Monday, June 28, 2021
The PIE homeland controversy: June 2021 status report
Archeologist David Anthony has made several appearances online recently to promote his theories about the origins of the Corded Ware and Yamnaya cultures and peoples.
In a clip on Youtube he reiterated his theory that the so called Iranian-related ancestry in the Yamnaya people actually came from what is now Iran, and, more precisely, that it was carried by hunter-gatherers who travelled relatively rapidly from the South Caspian region into the Volga Delta in what is now Russia.
It's still a complete mystery to me as to why a group of hunter-gatherers from the South Caspian would undertake such a migration, instead of, say, expanding their range gradually over thousands of years, first into the Caucasus and eventually into Eastern Europe.
But there's a more serious problem with Anthony's theory: it contradicts the currently available ancient DNA. That's because the so called Iranian-related ancestry in the Yamnaya people is most closely related to the Kotias and Satsurblia hunter-gatherers from what is now Georgia, and these hunter-gatherers form a separate clade from the earliest samples from what is now Iran. For instance, see here and here.
Also, in a podcast on Razib's blog, Anthony doubled down on his theory that Y-chromosome haplogroup R1a was closely associated with Yamnaya plebs who were excluded from Kurgan burials, and, as a result, their remains haven't yet been sampled.
At least this theory isn't yet contradicted by ancient DNA, but it's more complicated and less parsimonious than my theory, which posits that R1a, or rather R1a-M417, was simply a very rare lineage in the Yamnaya population, and that it only became a common and widespread marker thanks to the Corded Ware expansion (see here).
Intriguingly, my understanding is that there are several unpublished R1a samples from the Caspian and Volga steppes at Harvard's David Reich Lab that have been classified by its scientists as Yamnaya outliers. Of course, Anthony is collaborating on at least one major paper with this lab (see here).
Ergo, I strongly suspect that Anthony's theory is in part based on these Yamnaya outliers. However, I also believe that these samples are wrongly dated and probably represent Scythians and/or Sarmatians. I'll be able to look into that if they're ever published.
Speaking of the David Reich Lab, its leading scientists, David Reich and Nick Patterson, have also made appearances online recently, on Youtube and Razib's blog, respectively, to reveal that the Corded Ware and Yamnaya peoples aren't just very similar genetically, but in fact close cousins.
This is a very interesting finding. Apparently it's based on a relatively high level of Identity-by-Descent (IBD) segment sharing between Corded Ware and Yamnaya samples, but that's all I know. I'm guessing that the relevant paper is coming soon (that is, within the next five years).
However, the long-standing question that the readers of this blog want to see answered is not whether the Corded Ware and Yamnaya peoples are close cousins, but whether Yamnaya migrants founded the Corded Ware culture. The obvious way to prove that they did is to find at least one ancient population unambiguously classified as part of the Yamnaya horizon that is rich in the typically Corded Ware Y-haplogroups R1a-M417 and R1b-L151.
See also...
On the origin of the Corded Ware people
The PIE homeland controversy: January 2019 status report
The PIE homeland controversy: August 2019 status report
Subscribe to:
Post Comments (Atom)
532 comments:
«Oldest ‹Older 401 – 532 of 532@Rob
What I was saying is that just because the Yamnaya population mainly seem to coming from the Western side of the territory it does not follow that the NW Pontic was also colonized / transformed early. Now if the Yamnaya remains in that region are directly dated and there is no time gap, that is a different story. In this case my assumption was wrong.
Are they dated? And I am not talking about Ukraine in general, but specifically west of the Dnieper, the (former) Usatovo territory and the forest steppe north of it.
@Davidski
"I seem to remember that the earliest Yamnaya kurgans are actually in Ukraine and the Samara region.
This has always stuck out for me, because it seemed like Yamnaya appeared more or less at the same time at the western and eastern ends of the PC steppe."
Fast expansion then. Fast enough that carbon dating cannot differentiate. I'd say that Afanasievo was also the product of the initial expansion. It is not really an offshoot of Yamnaya, it is the result of the same expansion that created Yamnaya.
"I am aware of the theory that Yamnaya is a direct offshoot of Repin, but this is an archeological theory, and it might end up being contradicted by aDNA."
The alternative is that both Repin and Khvalinsk are overrun? While both of them seem to be good archeological precursors for Yamnaya?
BTW, are there Repin samples in the pipe?
@Andrzejewski
No I do not think it directly came from ANF. K1 and H13 were found in the Mesolithic Iron Gates hunter gatherers, so this leads me to believe that there’s some sort of common West Eurasian, Dzudzuana-like ancestry at play here, and it’s reflected in some mtDNA Haplogroups, like H, T, and K. Yeah I know, I’m not saying anything profound, in fact to many it may appear obvious, but I believe some of these linages show a link between Mesolithic European hunter gatherers (Iron Gates, Volosovo, and Karelia), and West Asian hunter gatherers, such as the ones found in Caucasia, the Zagros, and Anatolia.
@epoch
David Anthony doesn't even know that there are different subclades of R1b. That is, he actually seems unaware that the R1b in Yamnaya is different from the R1b in Corded Ware/Bell Beakers.
On top of that, he appears to be unaware that the Corded Ware spread wasn't just a migration, but also a very rapid population and genetic expansion.
So how is he going to tackle all of the intricacies of what happened there, if he doesn't know the basics?
@epoch, not on my computer so can't check the anno files but there at least 5 other Poland CW samples I can remember, so I don't know why of the 8 in Ringbauers plot they would have to be the SE Poland M269 samples.
@ Slumberry
Yes based on C14 data, Yamnaya is the same age East and West of the Dnieper ~ 3300 BCE because it was a ''horizon''. Yes, some have argued that the Repin is the forerunner for Yamnaya, but this isn;t really the case. Yamnaya represents a new cultural phase in the steppe, and in theory any number of late Eneolithic groups contributed to it (Lower Mikhailovka, kvitanya, Repin, Dereivka, Cernavoda III, Usatavo, Vikhanitsy, etc)
All we can say at the moment is, the earliest evidence for Yamnaya-related R1b-M269 is from Bulgaria . The earliest manifestation of Balkna-Yamnaya related I2a2 is in Hungary & the Dnieper. This is not consistent with 'rapidly expanded from the Don, or Volga' is it ?
I'm aware that from a genome-wide perspective, the characteristc 'steppe signature' is from well to the East of the Dnieper, but that admxiture needs to be clarified. This could have happened via a brisk migration (less likely IMO) or a more gradual admixture/ shifts in mating networks between 5000 and 3000 bce (albeit with a sharp upturn ~ 3500 bce).
@Matt,
These are the population pairs that I’m referring to with G25 distance runs.
Distance to: Udmurt
0.06070359 Tatar_Kazan
0.08601374 Tatar_Lipka
0.08885307 Tatar_Mishar
0.09797134 Bashkir
0.10966764 Tatar_Crimean_steppe
0.11544097 Russian_Pinega
0.12750545 Finnish_East
0.13114207 Turkmen
0.13310796 Tatar_Siberian
0.13366572 Turkmen_Uzbekistan
0.13509091 Russian_Kostroma
0.13773585 Tajik
0.14089777 Tajik_Shugnan
0.14153654 Tajik_Rushan
0.14393915 Iranian_Turkmen
0.15093452 Uzbek
0.15098617 Tajik_Ishkashim
0.15121243 Cossack_Kuban
0.15155844 Tatar_Siberian_Zabolotniye
0.15961422 Tajik_Yagnobi
0.15979505 Russian_Tver
0.16047252 Tlingit
0.16559742 Russian_Kursk
0.16919530 Russian_Orel
0.16977181 Cossack_Ukrainian
0.17017327 Estonian
0.17152629 Mansi
0.17241783 Nogai
0.17265718 Russian_Voronez
0.17442936 Ukrainian_B
0.17552127 Ukrainian
0.17796271 Russian_Smolensk
0.18020743 Swedish
0.18059071 Polish
0.18066006 Slovakian
0.18132243 Norwegian
0.18316431 Czech
0.18369906 Latvian
0.18579149 Khanty
0.18695227 Shetlandic
0.18789055 Orcadian
0.18817467 Scottish
0.18894553 Irish
0.19164537 English
0.19301456 Welsh
0.19550134 English_Cornwall
0.19868197 Karakalpak
0.20064421 Ket
0.21850976 Spanish_Galicia
0.22032703 Italian_Piedmont
0.22054734 Ashkenazi_Russia
0.23258082 Basque_Spanish
0.23862163 Ashkenazi_Germany
0.23968753 Italian_Calabria
0.24615103 Sephardic_Jew
0.30864498 Cree
Distance to: Mari
0.11547308 Bashkir
0.11670750 Tatar_Kazan
0.13300496 Tatar_Lipka
0.13857526 Tatar_Siberian
0.14194399 Tatar_Mishar
0.14622919 Tatar_Siberian_Zabolotniye
0.14750185 Tatar_Crimean_steppe
0.15879922 Mansi
0.16281645 Russian_Pinega
0.16609517 Turkmen
0.16666745 Turkmen_Uzbekistan
0.17214423 Tlingit
0.17229907 Uzbek
0.17348576 Khanty
0.17705239 Finnish_East
0.17783813 Nogai
0.18422090 Iranian_Turkmen
0.18522712 Russian_Kostroma
0.18790293 Tajik
0.19535284 Tajik_Shugnan
0.19748239 Tajik_Rushan
