    Is the above quote suggesting that Mesolithic Hunter Gatherers of Europe spoke languages that eventually gave rise to PIE but Neolithic Anatolian farmers spread into Europe and demographically dominated the regions they were present in . Thousands of years later Indo-European pastolralists from the Steppes north of the Caspian Sea overan Western and Eastern Europe replacing the language of the Anatolian farmers who had earlier replaced the languages (a proto PIE if you will) of the Neolithic hunter gatherers ??

    Read More
    • Replies: @Karl Zimmerman
    Proto-Indo-Europeans are thought to have been a roughly 50/50 hybrid between something from the Caucasus and Eastern European Hunter Gatherers (EHG) - with the latter supplying the patrilineage. But EHG is not the same thing as Western Hunter Gatherers, with the main difference being a much higher enrichment of Ancient North Eurasian. ANA is undetectable in most WHG, aside from Scandinavians, which had proportions coincidentally similar to modern Northern Europeans (around 10%). EHG had something in the range of 40% ANA admixture. Whether this was because they were a separate "race" at the time, or because they formed through a hybridization of WHG and ANA, is currently up in the air. It could even be that WHG was actually formed through hybridization of EHG and something else entirely.

    Regardless, they were not the same population during the time window examined, and there's no reason to think they spoke languages of the same family. There's arguments there is a pre-Uralic substrate in the Saami language, which may be the only relic of whatever Mesolithic European language family their was. But given the relative closeness geographically to the Karelian EHG samples, this may just be another EHG language family.
  78. @ohwilleke
    I wouldn't call myself a "crazy Afrocentrist", but I have explored a couple of links between African languages and extinct Eurasian isolated (and modern families).

    1. There is a significant similarity between Dravidian languages and some of the languages at the fringe of the Niger-Congo language area, coupled with the fact that the crop package that works in Southern India consists mostly of domesticates whose wild types and domestication happened in the Sahel region of India, some interesting cultural coincidences in terms of games, architecture and metaphysical views of the world between the Sahel and South India, a spike in Y-DNA T frequencies right around the most likely homeland within India of proto-Dravidian, the coincidence of time between the South Indian Neolithic and the emergence of a complete Sahel crop package in Africa, and known maritime transportation between the two regions in the relevant time period from Sumerian records.

    2. Maju has stumbled upon some surprisingly strong correspondences between Nubian Nilo-Saharan languages and Basque, with significant PIE lexical borrowing (Vasco-Nubian).

    3. I have long hypothesized that the ergative languages of Western Eurasia (e.g. Basque, Caucasian languages, Sumerian, Elamite) have some deep level linguistic family relationships. There also appears to be an ergative substrate in NW African Berber languages, and a couple of Nilo-Saharan languages that show ergativity.

    4. I have also explored possible connections between languages traceable through glottal consonants, some of which tend to replicate (3).

    5. I am inclined to think that in instances where we can't be sure, that the major language families of African populations that have herders or farmers now have roots in the Neolithic. But, it is important to note that most of Africa received herding much sooner than they received farming. So, the fact that a proto-language lacks words for cultivation, but has words for crops, could indicate a Neolithic herder origin that traded for the products of domesticated crops but didn't grow them themselves, rather than a Mesolithic community of proto-farmers. Indeed, we now know that Western Europe also received the products of domesticated crops through trade a thousand years or so before the Neolithic revolution reached it.

    This said, there has been considerable improvement in our knowledge of absolute dates and relative time sequences of many African language families (e.g. Berber, Chadic, Semitic, Nilo-Saharan, Cushitic).

    I have to disagree on the similarities between Dravidian and any African language. Having Tamil as my mother tongue I’ve been reading the theories of Linguists regarding Dravidian languages ; I’ve not seen any mention from reputable academic sources that Dravidian was related to any language outside the subcontinent . McAlpin was the only one to suggest (Elamo-Dravidian) but his hypothesis has not been accepted as theory.

    There is also no genetic commonality between Southern India and Africa.

    Read More
  79. @ohwilleke
    I wouldn't call myself a "crazy Afrocentrist", but I have explored a couple of links between African languages and extinct Eurasian isolated (and modern families).