0.19884563 Cossack_Kuban
0.19996889 Karakalpak
0.20260554 Tajik_Ishkashim
0.20267386 Ket
0.21081083 Russian_Tver
0.21232139 Tajik_Yagnobi
0.21549021 Russian_Kursk
0.21763448 Estonian
0.21906474 Russian_Orel
0.21984970 Cossack_Ukrainian
0.22059257 Russian_Voronez
0.22160807 Ukrainian_B
0.22486058 Ukrainian
0.22557314 Russian_Smolensk
0.22981453 Polish
0.23101851 Slovakian
0.23128921 Latvian
0.23172467 Swedish
0.23388498 Norwegian
0.23404842 Czech
0.23920761 Orcadian
0.23965138 Shetlandic
0.23980132 Scottish
0.24118352 Irish
0.24241418 English
0.24430668 Welsh
0.24634058 English_Cornwall
0.26224418 Ashkenazi_Russia
0.26265285 Spanish_Galicia
0.26465765 Italian_Piedmont
0.27549337 Basque_Spanish
0.27907793 Ashkenazi_Germany
0.28108040 Italian_Calabria
0.28515978 Sephardic_Jew
0.31489922 Cree
Most East Asian-shifted Udmurt individual:
Distance to: Udmurt:udmurd8
0.07878436 Tatar_Kazan
0.08187760 Bashkir
0.09175964 Tatar_Lipka
0.10852740 Tatar_Crimean_steppe
0.10922678 Tatar_Mishar
0.11499101 Tatar_Siberian
0.12768661 Turkmen
0.12839383 Turkmen_Uzbekistan
0.13030726 Tatar_Siberian_Zabolotniye
0.13792423 Russian_Pinega
0.14008865 Uzbek
0.14366467 Tajik
0.14503560 Iranian_Turkmen
0.14931869 Tlingit
0.15002433 Finnish_East
0.15051991 Mansi
0.15082985 Tajik_Shugnan
0.15262305 Tajik_Rushan
0.15733351 Nogai
0.15807654 Russian_Kostroma
0.15932898 Tajik_Ishkashim
0.16433979 Khanty
0.17150233 Tajik_Yagnobi
0.17424742 Cossack_Kuban
0.18039907 Ket
0.18225833 Karakalpak
0.18263746 Russian_Tver
0.18871031 Russian_Kursk
0.19204759 Russian_Orel
0.19219383 Cossack_Ukrainian
0.19334510 Estonian
0.19543075 Russian_Voronez
0.19648754 Ukrainian_B
0.19809225 Ukrainian
0.20046634 Russian_Smolensk
0.20108449 Swedish
0.20124156 Norwegian
0.20286088 Slovakian
0.20302204 Polish
0.20437588 Czech
0.20673314 Shetlandic
0.20715504 Latvian
0.20770837 Orcadian
0.20785913 Scottish
0.20848813 Irish
0.21127078 English
0.21262846 Welsh
0.21496495 English_Cornwall
0.23394254 Ashkenazi_Russia
0.23488790 Spanish_Galicia
0.23643305 Italian_Piedmont
0.24896226 Basque_Spanish
0.25155926 Ashkenazi_Germany
0.25329270 Italian_Calabria
0.25811284 Sephardic_Jew
0.29686940 Cree
Point taken.
Genome-scale sequencing and analysis of human, wolf, and bison DNA from 25,000-year-old sediment
Summary
Cave sediments have been shown to preserve ancient DNA but so far have not yielded the genome-scale information of skeletal remains. We retrieved and analyzed human and mammalian nuclear and mitochondrial environmental “shotgun” genomes from a single 25,000-year-old Upper Paleolithic sediment sample from Satsurblia cave, western Georgia:first, a human environmental genome with substantial basal Eurasian ancestry, which was an ancestral component of the majority of post-Ice Age people in the Near East, North Africa, and parts of Europe; second, a wolf environmental genome that is basal to extant Eurasian wolves and dogs and represents a previously unknown, likely extinct, Caucasian lineage; and third, a European bison environmental genome that is basal to present-day populations, suggesting that population structure has been substantially reshaped since the Last Glacial Maximum. Our results provide new insights into the Late Pleistocene genetic histories of these three species and demonstrate that direct shotgun sequencing of sediment DNA, without target enrichment methods, can yield genome-wide data informative of ancestry and phylogenetic relationships.
https://www.cell.com/current-biology/fulltext/S0960-9822(21)00818-6
@Davidksi
Pleb question, but are there any exciting ancient DNA papers and/or results coming down the pike in the next few months? I’m guessing so far just Volosovo?
@Ramber, I had some G25 euclidean distance vs f2 stats about for Udmurt, so here are a few for you, comparing moderns, to help visualise. Here: https://imgur.com/a/2UCjxHY
Unfortunately I was looking mainly at a different set of populations, so it doesn't show most of the closest population in G25 to them (even those in Human Origins), but you can see the general relationships (which is typical for f2 stats to G25 data).
You can see what I'm talking about maybe more clearly; the Global 25 euclidean distance has a non-linear relationship to the f2 statistic (in this case using Hill's sigmoidal, but there are other fits possible, like simple power relationship). And there is also some obvious extra drift in some populations which is not in g25 (Kalash, Karitiana, Ami, Nivh, Kusunda, some other populations on a low level) and there might be some degree to which I'm not using exactly the same population samples as G25 that causes some diferences (e.g. Yemeni / Algerian, for Ukrainian North I used G25 Cossack, who probably have some Udmurt related geneflow according to the distances you posted).
The last plot, which shows the rank order of population distances from Udmurt, is correlated pretty linearly at r^2 of 0.92 (which is pretty high). Some of the effects on rank are more dramatic than the raw-distance difference as a function of which populations I've chosen to include, so it's just for an impression.
So basically again to repeat myself (boringly) the G25 is fairly reliable for working out the order of how close populations are to each other... it might just not tell you perfectly about the actual distance. At least for modern people. But just being relatively correct is probably pretty useful anyway for most of what we want to use it for (e.g. things like Vahaduo).
Has anyone done an analysis of deep subclades to see whether the Balkan and Caucaus Z2103s are actually direct descendants of the Yamnaya groups we do have?
People tend to look over the fact that there are indeed A LOT of modern day Indo-European speaking people with Z2103 and just focus on R-L51 or R1a.
@Simon Stevin
There are a lot of new aDNA papers coming on things like the Eneolithic steppe, Rome, Greece, the Migration Period, the origins of Jews, the Uralic expansions, and pretty much anything else you can think of.
But I don't have a clue when they'll be published.
@Rambler,'
It is interesting you point out they are slightly closer to some Asians than Europeans. I never tested this and did not know this.
I suppose this can be explained by them having mixed European and Asian ancestry. And also due to hose Asians they are closer to also having mixed European, SW Asian, and Asian ancestry.
In bronze, iron age there was a lot of back and forth geneflow between Europe (Russia) and Asia. Indo European languages and Uralic languages are both Eurasian wide due to this.
You can also do distance for European Iranians (ie Scythians). They are closer to ancient Central Asian Iranians and modern Tajiks than to mainstream Europeans.
@Draft Dozen
"the numbers that I gave, are from the dissertation(2016) of Veselovskaya E.V."
I checked it, will prove useful to me in the future.
"Do you think, that the larger width than the more massive? Neither the large length nor the large width is an indicator of the massiveness of the skull. The level of massiveness is when the skull is heavy or light."
What? It is determined by circumference and most importantly dimensions, and a skull with a large dimension will also naturally have more bone volume.
@ Matt: many thanks for your efforts! What's your view on Tadjik vs. Udmurt relationship? Are they in reality close to each other? Taking into account the sizeable number of Indo Aryan loans in Uralic, one could easily imagine that there really is some kind of a connection.
Tajiks are Iranian speakers. The nearest Indo-Aryans are the Kalasha.
Anyway, there are two main reasons why Uralic speakers and Tajiks are similar:
- they both have a lot of Bronze Age Eastern Euro ancestry, via the same and different streams, like Andronovo
- they both have significant levels of East Eurasian ancestry, from Siberia in the case of Uralians, and from a variety of sources in the Tajiks.
@EX
‘There are more similar vague Turkish etymologies, see previous posts.’
I will try to explain this. There is a verb "kur" - to build (in some languages it sounds like "kor" (u>o). And there is a verb "kory" - to protect, strengthen (also "kuri" - u>o). Kurgan is translated as erected (object) , korigan is translated as "protected" (object, place) – fortress. (...)
The burial mounds / kurgans can be built only up / Go'Ra and inside a circle / ring / o+KRa"/oNG.
Primality G>K, see:
Onur Dincer said...
‘Go’R+Ka > Ko’R+GaN > Ko’R+HaN > Ko’R+GaN’
Even if the word kurgan is ultimately of IE origin (which is yet to be demonstrated), the g > k transformation should have happened in Turkic. So the word kurgan should have been formed in Turkic regardless of its ultimate origins, then was loaned to Old East Slavic and from East Slavic spread to other European languages. (...)
https://eurogenes.blogspot.com/2021/04/the-history-of-scythians-gnecchi.html?showComment=1623113224564#c3658482474634565711
(…) In the Old Russian language and in many Turkic languages "kurgan" means both (a burial mound and a fortress : there are recorded forms - kurgan, korgan and korygan). (…)
Can you prove it and give any quote and its source that allegedly in the Old Russian language ‘kurgan’ means both a burial mound and a fortress, please?
For comparison, here is another word related to above meanings and Go'Ra:
Gród / GRo’D
https://en.wikipedia.org/wiki/Gord_(archaeology)
https://pl.wiktionary.org/wiki/gr%C3%B3d#pl
https://en.wiktionary.org/wiki/gr%C3%B3d
https://en.wiktionary.org/wiki/Reconstruction:Proto-Slavic/gord%D1%8A
https://en.wiktionary.org/wiki/%D0%B3%D0%BE%D1%80%D0%BE%D0%B4%D1%8A#Old_East_Slavic
Another example but secondary distorted see G>K, D>T:
kuorat / Ko’oRaT
https://en.wiktionary.org/wiki/%D0%BA%D1%83%D0%BE%D1%80%D0%B0%D1%82#Yakut
(…) It seems that there was a contamination of these two meanings in the ancient Turkic languages.(…)
No kidding! How is it even possible for some ancient Turks to have 'contaminated' their own two meanings?!! Caramba! :-)
‘As I can see, you do not question anymore the Slavic etymology for the word Altay / (Z)+aLTay’
(…) No. (…)
It is no good... ;-)
(…) Altai has a good Mongolian etymology. The Turkic etymology is problematic because Turkic peoples use only form "altyn" (gold), while Mongols use two forms - altyn and alt (gold). But in ancient times, it could be different. The etymology of altyn is not entirely clear (possibly from the turkic word "alty" or "altyn" – low) (...)
It follows from the above that you can not undermine my Slavic / IE etymologies, but you are undermining Turkish ones.
So when did the Proto-Mongols first come to Z+Altay? According to you, who were they?
According to Onur Dincer, the Proto-Mongols were Donghu people see:
https://eurogenes.blogspot.com/2021/04/the-history-of-scythians-gnecchi.html?showComment=1619163069510#c4057193459287833651
‘Proto-Mongols, on the other hand, might have been the Donghu people, their eastern neighbor.’
https://en.wikipedia.org/wiki/Donghu_people
@Matt
Thanks for the data and detailed explanation! I know you are looking a different set of populations, but can you add some more populations such as Tatar_Siberian, Tatar_Siberian_Zabolotniye, Uzbek, Turkmen, Nogai, Karakalpak, Khanty and Iran_Turkmen into your f2 stats and G25 data if you also have them? I want to see if they would be closer to Udmurts than Tajiks, Finns and Russians are or not. I’m guessing the Russian sample in the f2 stats have a lot of Uralic ancestry? Yep the Cossack samples I used could have some Udmurt or other Finno-Permic gene flow.
Btw in the final graph, does it show the Mansi being genetically further away from Udmurts than Hungarian, Kabardin and Czech are? It’s interesting because in this G25 distance, it seems the opposite.
Distance to: Udmurt
0.11234591 Bashkir
0.13310796 Tatar_Siberian
0.17017327 Estonian
0.17152629 Mansi
0.17442936 Ukrainian_B
0.17552127 Ukrainian
0.18103482 Hungarian
0.18132243 Norwegian
0.18305190 Kabardin
0.18316431 Czech
Would you say G25 euclidean distance run is more accurate overall than Dodecad K12b, Eurogenes K13 Vahaduo distance runs for Uralics?
Also can you also do the same G25 and f2 states for the Mari?
@Andrzejewski
I think the young Dolph Lundgren phenotype did not become more common until the middle or late bronze age. However did PIE have the same craniometrics as any modern pops? Do you know?
Didn't see this mentioned while skimming the comments, but there's a lot of ink spilled here so maybe I missed it (apologies if so): on Harald's IBD heatmap, the numbers in brackets next to the axis pop labels are the number of samples in that pop, and the fraction labels within each cell like "1/10" means that 1 of the 10 possible pairs between the x- and y-axis pops displayed at least 1 IBD segment of length 16-20 cM.
Sharing 1 segment of 16-20 cM is a low-ish bar to clear, especially for a pair of low-depth, crappily imputed samples. And with a small number of pairs, maybe you got "unlucky" and happened to sample an outlier with an unusual migrant ancestor in one of the two pops. All of which is to say that I wouldn't start spending hours theorising about zany scenarios to explain the 1/10 for Lithuania_LN and Russia_Volga_Yamnaya just yet.
@Davidski “ These Volosovo samples haven't been published yet, but yes, they're what we'd expect anyway, and rich in R1b.”
Yes, right. And since they have no R1a1 in them, it corroborates my gut feeling that PIE grew independently out of some more westerly cluster of R1a1 carriers, rather than among any EHG forager (like Volosovo or Combed Ceramic), neither a CHG one. Fledgling PIE was the language of Sredny Stog (or maybe Dnieper Donets before it) with many influences of GAC and to a much lesser degree Tripolye and Narva HG.
@Genos Historia
Yes, I found it fascinating as well how Udmurts, Maris seem to be closer to Central Asians and Turkics including Tajiks, Uzbeks, Turkmens, Nogais, Siberian_Tatars and Ugrics (Khanty, Mansi) than to most Europeans.
Yup I think it is due to them having both European and East Asian ancestry. I don’t think they have SW Asian but rather Iranic/Central Asian-related ancestry which pulls them closer to Tajiks and Turkics. In Vahaduo G25 and Gedmatch models, you would need to add an Iranic source such as Iranian_Zoroastrian or Tajik_Yagnobi (the latter give better distance fits it seems) to give these Uralics a good fit.
I think you are right regarding the back and forth gene flow between the two regions.
I will try it out. I also think they will be closer to those ancient Central Asian Iranians and modern day Tajiks.
@Rob
OK, I accept then that Yamnaya indeed covered all corners.
However about this part:
"All we can say at the moment is, the earliest evidence for Yamnaya-related R1b-M269 is from Bulgaria . The earliest manifestation of Balkna-Yamnaya related I2a2 is in Hungary & the Dnieper. This is not consistent with 'rapidly expanded from the Don, or Volga' is it ?
It does not support it, but as far as "consistency" goes it is perfectly compatible with it. Just because these lineages were present here and there before the expansion it does not follow that they could not expand rapidly from a specific sub-region. The estimated TMRCA of R1b-Z2103 is older than Yamnaya, but it was not necessarily frequent, let alone dominant, in any bigger region before it and the center of expansion can be different from the birthplace.
@ALL
It 's a bit out of topic, but i would like to ask you guys a question. Someone just told me that David Reich said once it is not possible to distinguish between Insular Celts and Anglo-Saxons ancesrry component.
Is it true or false we can't?
@J.S.
Take a look at the second plot here.
https://eurogenes.blogspot.com/2019/12/avalon-vs-valhalla-revisited.html
So yes, it's possible, and it's also possible to do it in admixture tests using haplotypes.
Of course, whether David Reich actually claimed that it's not is a different matter.
@Davidski
Thank you very much.
@all
https://twitter.com/benmpeter/status/1414946817373057026
"Modelling complex population structure using F-statistics and Principal Component Analysis"
https://www.biorxiv.org/content/10.1101/2021.07.13.452141v1
@Ramber
Would G25 be a lot better to utilize for modelling?
Definitely!
Would the "Vahaduo tools for G25 custom calculators" be much more suitable for distance comparisons than those of "Gedmatch calcs"?
Yes. Although I strongly recommend using original G25 scaled spreadsheets and original tools that you can find here:
https://vahaduo.github.io/
Why shouldn't they be used for modelling?
"Components" in the so called "calculators" overlap to a certain degree and distances between them are different and unknown. Without the knowledge how they relate to each other it's impossible to correctly translate results from a "calculator" into genetic distances.
p1. @Huck and Ramber, OK, since you're interested I will do that with as many of those as are available. Just as a point though, this takes a long time as computing the f2 and fst have to be done for any "new" populations I'd want to add on the same set of SNPs as the old f2 or fst, or they would become slightly non-comparable at fine levels. So I have to recompute the whole set which takes quite a long time.
(I won't go into a full on attempt at explanation of why this happens, partly because I'm not 100% sure why myself, though I have some ideas, basically thinking the choice of including capture ancient dna and different numbers of populations can have some effects on how many polymorphic/monomorphic SNPs are available, and this affects f2).
#Here's a plot to show the f2 scores, comparing f2 and fst, and also plotting f2 against a random number (just so you can see the f2 without population labels overlapping): https://imgur.com/a/xz3lgGN
(Fst and F2 are pretty correlated. Fst is approximately like ((f2(between_population)-f2(within_population)) / f2(between_population)). Like a ratio of f2 statistics where 1 indicates all the allele frequency difference is between population and 0 that the allele frequency difference between populations is indistinguishable from difference within the population. This largely means that they're linearly correlated and only where there is much more difference within population, with African populations, or much less like with Native Americans that there are significant differences).
The lowest f2 and fst in the set I checked was Tatars from the Volga (Volga / Kazan). There is still quite a bit of distance because Udmurt seem to have quite a significant founder effect in this data, but that just moves them away from all populations, doesn't change the order of distances.
Also pastebin in case you wish to try and compare directly and plot in Past or another graph software: https://pastebin.com/a3g3KNPf
I will try and do a comparison with G25 in a bit today or tomorrow if I can - it's a bit trickier since the populations don't automatically have the same labelling so it takes some time and I might not match up exactly right (like for example, not sure how to really match up "Russian" here, since it doesn't match clearly to any of the different Russian populations in G25). And I didn't really set it up last time I did it with a mind to doing it again, unfortunately.
p2. Re your question about Mansi, I would have to say on that point, got to stress again the direct f2 distances can be heavily affected by founder effects, while these can disappear in projection onto PCA that doesn't include/heavily sample the population. If you have two populations who are closely related and they have founder effects, then the direct distance between them can be high even though they shared ancestry until recently. I think that's what happened here from what I can remember looking at the f2 tree, where the Mansi had quite a big branch, i.e. probably a founder effect (https://imgur.com/a/RyNpIi0). Sometimes when people talk about the distance between populations what they really mean is something like "The distance between these populations *disregarding* any recent founder effects", and something like the PCA projection method is maybe a better way at getting to that than the direct distance itself.