    1. There is a significant similarity between Dravidian languages and some of the languages at the fringe of the Niger-Congo language area, coupled with the fact that the crop package that works in Southern India consists mostly of domesticates whose wild types and domestication happened in the Sahel region of India, some interesting cultural coincidences in terms of games, architecture and metaphysical views of the world between the Sahel and South India, a spike in Y-DNA T frequencies right around the most likely homeland within India of proto-Dravidian, the coincidence of time between the South Indian Neolithic and the emergence of a complete Sahel crop package in Africa, and known maritime transportation between the two regions in the relevant time period from Sumerian records.

    2. Maju has stumbled upon some surprisingly strong correspondences between Nubian Nilo-Saharan languages and Basque, with significant PIE lexical borrowing (Vasco-Nubian).

    3. I have long hypothesized that the ergative languages of Western Eurasia (e.g. Basque, Caucasian languages, Sumerian, Elamite) have some deep level linguistic family relationships. There also appears to be an ergative substrate in NW African Berber languages, and a couple of Nilo-Saharan languages that show ergativity.

    4. I have also explored possible connections between languages traceable through glottal consonants, some of which tend to replicate (3).

    5. I am inclined to think that in instances where we can't be sure, that the major language families of African populations that have herders or farmers now have roots in the Neolithic. But, it is important to note that most of Africa received herding much sooner than they received farming. So, the fact that a proto-language lacks words for cultivation, but has words for crops, could indicate a Neolithic herder origin that traded for the products of domesticated crops but didn't grow them themselves, rather than a Mesolithic community of proto-farmers. Indeed, we now know that Western Europe also received the products of domesticated crops through trade a thousand years or so before the Neolithic revolution reached it.

    This said, there has been considerable improvement in our knowledge of absolute dates and relative time sequences of many African language families (e.g. Berber, Chadic, Semitic, Nilo-Saharan, Cushitic).

    3. I have long hypothesized that the ergative languages of Western Eurasia (e.g. Basque, Caucasian languages, Sumerian, Elamite) have some deep level linguistic family relationships. There also appears to be an ergative substrate in NW African Berber languages, and a couple of Nilo-Saharan languages that show ergativity.

    you might be onto something here.

    It looks that Hurro-Urartian languages were relexified into some North-Western Iranian Kurdish languages, like Zazaki, Gorani, Sorani, Kurmanji and Elamite into some South-Western Iranian languages like Davani [1]

    [1] https://en.wikipedia.org/wiki/Ergative%E2%80%93absolutive_language#Distribution_of_ergative_languages

    Read More
    • Replies: @Asya Pereltsvaig
    There are also ergative languages in the Americas and Australia -- are they related too? (My point is that ergativity and such properties develop independently and are not necessarily a sign of relatedness)
  80. @Karl Zimmerman
    I'm hardly an expert in genetics, but even if you presume that "ancestral neoafricans" had disadvantageous immune systems, due to the effects of recombination the SSA and West Eurasian chunks would become smaller and smaller with each generation post-admixture, meaning progressively less and less SSA genes would be "taken along for the ride" with the disadvantageous alleles. To give an example, Razib has noted that after 1,000 years of admixture, Uighurs' phenotype no longer match their genotype. That is to say, West Eurasian and East Eurasian looking Uighurs have similar admixture, it's just that the former happened to chance upon genes in their highly blended genome for a West Eurasian phenotype (light hair/eyes, lack of EDAR, etc), while the latter did not. The genomes for African-Americans and Latinos are a bit chunkier (since the admixture cannot be much more than half the value of the Uighurs) but it's still highly recombined. Unless there were hundreds of genes affecting immune system performance, I'm not sure the whole genome proportion which was SSA would really matter - particularly if it took a few hundred years for the population density to rise enough for plagues to matter.

    More generally though, there are interesting ways this could be tested as a general hypothesis. For example, extant Mayans and Quecha almost always have some European admixture. Has this proportion been selected for? I also wonder if this happened in Madagascar, with the introduction of malaria causing a decline in Southeast Asian ancestry everywhere but the highlands.

    I was presuming the opposite, that the Levantine population would have had significant genetic immune system advantages, and I certainly get that the ibd chunks would get smaller and smaller, especially by 1,000 years. But the disease effects may have been immediate and dramatic over the first 8-10 generations. It took very little admixture to utterly muddy the genetics of Cherokees, in part because of the toll taken by Old World diseases.