Re; other tools, I would think the G25 euclidean distance is definitely more meaningful as a distance. The other calcs based on ADMIXTURE might still be pretty good at finding the closest neighbours between populations - I haven't really tested it, but they often sample lots of modern populations quite well and the ADMIXTURE model works fairly well with good quality modern data (more problems with ancient dna). Obviously G25 has more dimensionality in it to cover the structure and that's in its favour. But anyway, for the other calcs based on ADMIXTURE components, I would say any results that shouldn't be interpreted as a genetic distance really, because it's just something like a statistical summary of how much difference there is in amounts of the ADMIXTURE components, not anything that's actually like a meaningful distance or takes into account how different the components are from each other. (There are ways that can be done but it's more complicated than the calcs use). If that makes sense.
Unfortunately I can't do this for the Mari as the Human Origins panel I'm using seems to only have 1x Mari and you need at least 2x in a population to run these stats. (If there's only one sample it can't distinguish between what is the difference between an individual and their population, so you get a lot of high differentiation that's not really characteristic of a population level).
@Huck, to be honest, I don't know enough about the populations to say; it seems to me like this is mostly due to fairly deep, fairly ancient components having similarities in proportion but I don't know enough to speculate.
@ Slumberry
'' but as far as "consistency" goes it is perfectly compatible with it. Just because these lineages were present here and there before the expansion it does not follow that they could not expand rapidly from a specific sub-region. The estimated TMRCA of R1b-Z2103 is older than Yamnaya, but it was not necessarily frequent, let alone dominant, in any bigger region before it and the center of expansion can be different from the birthplace.''
Yes I know how TMRCA works & what it entails
You might be in theory correct (that R1b-M269 was resting in the Don bend for ex until 3000 BCE) ; but i suspect not
In the (2020) book The World of the Oxus Civilization, in Chapter 26 untitled "The Oxus Civilization and the Northern Steppes", Gian Luca Bonora comments that east of Samarkand, in Zeravshan valley, Uzbekistan, there's a place called "Zhukov sanctuary" that in the late 4th to early 3rd millennium BCE became the meeting point of diverse Eurasian cultures like Kel'teminar, Pit-Grave (Yamnaya), and Afanasievo. He writes: "the predominant Afanasievo cultural component in the pottery assemblage seems to indicate a hypothetical movement of an Afanasievo group through Central Asia in search for new resources and pastures. Nevertheless, the aforementioned Kel'teminar component, as well as some cultural traits belonging to the Pit-Grave culture of the Cis-Urals steppes, cannot be ignored."
Is it possible the presence in (southern) Central Asia of Pit-Grave or Yamnaya culture at such an early date around 3000 BCE?
I haven't yet seen any evidence that Yamnaya was present in Turan.
The closest we have is a Yamnaya sample from a burial in Karagash, Kazakhstan.
That's still basically on the steppe.
@Davidski
"I haven't yet seen any evidence that Yamnaya was present in Turan.
The closest we have is a Yamnaya sample from a burial in Karagash, Kazakhstan. That's still basically on the steppe."
Thanks for the reference.
Gian Luca Bonora's chapter on Zhukov sanctuary was based on an earlier paper by Nona A. Avanesova (2013), in French, published by journal Paléorient. The full paper here:
https://tinyurl.com/4zzktr5k
@vahaduo, nice paper; I've been thinking about how f stats relate to PCA from a more amateurish perspective for a while so very cool to see someone (who has been responsible for many of the more fundamental papers about f interpretation) tackle it from a more theoretical and mathematical way. (It's probably too arrogant of me to think there could have been feedback!)
Interpretations that f3 is equivalent to a distance from a midpoint between two points on a PCA and that being within an n dimensional sphere around that point will define if the stat is negative/positive, and f4 equivalent to an angle between populations on PCA (i.e. a measure of relative position that does not include distance only the direction.)
("PCA and F-statistics are closely related. In fact, the principal components can be directly calculated from F-statistics using multidimensional scaling, which, for squared Euclidean (F2)-distances, leads to an identical decomposition to PCA (Gower, 1966)" - close correspondence being what I found from PCoA MDS ON F2 and fst so useful to see this was totally theoretically predicted
"The F2-statistic is an estimate of the squared Euclidean distance between two populations. It thus corresponds to the squared distance between populations in PCA-space, and reflects the intuition that closely related populations will be close to each other on a PCA-plot" - exactly and approximately what I found from plots and trying to fit relationships using G25 data, and what Florian Price also described comparing PCA distance to the closely related fst measure, which we've shown are fairly linear).
I think one of the cool things is that bridging the interpretation gap between PCA and f-statistics can give more information about how much a PCA is capturing variation relating to projected ancient DNA, or whether projection misses things important to the adna. Also can cut the other way and identify illusory dimensionality in raw f statistics from damage and sequencing error which is avoided in PCA. If between this they can get to qualitatively better means of projection for ancient DNA that would be meaningful for use with these linear combination PCA distance minimisation methods we're all using (and which you've so nicely implemented).
It's a big ask but do you think these methods (i.e. calculating f3, f4 from squared distance data) are implementable in your tools?
SAT29
numsnps:
HO - 5805
1240k - 11701
https://drive.google.com/file/d/1VhrHiCKRE79P2M9D99qSqLxUA47Pacd0/view?usp=sharing
SKRiBHa said...
It follows from the above that you can not undermine my Slavic / IE etymologies, but you are undermining Turkish ones.
There are many place names, whose etymology is very problematic (Greece, Macedonia, Georgia, Armenia, etc.) because toponyms are often very conservative and carry archaic grammatical / syntactic / phonetic forms (often not recorded) and vocabulary.
The example: some Turkic peoples do not understand the meaning of the "kurgan" because they no longer use suffix "gan".
According to Onur Dincer, the Proto-Mongols were Donghu people see:
I think Donghu=Tungusic people.In the old days, Tungusis peoples were a powerful and numerous tribes and were even able to conquer China.
Re: https://www.cell.com/current-biology/fulltext/S0960-9822(21)00818-6
Hi Folks,
I assume this means the Dzudzuana2 genome has been published?
However it does note that "We acknowledge David Reich and Iosif Lazaridis for sharing the data of the Dzudzuana2 genome and helping us with its analysis." so that might be a bad sign.
Cheers,
Guy
Hey @vahaduo.
I'm wondering if it's possible to write in a Chebyshev mode for the g25 Nmonte + distance runner? Just something I've been really hoping to see for some while now.
On another note whats going on wit SAT29!? Isn't it identical to Dzudzuana? Have my prayers been answered!?
SAT29 doesn't have enough data to be run in the G25, if this is where the conversation is going.
Would you guys be interested in substracted ANA genos?
@Ex
There are many place names, whose etymology is very problematic (Greece, Macedonia, Georgia, Armenia, etc.) because toponyms are often very conservative and carry archaic grammatical / syntactic / phonetic forms (often not recorded) and vocabulary. The example: some Turkic peoples do not understand the meaning of the "kurgan" because they no longer use suffix "gan".(…)
Thank you for confirming the problematic nature of the Turkish etymologies for the word kurgan. Again it follows from the above that you can not undermine my Slavic / IE etymologies, i.e. the Slavic etymologies that I presented earlier remain valid.
(…) I think Donghu=Tungusic people. In the old days, Tungusis peoples were a powerful and numerous tribes and were even able to conquer China. (…)
This is irrelevant. Unfortunately, you did not answer my questions, see:
So when did the Proto-Mongols first come to Z+Altay? According to you, who were they?
@Ramber, here are the G25 distance vs f2 crossplots you were interested in: https://imgur.com/a/Cl0dauZ
The samples are generally of the same relationship (strong curvilinear relationship with G25 euclidean distance, linear relationship with G25 euclidean squared distance), though there are some more outliers which I think again are driven by extreme drift that isn't present under G25 (would show up in some high dimension specific to the population that isn't present). Small populations in Siberia, often very high founder effects, I'm guessing.
PAST datasheet format: https://pastebin.com/SDnDb09g
Re: SAT29
f-stats-based PCA:
https://i.postimg.cc/kJ4xN3NT/SAT29.png
Dzudzuana sample is very important, especially to figure out the chg/Iran question for the steppe_en samples. Why is the paper and sample isn't being published? The preprint has been out for so long.
There are 4 samples in this SAT29 paper that cluster with Oase1? What are those? ZKU022 is one of them.
@vahaduo and others, some quick graphics inspired by the paper by Peter, above: https://imgur.com/a/1oou11Z
The first 4 plots show the comparison of Global25^euclidean distance with f2 distance. These are pretty correlated but there are some points off the line.
The next 3 plots show a comparison between an outgroup f3 statistic computed with either f2 distance or the squared euclidean distance from G25. These are even more correlated, which implies that the differences seen in the first 3 are due to some kind of unmodelled drift in Global 25 (e.g. its "extra drift" rather than the G25 model being in any sense wrong). This is also supported by the population's "self" outgroup f3 value generally being higher if computed with f2 statistic rather than Global 25, particularly for Karitiana.
These are all for modern populations so it would be interesting for me to see how ancient populations behave. If there are more differences, that might imply something about how well projection is performing.