    Native American languages are mostly lost now. Even Cherokee may not survive. But if similar population processes had worked themselves out in the absence of modern transportation and communications networks, with the balance tipped just a tiny bit more in favor of natives, one could easily imagine Cherokee and the 5 Iroquois languages surviving, and one day, someone would find it bizarre that the genetics of Iroquoian populations go back to a moment of Euro-native admixture in the late 1600′s/1700′s, even though the linguists claim the Iroquoian languages have a much older history.

    I was just trying to explain how the linguistic and archaeological evidence of Afro-Asiatic might say something different from the genetics, without actually contradicting it. Both may be correct.

    Read More
    • Replies: @Karl Zimmerman
    I get your point here, and it is plausible conjecture. I can make a few quibbles with the idea though. First, I think it's wrong to equate the meeting of Europeans and Native Americans to what would have been happening in Africa, as the gap in performance between the first "real farmers" and neighboring herders/agriculturalists would be smaller. Secondly, I think epidemic diseases would need more than a few generations to become endemic. The population would need to become both large enough and cramped enough to serve as a disease reservoir. Also, contagious diseases would need time to spread from the Near East into Africa (from trade routes) or else develop locally (through mutations of diseases already present in local livestock populations). Also, at the same time populations were building up, Malaria would become more of an issue, which would presumably create a strong selection against LBK-like ancestry.

    Oh, and as one aside, you do realize you actually gave Afrocentrists a rationale to explain how ancient Egypt could be "black" while modern Copts are not, right? ;)
  81. @Matt_
    I'll have to look at this new paper in full but what is the method used to find the Eurasian signal / proportions? F4 estimation?

    It seems strange that, as far as I remember, d(chimpanzee outgroup, african;Sardinian,east Asian) stats are not statistically different from zero, if many Africans have LBK related ancestry in substantial proportion.

    Dstats do show a small, but significant, African/Euro affinity:

    Chimp Yoruba French Han -0.0053 -2.808
    Chimp Yoruba French Dai -0.0051 -2.628
    Chimp Yoruba Sardinian Han -0.0054 -2.786
    Chimp Yoruba Sardinian Dai -0.0053 -2.657
    Chimp Mandenka French Han -0.007 -3.656
    Chimp Mandenka French Dai -0.0069 -3.518
    Chimp Mandenka Sardinian Han -0.0074 -3.775
    Chimp Mandenka Sardinian Dai -0.0073 -3.678

    Read More
  82. @greysquirrell

    I believe that Indo-European and Afro-Asiatic may both be language families of hunter-gatherer populations who absorbed migrating farmers into a radically different ecology where the competitive playing field was much more level than in the initial zone of expansion.
     
    The above quote and the follow-up comments by posters has me confused.

    Is the above quote suggesting that Mesolithic Hunter Gatherers of Europe spoke languages that eventually gave rise to PIE but Neolithic Anatolian farmers spread into Europe and demographically dominated the regions they were present in . Thousands of years later Indo-European pastolralists from the Steppes north of the Caspian Sea overan Western and Eastern Europe replacing the language of the Anatolian farmers who had earlier replaced the languages (a proto PIE if you will) of the Neolithic hunter gatherers ??

    Proto-Indo-Europeans are thought to have been a roughly 50/50 hybrid between something from the Caucasus and Eastern European Hunter Gatherers (EHG) – with the latter supplying the patrilineage. But EHG is not the same thing as Western Hunter Gatherers, with the main difference being a much higher enrichment of Ancient North Eurasian. ANA is undetectable in most WHG, aside from Scandinavians, which had proportions coincidentally similar to modern Northern Europeans (around 10%). EHG had something in the range of 40% ANA admixture. Whether this was because they were a separate “race” at the time, or because they formed through a hybridization of WHG and ANA, is currently up in the air. It could even be that WHG was actually formed through hybridization of EHG and something else entirely.

    Regardless, they were not the same population during the time window examined, and there’s no reason to think they spoke languages of the same family. There’s arguments there is a pre-Uralic substrate in the Saami language, which may be the only relic of whatever Mesolithic European language family their was. But given the relative closeness geographically to the Karelian EHG samples, this may just be another EHG language family.