Generally it seems like the pseudo-f3 statistic (and then probably also pseudo f4 statistics) would work very well for Global 25 and be very comparable to f3 statistics generally, at least for modern:modern comparisons. (Ancients might need to be checked first).
Davidski,
according to this new paper, "SAT29 belongs to mtDNA haplogroup N like Dzudzuana3 from Dzudzuana cave. SAT29 shares more drift with Villabruna and Dzudzuana2, showing higher genetic affinity to Northern and Western Europeans. 1% Neanderthal ancestry in the SAT29 sample is lower than that of Palaeolithic Europeans due to dilution from Basal Eurasian ancestry, which is similar to Dzuzuana2." Why is SAT29 closer to Northern and Western Europeans instead to Southern Europeans or Middle Easterners? Some folks are arguing for Basal Eurasians to be in fact an "ANA" lineage. Another genetic study concludes that the Iberomaurusians are so far the best proxy for the Basal Eurasians. So I would expect SAT29 with high Basal Eurasian input to be closer to North Africans or Arabs, for instance. Have you any idea?
SAT29 isn't really closer to Northern and Western Europeans. This result is just an artifact of the methods used.
Formal stats have this odd property of making samples more attracted to references that are closer to the extremes of mixture clines, because they're more homogeneous, less admixed etc.
Northern and Western Euros have less complex ancestry than Southern Euros and Near Easterners, so they're more attractive references in these type of tests.
I would have thought that the authors of this paper would've been aware of this by now, but nope.
@Ebizur
Will you comment on the EX’s claims from July 14th, 2021 at 4:09 AM about the problematic nature of Turkish etymologies for the word kurgan?
I remind you that you still have not replied to my reply from July 11th, 2021 at 2:15 AM, where also I questioned your claims on the same subject.
If you do not address to the above, I will consider that you are unable to defend your claims about the validity of Turkish etymologies for the word kurgan.
This will automatically mean that my Slavic / IE etymologies for the word kurgan are most likely correct, unlike the Turkic etymologies which are incorrect, see my Slavic / IE etymologies for the word Z+Altay.
@Romulus
Zlaty Kun and the three oldest Bacho Kiro samples.
@Arza or Matt
Can you do a quick list of f2 similarity stats for SAT29?
@Wise dragon
The paper actually says "Among present-day Eurasian populations, SAT29 shows higher genetic affinity to Northern and Western Europeans rather than Central and Southern Asians." Nothing about Southern Europeans or Middle Easterners.
@epoch
Thanks
I find it interesting they say X, W are under the same N lineage. What mutation do they share?
These are two of the three West Eurasian basal N lineages. They are supposed to be two separate branches.
If 2/2 mtDNA samples from Paleolithic Caucasus belong to a basal lineage related to them that would be quite interesting.
We all know X, W are good candidates for Basal Eurasian mtDNA.
@Matt
Thank you for your efforts. From the plots you did, it seems the Udmurt are indeed genetically closer to several Turkics/Central Asians especially Bashkir, Siberian Tatar, Uzbek, Tajik, Turkmen and interestingly, Forest Yukagir than to most Europeans which really corresponds with the G25 runs that I have did. Can you elaborate more regarding your statement, “Udmurts seem to have a quite significant founder effect in this data”? Do I need to download Past? Actually I plot those coordinates in Vahaduo and it aligns with the data you show me. Hmm the fact that the “Russian” comes closer to Udmurts more than Finnish in the data suggest that the sample seemed to possess significant Finno-Permic gene flow. It doesn’t even match Russian_Pinega or Russian_Kostroma on G25 who have substantial Finnic ancestry? Also could you include some more populations into your plots such as Khanty, Tatar_Crimean_Steppe, Nogai_Karachay_Cherkessia, Iranian_Turkmen, Kho_Singanali, Nogai_Stavropol, Uyghur.DG, Turkish_Northwest, Hazara_Afghanistan (I think the one you have is from Pakistan which seems to be a bit distinct from those from Afghanistan according to G25) and Afghan Turkmens if you have them?
Right. Can you expand more on what founder effects mean? Is founder effects similar to genetic driftness? Overall direct distance is much better at telling the overall story than PCA projection?
I see. I was confused because another user in another forum told me the opposite that, “G25 overestimate the difference between components. “If a population has a considerable amount of Natufian, Eastern Eurasian or South Asian let's say, they will be heavily drifted towards the populations who bear the same outlier admixture as them no matter what other profile(like Slavic, Anatolian...) they have. Dodecad K12b unquestionably has the most solid components for Eurasian population, and the best solution to see actual distances should be to use K12b by taking into account fst scores between its components. This will function same as G25's dimension method, but not going to overestimate the differences as it does due to the fact that being based on a PCA.” As a result, I become perplexed after learning that from him.
I see. Can you create these same f2/fst/G25 plots for the Saami or Chuvash or this Udmurt individual from G25 instead?(he/she is the most East Asian-shifted Udmurt, would be great to compare him/her to the Udmurt average)
Udmurt:udmurd8,0.106994,-0.040621,0.079572,0.059755,-0.0397,0.007251,0.00846,0.008769,-0.007772,-0.032438,0.024521,-0.008393,0.012785,-0.034543,-0.00855,-0.006629,-0.009127,-0.002154,-0.014204,-0.012756,0.003743,0.006925,0.000493,0.011086,-0.006347
I really greatly appreciated your generosity and patience in answering my questions! Thank you very much!
@ Davidski
Moderns, HO (~3300 SNPs IIRC)
https://pastebin.com/eQ4q2WiX
Lowest f2:
0.130326749989335 Italian_South
0.130375340296211 Kurd
0.130487423395754 Spanish_North
0.130504418248622 Moldavian
0.130583129121570 Italian_North
0.130657147669688 Iranian_Bandari
0.130670593465457 Finnish
0.130694496601790 Greek
0.130896160594226 Tabasaran
0.130955720236102 Albanian
0.131022208995646 Ukrainian
0.131081563709389 Abkhasian
0.131083905643475 Yemeni_Highlands_Raymah
0.131222647496161 Russian_Archangelsk_Krasnoborsky
0.131250269089922 Jew_Turkish
More or less it matches the Ozera_o-like position on the WE PCA, just way "below" the plane, akin to other samples with comparable age.
@Davidski, got to be honest, my R skills are still pretty limited and I haven't actually gone beyond simply running admixtools on the Human Origins files without adding any other resources into them - I'll keep it in mind to have a look into though. It is a good idea to look at the f2 though, as that's more like a direct distance? It might be inflated generally but probably not too much inflated for any specific part of modern West Eurasia.
@Ramber, Part 1:
“Udmurts seem to have a quite significant founder effect in this data”, yeah, I'm just saying they look to have a bit more genetic drift on top of what G25 would show. Not genetic drift that makes them relatively closer to any other population, just extra genetic drift. That's what I mean by founder effect. Though I might be using that term sloppily, I'm still an amateur on genetics - https://en.wikipedia.org/wiki/Founder_effect - and technically it could be that it's better to call these "bottlenecks" (if its just one population descreasing in size and not a population splitting then my understanding is "bottleneck" is more technically correct). More genetic drift particular to the population anyway. It's just my guess but since they basically look to be in the same relative positions as the G25 but + extra distance, I think that's what's the case. For Udmurt specifically it isn't super heavy like Kalash or something (where they have a big extra bottleneck that isn't captured by G25 because its designed not to) but it still looks quite significant.
Re; "Russian"m I've just used the standard set that the Human Origins panel has lumped together as "Russian" and a finer scale set might behave differently - it's sort of the problem of me getting the data from that and Davidski's G25 coming from his personal file. I don't totally know who's in that and may it includes some people who are minorities from the Volga region or something. (Ideally both the PCA data and F2 data would come from the same source but it's not always possible as he probably doesn't have time to run these off).
Re; "Overall direct distance is much better at telling the overall story than PCA projection?", yes. But caveat is, I think it's also like, what do you want to know? If you want to know, just how exactly different in average / expected genetic allele frequencies the two populations are, the distance will tell you that directly. But as the distance on the PCA projection for a PCA that doesn't take private genetic drift to the population into account then alternatively that will kind of tell you another kind story that is more like "How different would these populations be, disregarding their own particular extra genetic drift?". And that might be what you want to get an idea of. Plus, for ancient dna there is some questions about if ancient dna damage affects stats I guess, so to some degree looking at the direct f2 distance (rather than an f3 distance or a projected PCA distance) might be confounded by damage or missingness and things like this.
@Ramber, part 2:
Re; adding extra populations to the Udmurt comparison, I don't think
I'd have time for that any time soon, sorry. Most of those populations don't seem to be in the v44.3 Human Origins (so I surmise Davidski has integrated them from other studies), so I couldn't without chasing down those etc anyways.
Re; running for extra populations and even individuals, well it's possible, though I think the f2 distances benefit from having at least a couple of individuals (as they would be inflated for an individual from what I can tell). Again, sorry, I don't really have the time right now.
Re; what the other user told you, I think the G25 seems to reflect real population differentiation pretty well, provided we look at the squared euclidean distance (and not the simple euclidean distance, though that tells them same story in terms of rank order). That squared distance seems to relate more to distances like Fst and F2. I think that user might have looked at how G25 raw/simple euclidean distances seemed overly high between local populations compared to the Fst, and then drawn a correct conclusion that euclidean distance on Global 25 is overestimating distances between close populations... But that person's solution and theory is not really necessary (and may not make sense, I couldn't decipher it) and actually its just the problem that is more general to euclidean distance on PCA not matching to Fst / F2 distance as well as squared euclidean distance, which is easily solved by squaring the distance. They seem to be thinking there's something inherently wrong with PCA for estimating distances, but as the above paper from Benjamin Peter shows for f2, actually this is not the case. (Florian Prive also showed the good relationship of squared PCA euclidean distance to Fst in another paper from last year- https://www.biorxiv.org/content/10.1101/2020.10.06.328203v2.full.pdf)
(If it's the user/poster I'm thinking of, then what he proposes with using Dodecad taking the component Fst scores into account would be possible... But from my experience the method that's required to make that work would be more complex than simply matrix multiplication. We'd have to transform the component Fst matrix into dimensional data using MDS methods, *then* do matrix multiplication with populations' component levels, *then* detransform back into squared euclidean distances on that. It's a lot of effort when just squaring the G25 distance gives you something highly correlated with the real f2 and fst distances, and going through all the effort wouldn't actually give you anything better than G25 gives you, and quite possibly lower resolution.)
These questions are pretty complicated so I might be pushing my limit of explanation here, but this is all what seems to be the case to me.
@ Skirbha
“ mean the founding lie which Curta perfectly formulated at the level of archeology and linguistics, see the alleged formation of the Proto-Slavic in the 8th century as alleged ‘a lingua franca of Avar kaganate’, etc.”
I wouldn’t discredit the theory in total
The first phase is creolisation as various Balto-Slavic warriors moved toward the carpathian basin during the avar period
Then Slavic becomes they established as a LF
Then Slavo-phone duchies are established exactly on the periphery of the Avar empire (Moravia, Pannonia, South Poland).
This explains the Baltic toponymic substrate in the Kiev culture zone- it was not the Common Slavic “homeland”
On another topic, as mention / warned previously, I was gonna compare outgroup f3 stats computed on an f2, with computing them on the G25 euclidean squared distance, for ancients.
Here are some results for linear plots: https://imgur.com/a/UYIWoIW
In general there's a *very* high correlation (0.99) between them. Very close.
However the slight differences do also have the ancient populations breaking the regression lines. It seems like reflect that ancients are further from modern people and to some extent closer to each other when looking in the f2 rather than G25 squared distance.
This could be for multiple reasons - I'd have to guess its because the PCA projection method in the PCA algorithm used to compute G25 is very, very good, but not quite perfect? It could also be due to some characteristics in the ancient dna vs modern dna in some other sense (and so G25 is more accurate) but I think that feels less likely to me at least.
One other way to look at the outgroup f3 / outgroup pseudo-f3 stats computed using PCA distance (using Peter's term) is to do what I always do with any data and bung em into another PCA.
So G25 pseudo-f3 first: https://imgur.com/a/RmuW4qR
Recreates a lot of familar patterns that I don't need to talk about!
For comparison, real f3 : https://imgur.com/a/Mht15oV
Although they're very similar, I get the impression that under the real f3, the ancients "contain" modern populations more, and the dimensions seem to split in ways that are cleaner and more orthogonal. (e.g. in the last of the real f3 outgroup PCA, the Spain_MLN has a very clean dimension where only Western European and mainly SW European populations "break" towards it, which is very logical. And in particular there isn't really an "anti-polarity" associated with Latvia_BA as in the G25 outgroup pseudo-f3s! Which is again probably more logical than some kind of anti-correlation with Latvia_BA)
@Matt
It is a good idea to look at the f2 though, as that's more like a direct distance?
Yep, more comparable to G25 distances and Identical-by-State (IBS) similarity.
@Arza
Lowest f2:
0.130326749989335 Italian_South
0.130375340296211 Kurd
0.130487423395754 Spanish_North
0.130504418248622 Moldavian
0.130583129121570 Italian_North
Thanks, this makes sense.
So SAT29 ain't no Northern or Western European.
@ Rob
“ mean the founding lie which Curta perfectly formulated at the level of archeology and linguistics, see the alleged formation of the Proto-Slavic in the 8th century as alleged ‘a lingua franca of Avar kaganate’, etc.”
I wouldn’t discredit the theory in total
The first phase is creolisation as various Balto-Slavic warriors moved toward the carpathian basin during the avar period
Then Slavic becomes they established as a LF
Then Slavo-phone duchies are established exactly on the periphery of the Avar empire (Moravia, Pannonia, South Poland).
This explains the Baltic toponymic substrate in the Kiev culture zone- it was not the Common Slavic “homeland”
Thank you very much for this comment. Note that I pay special attention to the 'alleged formation of the Proto-Slavic in the 8th century as alleged' a lingua franca of Avar kaganate''.
Can you please focus on this point and explain what you meant by writing this:
‘The first phase is creolisation as various Balto-Slavic warriors moved toward the carpathian basin during the avar period
Then Slavic becomes they established as a LF’
@Arza
Could you process TAF011-TAF014 to bam files, from which I would substract some ANA sequences. https://www.ebi.ac.uk/ena/browser/view/PRJNA422662
@all
an R1a sample has been found in the eastern baltic with a genome wide ancestry all on the WHG/EHG side. Timing: 3000 BC.
Teepean47 on Anthrogenica has done the genotypes
https://anthrogenica.com/showthread.php?24220-A-5-000-year-old-hunter-gatherer-already-plagued-by-Yersinia-pestis
But what's the big deal here? It's just Latvia HG with R1a instead of R1b.
From the looks of it it's a Latvia MN type individual with a clade shared by PES001 and other EHGs. Actually it is from the Eneolithic and is the youngest of such people so far.
Latvia MN aka new wave of hunter gatherers from Russia. This is for sure where the R1a is from originally.
In itself it is not a big deal but it is significant that we do not find an R1a sample in the northern caucasus or in the Volga. If anything it is another confirmation that PIE emerged from the mixed HG WHG/EHG that populated the region between the eastern baltic and the north pontic area. Obviously the most likely candidate at this point are the Dneper-Don foragers. The Western Steppe Herder ( aka the Yamnaya) were a mix of Ukraine neolithic (20%) and Progress like ancestry ( around 80%). So they were IE speakers for the first component not the second. Paradoxically there was no steppe ancestry among PIE. Again the greek paper confirms that.
In fact, there's R1a-M459 in Khvalynsk and Steppe Maykop.
So, broadly speaking, it's the same line as in this new Latvia MN sample.
But the Indo-European R1a is obviously R1a-M417, and you won't find that in the East Baltic until Corded Ware spreads there from the steppe.
@old europe “ In itself it is not a big deal but it is significant that we do not find an R1a sample in the northern caucasus or in the Volga. If anything it is another confirmation that PIE emerged from the mixed HG WHG/EHG that populated the region between the eastern baltic and the north pontic area. Obviously the most likely candidate at this point are the Dneper-Don foragers. The Western Steppe Herder ( aka the Yamnaya) were a mix of Ukraine neolithic (20%) and Progress like ancestry ( around 80%). So they were IE speakers for the first component not the second. Paradoxically there was no steppe ancestry among PIE. Again the greek paper confirms that.”
Yes, I agree with most of what you are saying, except for why would a 20% component impose its language on the other 4/5?
I think that PIE was mostly Dnieper-Donetsk language with a strong GAC and to a lesser extent Tripolye and Narva HG linguistic influences.
@Davidski “ In fact, there's R1a-M459 in Khvalynsk and Steppe Maykop.”
My guess is that these R1a samples migrated from the Don area in Ukraine.
Do we have a confirmation that Khvalynsk or Progress speak anything resembling IE? If not, then it’s corroborating PIE’s putative origins as a Western Ukraine autochthonous language rather than anything “EHG” or “CHG”?
@ Skribha
''Can you please focus on this point and explain what you meant by writing this:
‘The first phase is creolisation as various Balto-Slavic warriors moved toward the carpathian basin during the avar period
Then Slavic becomes they established as a LF’''
It means that various individuals & groups from various places in northeastern Europe were settling in the Carpathian basin (but also parts of the Lower Danube). Given they spoke related but not identical languages, they would have first underwent koinization (sorry, NOT Kreolization; word slip there).
@ old europe
''So they were IE speakers for the first component not the second. Paradoxically there was no steppe ancestry among PIE. Again the greek paper confirms that.''
that's due to the odd way they presented their ADMIXTURE analysis
"In itself it is not a big deal but it is significant that we do not find an R1a sample in the northern caucasus or in the Volga."
This region is highly undersampled
Rob, Babik says that linguistic disintegration, which consists in the accumulation of innovations in small areas, is a long-term process. Therefore, it is not possible for the disintegration of the Baltic and Slavic branches to take place in the short Avar episode.
Babik also considers the Slavic language as a koine formed on a linguistic basis with a close degree of mutual kinship, but rejects the Avars as the driving force of such linguistic integration, due to the much wider range of Slavic in those times (going far beyond the reach of direct Avar influences) and the presumption of this integration before the period of Slavic migrations.
@ Amrbon
Sorry who's Babik and why should his view be Gospel ?
Before the Avars arrived, those parts of East Central Europe which still enjoyed some semblance of population were 'Germanic', clearly linked with Lombards & Thuringians, Gepids, etc. It is only after ~ 600 that power shifts and population movements enabled Slavs to take their place.
Of course, Slavic reached regions where Avars never truly controlled, but that's later, once relevant power centres from secondary expansion could occur had been established. This is clear in the early Slavic burials in Moravia, Pannonia & Dalmatia, where burials were obviously Europeoid but military elites had Avar type weapons and ornaments. It shows that Slavs could become regional elites within larger power structures (Avars, Franks, Byzantium), liable to proclaiming independence should they find it better to their circumstances.