    Read More
    • Replies: @greysquirrell
    My readings on population genetics has lapsed a bit so what is " ANA " ? Thanks.

    Going off on a tangent: Is there any consensus on whether R1b was introduced by Indo-Europeans? Since R1b is also found in some areas of Sub-Saharan Africa and there is no indication it had anything to do with Indo-European speakers I've been skeptical on whether it like R1a introduced IE languages.
  83. @ryanwc
    I was presuming the opposite, that the Levantine population would have had significant genetic immune system advantages, and I certainly get that the ibd chunks would get smaller and smaller, especially by 1,000 years. But the disease effects may have been immediate and dramatic over the first 8-10 generations. It took very little admixture to utterly muddy the genetics of Cherokees, in part because of the toll taken by Old World diseases.

    Native American languages are mostly lost now. Even Cherokee may not survive. But if similar population processes had worked themselves out in the absence of modern transportation and communications networks, with the balance tipped just a tiny bit more in favor of natives, one could easily imagine Cherokee and the 5 Iroquois languages surviving, and one day, someone would find it bizarre that the genetics of Iroquoian populations go back to a moment of Euro-native admixture in the late 1600's/1700's, even though the linguists claim the Iroquoian languages have a much older history.

    I was just trying to explain how the linguistic and archaeological evidence of Afro-Asiatic might say something different from the genetics, without actually contradicting it. Both may be correct.

    I get your point here, and it is plausible conjecture. I can make a few quibbles with the idea though. First, I think it’s wrong to equate the meeting of Europeans and Native Americans to what would have been happening in Africa, as the gap in performance between the first “real farmers” and neighboring herders/agriculturalists would be smaller. Secondly, I think epidemic diseases would need more than a few generations to become endemic. The population would need to become both large enough and cramped enough to serve as a disease reservoir. Also, contagious diseases would need time to spread from the Near East into Africa (from trade routes) or else develop locally (through mutations of diseases already present in local livestock populations). Also, at the same time populations were building up, Malaria would become more of an issue, which would presumably create a strong selection against LBK-like ancestry.

    Oh, and as one aside, you do realize you actually gave Afrocentrists a rationale to explain how ancient Egypt could be “black” while modern Copts are not, right? ;)

    Read More
  84. @Dmitry Pruss

    Good point about Proto-Kartvelian. I will ask Johanna Nichols a bit more
     
    As I understand, she noted that in historic times, languages kept spreading West across Central Asia, and only very rarely in different directions. She didn't think that the conditions for the earlier migrations differed all that much [and I'm not going to argue with that, for the sake of brevity]. Ergo, Indo-Europeans, Kartvelians, etc. also must have spread West across Central Asia, in her opinion?

    I don’t live in Prof. Nichols’ head, so I can’t say what she thinks. If you’re interested, I can report what she says in her guest lecture.

    Read More
  85. @LevantineJew


    3. I have long hypothesized that the ergative languages of Western Eurasia (e.g. Basque, Caucasian languages, Sumerian, Elamite) have some deep level linguistic family relationships. There also appears to be an ergative substrate in NW African Berber languages, and a couple of Nilo-Saharan languages that show ergativity.

     

    you might be onto something here.

    It looks that Hurro-Urartian languages were relexified into some North-Western Iranian Kurdish languages, like Zazaki, Gorani, Sorani, Kurmanji and Elamite into some South-Western Iranian languages like Davani [1]


    [1] https://en.wikipedia.org/wiki/Ergative%E2%80%93absolutive_language#Distribution_of_ergative_languages

    There are also ergative languages in the Americas and Australia — are they related too? (My point is that ergativity and such properties develop independently and are not necessarily a sign of relatedness)

    Read More
  86. @greysquirrell
    I have to disagree on the similarities between Dravidian and any African language. Having Tamil as my mother tongue I've been reading the theories of Linguists regarding Dravidian languages ; I've not seen any mention from reputable academic sources that Dravidian was related to any language outside the subcontinent . McAlpin was the only one to suggest (Elamo-Dravidian) but his hypothesis has not been accepted as theory.

    There is also no genetic commonality between Southern India and Africa.

    Agreed.