@ Norfern-Ostrobothnian
You can download original ones:
https://pastebin.com/aux5fYg2
@Arza, Norfern et al.
Do we have these Medieval Lubeck samples?
https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3762113
No, I don't think we do.
@Matt,
Sorry I’m still confused. Can you explain more on this part, “they look to have a bit more genetic drift on top of what G25 would show. Not genetic drift that makes them relatively closer to any other population, just extra genetic drift.”? You are still an amateur in genetics? Wow you are very knowledgeable and well-versed in the subject tbh. Agreed, maybe bottleneck is a better term. I search founder effect on wikipedia and it says “the loss of genetic variation that occurs when a new population is established by a very small number of individuals from a larger population. As a result of the loss of genetic variation, the new population may be distinctively different, both genotypically and phenotypically, from the parent population from which it is derived. In extreme cases, the founder effect is thought to lead to the speciation and subsequent evolution of new species”. Interesting definition. So Udmurts were the new population who branched off from a larger population?
I see. So I guess the Human Origins didn’t label where specifically the “Russian” sample was from in the nation?
Would you say that taking genetic drift into account is more important than just comparing and contrasting between average/expected alleles of two populations? Oh I didn’t know that ancient DNA damage can also affect stats. In this case would an f3 tell the story better if the ancient DNA is damaged?
Since you don’t really have the time, Is it possible for me to add those extra populations into the Udmurt data by myself? Or do you know anyone else who could help adding these populations?
It’s interesting because this is what another member in the same forum ( not the one that told me about the G25 not being effective and that Dodecad K12b/Eurogenes K13 is better for genetic distance) did using the Harvard 1240K+HO dataset: https://reich.hms.harvard.edu/allen-ancient-dna-resource-aadr-downloadable-genotypes-present-day-and-ancient-dna-data.
That user created a heatmap based on the FST distance of Udmurts to various populations from that dataset. It seems to fairly correspond to the data for the Udmurts that you created and show Udmurts are also genetically closer to Nogai_Karachay_Cherkessia, Nogai_Stavropol along with Bashkir, Siberian Tatar, Uzbek, Tajik, Forest Yukaghir and Turkmen than to most Europeans.
These information seems rather complicated for me. Regardless, thank you very much for taking your time to answer me. It will take quite a while to fully grasp all of the concepts that you wrote as these seem really advanced for the a layman like me.
Btw you are very knowledgeable and have a lot of insights about this subject, do you happen to be a mathematician/statistician? Just asking out of curiosity. Because population genetics seem to require a lot of statistics to do all these formal stat tests.
Rob, professor Zbigniew Babik is a contemporary linguist, specialist in Slavic toponomastic.
@Matt
Is it ok if I used the data you made on the Udmurts on some discussions on Anthrogenica?
@Ramber, yeah sure, feel free to reference any of the posts, no problem with me.
@Rob
It means that various individuals & groups from various places in northeastern Europe were settling in the Carpathian basin (but also parts of the Lower Danube). Given they spoke related but not identical languages, they would have first underwent koinization (sorry, NOT Kreolization; word slip there).
Thank you for clarifying what you wrote above. I propose to establish a discrepancy protocol.
1.
Can you describe your version of the origin (place, time, etc,) of the languages, :
- PIE,
- Indian (Vedic Sanskrit)
- Iranian, (Avestan)
- Anatolian,
- Hellenic,
- Armenian,
- Celto-Italian,
- Germanic,
- Baltic,
- Slavic?
2.
It is unknown who the Avars were, but they could logically be a mixture of Turkic tribes, see:
https://en.wikipedia.org/wiki/Pannonian_Avars
(…) Several historians, including Peter Benjamin Golden, suggest that the Avars are of Turkic origin, likely from the Oghur branch.[30] Another theory suggests that some of the Avars were of Tungusic origin.[31] A study by Emil HerÅ¡ak and Ana Silić suggests that the Avars were of heterogeneous origin, including mostly Turkic (Oghuric) and Mongolic groups. Later in Europe some Germanic and Slavic groups were assimilated into the Avars. They concluded that their exact origin is unknown but state that it is likely that the Avars were originally mainly composed of Turkic (Oghuric) tribes.[32] (...)
Do you claim that the Avars themselves spoke Turkish like their victorious Turkish ancestors, or that they changed their language (like the Normans from Norwegian to French) and spoke Slavic like the peoples they conquered,.. because it was easier and more convenient for them?
I remind all that the word khagan itself has unchallenged Slavic / IE origins, see:
https://eurogenes.blogspot.com/2021/04/the-history-of-scythians-gnecchi.html?showComment=1622294665251#c1424616746798001699
(…) By the way, there are words in the Polish language such as Kagan / KaGaN, Kaganiec / KaGaN+ieC and Kaganek / KaGaN+eK, see:
Kaganiec / KaGaN+ieC
https://en.wiktionary.org/wiki/kaganiec
Kaganek / KaGaN+eK
https://en.wiktionary.org/wiki/kaganek
When you drop suffixes and focus on the root od the word, you could understand the true meaning of this:
https://en.wikipedia.org/wiki/Khagan
(…) The term is of unknown origin and possibly a loanword from the Ruanruan language. Pulleyblank (1962) first suggested that a Xiongnu title, transcribed as è·äºŽ (Old Chinese: *hÊ·aÊ”-hÊ·aÊ°) might have been behind Proto-Turkic *qaÉ£an ~ *xaÉ£an. (…)
@ Ambron
Well you & Babick need to understand that Baltic-Slavic did not “suddenly disintegrate” between 600& 800 AD
I Baltic-Slavic dialects were already variable; which is the very reason that koinization was required.
@Rob
3.
Koine is a written language created by mixing different dialects of the same language, see:
https://en.wikipedia.org/wiki/Koine_Greek
(…) Koine, also known as Alexandrian dialect, common Attic, Hellenistic or Biblical Greek, was the common supra-regional form of Greek spoken and written during the Hellenistic period, the Roman Empire and the early Byzantine Empire. It evolved from the spread of Greek following the conquests of Alexander the Great in the fourth century BC, and served as the lingua franca of much of the Mediterranean region and the Middle East during the following centuries.
(…) Literary Koine was the medium of much of post-classical Greek literary and scholarly writing, such as the works of Plutarch and Polybius.[5] Koine is also the language of the Christian New Testament, of the Septuagint (the 3rd-century BC Greek translation of the Hebrew Bible), and of most early Christian theological writing by the Church Fathers. (...)
4.
Definition of the koineisation:
https://www.thoughtco.com/koineization-dialect-mixing-1691093
In sociolinguistics, koineization is the process by which a new variety of a language emerges from the mixing, leveling, and simplifying of different dialects. Also known as dialect mixing and structural nativization.
(…)
In some cases, the koine is a regional lingua franca which does not replace the already existing dialects."
(…)
"As Hock and Joseph (1996:387,423) point out, koineization, the convergence between languages, and pidginization usually involve structural simplification as well as the development of an interlanguage. Siegel (2001) argues that (a) pidginization and koineization both involve second language learning, transfer, mixing and leveling; and (b) the difference between pidginization and creole genesis, on the one hand, and koineisation, on the other, are due to differences in the values of a small number of language-related, social, and demographic variables. Koineisation is usually a gradual, continuous process which takes place over a long period of sustained contact; whereas pidginization and creolization are traditionally thought of as relatively rapid and sudden processes."
(...)
"[T]he social contexts of the two processes differ. Koineization requires free social interaction between speakers of the various varieties in contact, whereas pidginization results from restricted social interaction. Another difference is the time factor. Pidginization is most often considered a rapid process in response to a need for immediate and practical communication. In contrast, koineization is usually a process which occurs during prolonged contact between speakers who can almost always understand each other to some extent."
(…)
How could the koineisation of the Slavic language come about if the Avars spoke Turkish and the Proto-Slavs / Slavs did not use the script, like the Aryans on Z+Altay, etc.?
5.
Definition of lingua franca:
https://en.wikipedia.org/wiki/Lingua_franca
(…) The term "lingua franca" derives from Mediterranean Lingua Franca (also known as Sabir), the pidgin language that people around the Levant and the eastern Mediterranean Sea used as the main language of commerce and diplomacy from late medieval times to the 18th century, most notably during the Renaissance era.[15][8] During that period, a simplified version of mainly Italian in the eastern and Spanish in the western Mediterranean that incorporated many loan words from Greek, the Slavic languages, Arabic, and Turkish came to be widely used as the "lingua franca" of the region, although some scholars claim that the Mediterranean Lingua Franca was just poorly used Italian.[13] (…)
@Rob
6.
Curta's claims:
https://www.academia.edu/43351673/Slavs_in_the_Making._History_Linguistics_and_Archaeology_in_Eastern_Europe_ca._500_ca._700_
Slavs in the Making takes a fresh look at archaeological evidence from parts of Slavic-speaking Europe north of the Lower Danube, including the present-day territories of the Czech Republic, Slovakia, Poland, Belarus, Ukraine, and Russia. Nothing is known about what the inhabitants of those remote lands called themselves during the sixth century, or whether they spoke a Slavic language. (…)
https://indo-european.eu/2020/06/slavs-in-the-making-history-linguistics-and-archaeology/
(…) Language
The heavy influence of Germanic on Slavic is almost exclusively lexical, which has been interpreted as an indication of unequal relations, with speakers of Slavic imitating the supposedly prestigious language patterns of speakers of Germanic. Similar sociopolitical conditions seem to have existed at the moment and in the place of Altaic–Slavic contact, as scholars assume a pidgin was created by (or for) speakers of Altaic languages to communicate with speakers of Slavic, who in turn strove to imitate the speech habits of their masters.