    Read More
  87. @Dmitry Pruss
    It looks like moderation is taking forever, so I'm going to take my time before commenting on the PIE family structure, on Johanna Nichols's updated migration-routes theory, and more topics relevant for Kartvelian-like female-driven admixture story. Just because my understanding has a chance to improve before the comments are published :)

    So just a short tidbit for now: vowel phonology hints at PIE - Proto-Kartvelian interaction
    http://www.cranberryletters.com/home/2014/4/4/phonology-in-the-caucasus

    Dmitry, I don’t know about you, but I take time off work on Shabbat.

    Read More
    • Replies: @Dmitry Pruss

    but I take time off work
     
    :) oh, I'm sure it was the weekend stuff which slowed down publication of replies on this site (including yours) - but more specifically the fact that it was the ASHG weekend, with its flow of exciting news (including 7000 years old onset of intermixing of Eastern Hunter-Gatherers with the Caucasus-like agriculturalists, perhaps East of the Caspian Sea; Srubnaya culture being linked to the origin of the Indo-Aryans, and many, many new datapoints from the Steppes) .

    Specifically with respect to Johanna Nichols evolving views on the early Steppe migration routes ... I got my clues from the following recent abstract ( http://conferences.saxo.ku.dk/nordic-tag-2015/documents/N-TAG_session_list_updated__2_.pdf/Final_NTAG_booklet__1_.pdf ), which cites evidence for the crucial role of interrelations between the Steppe people and the hunter-gatherers of the fringe woodland zone for the successful migrations (including the notion of "The Fur Road" along the Northern boundaries of the Steppe, and a "catalytic role" of Proto-Uralic in the early spread of IE languages).

  88. […] It All Began at Ararat, and Esau’s Revenge – “Ancient DNA in Europe strongly indicates massive replacement [of hunter-gatherers by agriculturalists]. But, there is also suggestion of admixture with the local substrate. And, unlike the stylized model of Bellwood, it seems that there were multiple migrations after the initial pulse which reshaped the genetic and cultural landscape of human societies in the wake of agriculture.” – from razib. […]

    Read More
  89. @Karl Zimmerman
    Proto-Indo-Europeans are thought to have been a roughly 50/50 hybrid between something from the Caucasus and Eastern European Hunter Gatherers (EHG) - with the latter supplying the patrilineage. But EHG is not the same thing as Western Hunter Gatherers, with the main difference being a much higher enrichment of Ancient North Eurasian. ANA is undetectable in most WHG, aside from Scandinavians, which had proportions coincidentally similar to modern Northern Europeans (around 10%). EHG had something in the range of 40% ANA admixture. Whether this was because they were a separate "race" at the time, or because they formed through a hybridization of WHG and ANA, is currently up in the air. It could even be that WHG was actually formed through hybridization of EHG and something else entirely.

    Regardless, they were not the same population during the time window examined, and there's no reason to think they spoke languages of the same family. There's arguments there is a pre-Uralic substrate in the Saami language, which may be the only relic of whatever Mesolithic European language family their was. But given the relative closeness geographically to the Karelian EHG samples, this may just be another EHG language family.

    My readings on population genetics has lapsed a bit so what is ” ANA ” ? Thanks.

    Going off on a tangent: Is there any consensus on whether R1b was introduced by Indo-Europeans? Since R1b is also found in some areas of Sub-Saharan Africa and there is no indication it had anything to do with Indo-European speakers I’ve been skeptical on whether it like R1a introduced IE languages.

    Read More
  90. @Asya Pereltsvaig
    Dmitry, I don't know about you, but I take time off work on Shabbat.

    but I take time off work

    :) oh, I’m sure it was the weekend stuff which slowed down publication of replies on this site (including yours) – but more specifically the fact that it was the ASHG weekend, with its flow of exciting news (including 7000 years old onset of intermixing of Eastern Hunter-Gatherers with the Caucasus-like agriculturalists, perhaps East of the Caspian Sea; Srubnaya culture being linked to the origin of the Indo-Aryans, and many, many new datapoints from the Steppes) .

    Specifically with respect to Johanna Nichols evolving views on the early Steppe migration routes … I got my clues from the following recent abstract ( http://conferences.saxo.ku.dk/nordic-tag-2015/documents/N-TAG_session_list_updated__2_.pdf/Final_NTAG_booklet__1_.pdf ), which cites evidence for the crucial role of interrelations between the Steppe people and the hunter-gatherers of the fringe woodland zone for the successful migrations (including the notion of “The Fur Road” along the Northern boundaries of the Steppe, and a “catalytic role” of Proto-Uralic in the early spread of IE languages).

    Read More
    • Replies: @Asya Pereltsvaig
    Re: Nichols' work, it looks very interesting, but it's hard to get much out of a short abstract like that. I'd still wait for a longer expose.
  91. @Dmitry Pruss

    but I take time off work
     
    :) oh, I'm sure it was the weekend stuff which slowed down publication of replies on this site (including yours) - but more specifically the fact that it was the ASHG weekend, with its flow of exciting news (including 7000 years old onset of intermixing of Eastern Hunter-Gatherers with the Caucasus-like agriculturalists, perhaps East of the Caspian Sea; Srubnaya culture being linked to the origin of the Indo-Aryans, and many, many new datapoints from the Steppes) .

    Specifically with respect to Johanna Nichols evolving views on the early Steppe migration routes ... I got my clues from the following recent abstract ( http://conferences.saxo.ku.dk/nordic-tag-2015/documents/N-TAG_session_list_updated__2_.pdf/Final_NTAG_booklet__1_.pdf ), which cites evidence for the crucial role of interrelations between the Steppe people and the hunter-gatherers of the fringe woodland zone for the successful migrations (including the notion of "The Fur Road" along the Northern boundaries of the Steppe, and a "catalytic role" of Proto-Uralic in the early spread of IE languages).

    Re: Nichols’ work, it looks very interesting, but it’s hard to get much out of a short abstract like that. I’d still wait for a longer expose.

    Read More
  92. @Davidski
    We don't yet have any Hittite genomes, but it's likely that at some point they will be available.

    TC genomes might also be available soon. But since TC is partly of Starcevo origin, and we already know that Starcevo people were in large part of Neolithic Anatolian origin, then it's unlikely that they'll fit the bill as ancestral to Yamnaya and related groups on the steppe.

    Still fascinated by the R1b transition to R1a-Z93 in Poltavka and Srubnaya which coincides with the emergence of European Neolithic ancestry in the East of the region (and obviously reinforces the opinion that the Indo-Iranians descend from these Steppe cultures).

    Davidski – what do you think of the earliest R1a-Z93, Poltavka male 10432 / SVP42? He was buried at Potapovka I, a cemetery which continued to be used in later Bronze Age and in a grave which intersected with a later burial, in a way which removed 60% of the bones, but apparently was directly carbon-dated to 2900-2500 BCE. Note that this Poltavka sample was an outlier similar to Srubnaya in autosomal composition, while the other 3 mapped with Yamnaya. Could it be a misplaced cranium from an overlapping later-age burial? Do we know what material was used for carbon-dating?

    Archaeologically, Srubnaya burials has long been likened to the enclosed ossiaries in Central-East Europe’s Globular Amphora Culture, of 3400-2800 BC (and perhaps both explained by a proto-Avestan belief against contamination of sacred Earth). Are there any Globe Amphora DNA data?

    Of course another theory (Sarianidi etc.) links the emergence of Avestan beliefs to Bactria-Margiana and the city culture of Gonur-Depe (2500 BC and on) (purity of flame and cinders, human bones stripped of flesh for burial, special burials of “impure” humans, ritual sacrifice of colts with cut heads and tails…) Soghd and Margiana appear to be mentioned right after the mystic urheimat of Avestans in the myth. Andronovo-type culture existed in the surrounding lands; Andronovo ceramics becomes frequent in town.The non-IE language of Margiana has been hypothesized by Lubotsky to be the source of numerous borrowings shared in Proto-Indic and Iranian languages but not elsewhere in the IE family, which refer to construction, irrigation, clothes, utensils, and cult, as well as to animals and birds many of which – like donkey and camel – were domesticated in Margiana-Bactria. The group origin of the Margiana language can’t be reconstructed but its features are similar to the borrowings specific to Proto-Indic but not Iranian, indicating longer or more extensive contact of the Proto-Indians with this unknown language. But I’m sure there is no aDNA data from anywhere in Margiana…

    Read More

Comments are closed.