A very different explanation is needed therefore for the spread of the koiné within the Carpathian Basin, and beyond. Slavic reached East Central Europe as a consequence of migration, most likely that of the Avars. Various groups of population from Eastern Europe, most likely from Right-Bank Ukraine, accompanied the Avar movement from the steppe lands north of the Black Sea into the Middle Danube region (see Chapter 4). Some of them may well have been speakers of (the) Slavic (koiné). The evidence of language contact strongly suggests that speakers of that language were in a position that was socially inferior to that of speakers of both Germanic and Altaic languages. The archaeological evidence, on the other hand, suggests that groups such as Lazuri-Pişcolt in northwestern Transylvania were intrusive, and possibly the result of forced movements of population from other areas controlled by the Avars.
By contrast, there may have been a short-distance migration from the northern parts of the Carpathian Basin to southwestern Slovakia and to Moravia. Whether or not the migrants were speakers of Slavic, the local population that the Avars found in place may have shifted to that koiné, which would have thus effectively become a lingua franca inside the Avar qaganate, as the Avar elites may have (occasionally) used Slavic in order to communicate with their subjects.
Language shift may have also taken place in central Bohemia as well as in the lands of southeastern and eastern Germany, although forced movements of populations under Avar rule cannot be excluded. Such developments took place most likely after ad 600, and reached more peripheral areas only in the late 7th or early 8th century. Assuming that the population that established the supposedly Prague-culture settlements in the Upper Vistula and Upper San region spoke the Slavic koiné, the spread of that language throughout Poland was a slow process that began in the second half of the 7th century and continued long after ad 700.
A parallel process may have brought speakers of the koiné across the Middle Dnieper river into Left-Bank Ukraine, as well as deep into the forest belt (the Polesie region) of northern Ukraine and southern Belarus. It is difficult, if not impossible, to tell whether it was language shift or migration (or both) that was responsible for the linguistic spread in all those regions and what the duration of the process may have been. This scenario is of course based on the idea of koiné, which is one of many possible explanations for the rise and spread of the language called (Common) Slavic. It is, in other words, little more than a hypothesis. However, when confronted with the archaeological data, that scenario seems more likely than others, at least to me. (...)
@Rob
7.
Here are some comments on Curta's claims:
https://linguistics.stackexchange.com/questions/35412/could-have-inflected-proto-slavic-really-been-created-as-a-lingua-franca-among
AmazingWouldBeGreatBut Mar 6 '20 at 11:53
My point is that even nowadays people formally studying Slavic inflection in the process of a second language learning fail to master it. Usually pidgins tend to simplify things. I can't see such a complex language such as Slavic, without literary tradition, not only not simplifying, but keeping the complexity among Turkic peoples.
Atamiri edited Mar 7 '20 at 19:10 answered Mar 7 '20 at 1:43
Curta’s hypothesis sounds a little far-fetched but there’s a more plausible theory that a form of Slavic spread very rapidly — maybe with the Avars — across the Balkans, effectively smoothing out much of the dialectal differences that must have existed at that time. This theory assumes that Slavic was indeed used as a lingua franca but had been adopted rather than created. Nicolaos Trunte gives a historic account in his book on Old Church Slavonic.
@ Skribha
''Can you describe your version of the origin (place, time, etc,) of the languages, :
- PIE,
- Indian (Vedic Sanskrit)
- Iranian, (Avestan)
- Anatolian,
- Hellenic,
- Armenian,
- Celto-Italian,
- Germanic,
- Baltic,
- Slavic?
''
Let's not waster time with irrelevancies
2.''It is unknown who the Avars were, but they could logically be a mixture of Turkic tribes, see:''
No shit , Sherlock
3.,''Koine is a written language created by mixing different dialects of the same language, see:''
Or similar languages, like Balto-Slavic dialects, which is what I said, had you managed to understand what I wrote
4 ''Definition of lingua franca:''
(…) The term "lingua franca" derives from Mediterranean Lingua Franca (also known as Sabir), the pidgin language that people around the Levant and the eastern Mediterranean Sea used as the main language of commerce and diplomacy from''
That's an irrelevant definition.
Latin was LF of the Roman Empire, and it was certainly not a pidgin. Most historical LF have been 'natural languages, not pidgins
Pidgins are rare outside modern colonialist scenarios
6. ''How could the koineisation of the Slavic language come about if the Avars spoke Turkish and the Proto-Slavs / Slavs did not use the script, like the Aryans on Z+Altay, etc.?''
LOL are you still going on about Z+Altay ? yawn
Slavic became the koine of the majority of the population, which was European, and by 700 AD, eastern European.
Even the core group of 'Asian' Avars could have learned it, but obviously Turkic wasnt relevant in the koinization of Slavic, apart from a few loanwords
This isn't rocket science
..7.
nevermind your thoughts on Curta. I made my own deductions
But I will say, your understanding is pretty basic
Rob, professor Babik, is unlikely to take your advice, as he is one of the most outstanding contemporary linguists. However, I see that you completely do not understand the linguistic processes.
The Slavic and Baltic innovation centers were far away from each other, therefore today's Slavic and Baltic languages are completely incomprehensible to each other. Of course, there were undefined Balto-Slavic dialects at the borders of the isoglosses, which is clearly visible, for example, in the toponymy of Mazovia. However, such dialects could not be the basis of the Slavic koine, because there was already a Slavic dialect group at that time.
@ Ambron
That’s ok; I’m just bouncing ideas around. I don’t pretend to possess all the answers, but I do pay attention to details instead of preconceived ideas
Another quick experiment using the real f3 outgroup stats and pseudo-f3 outgroup stats derived from squared G25 distances, using them in Vahaduo for what we used to call "f3 nMonte".
First with the real f3 scores: https://imgur.com/a/j7SIm6t
Then with the pseudo-f3 scores (or "PCA distance f3"): https://imgur.com/a/n8hckC5
Qualitatively the results of both are similar and make pretty good sense. The differences are:
- f3s calculated from f2 distance seem to result in more Israel_C related ancestry estimates among present day populations. That suggests more of a wide spread of "East Med" ancestry that Israel_C represents. This is no doubt enriched by Israel_C being an unusual population with extra Barcin ancestry compared to later BA and IA populations, but still seems quite strong.
Under the "PCA distance" f3, only a few populations in Southern European take Israel_C.
- f3s calculated from f2 distance seem to put more "Latvia BA" into Lithuanians (50% vs 80%) and Finns, and there is some odd effect of Latvia BA going into Sicilian and Ashkenazi populations which is suspect. But general Latvia BA in Belarusians / Ukrainians is more similar (40-30%). Latvia BA might be slightly compressed towards Lithuanians or something in G25 possibly.
- Also in real f3 some low level China_NE in more populations (approximately 1%), falling off on distance from Western Europe, while in the f3 based on G25 there's none of this. Which seems plausible enough, presuming this just covers "ENA" related ancestry generally, though could also relate to some differences between ancient and modern dna somehow.
Not going back to saying the "f3 nMonte" method is better, or that these produce realistic models, more that this is just one way to look at how differences between f2 and squared PCA distances might affect things.
I'm seeing Slavic users writing on "Indo-Germanic theory" again. I said it once, but I have to repeat myself, so let's put this straight, once and forever: this is merely based on a misunderstanding of the German word "indogermanisch". In fact, there is no Indo-Germanic theory, there never was. In the German language, there are simply two words for "Indo-European": "indogermanisch" and "indoeuropäisch". They are perfectly synonymic and exchangable. They denote 100% the same. "Indogermanisch" doesn't imply any specific theory about IE origins, spread or phylogeny. The only correct English translation of "indogermanisch" is "Indo-European". "Indo-Germanic" is a false translation. If you don't believe me check a dictionary. The term "indogermanisch" was coined in order to name the IE language family after its known westernmost and easternmost languages, which at that time were Icelandic (Germanic) and Sanskrit (Indic). Tocharian hadn't been discovered yet. This labelling for sure is not perfect, but neither is "Indo-European", because not all European languages are IE, nor is Indic the only Asian IE language.
Rob, sure! Everyone here shares common interests, we all have some ideas and we exchange them with each other.
@Davidski
Can you process this genotype data?
https://www.mediafire.com/file/j0rgtvx52zocx4h/TAF014_A.zip/file
@Rob
I would argue that French, at least, was semi-creolized, at least phonologically. In French, for example, the Latin advocatus becomes avoué (while in comparison, the Spanish descendant of advocatus is abogado), and this continues even with the partially Viking Norman invaders of England, who when they began to learn the language of the natives began omitting consonant clusters not found in their native tongue, the reason yacht rhymes with pot (at least in my dialect), it has been shorn of its Germanic ch sound, which is retained in the nearby Dutch language.
"Our results also showed that the disease victims were people of northern and eastern European descent, confirming that Hanseatic Lübeck was an important trading centre for the Baltic region."
A stupid conclusion. Lübeck was founded by Slavs and later settled by North Germans. This explains northern and eastern Euopean descent of the local population much more parsimoniously than Hanseatic trade contacts across the Baltic sea.
Yes, why the sudden expansion so far and wide ? And why did all Steppe migrants not expand towards the Danube ? Seems like an explosion in the Henhouse. A desperate search for something that only very faraway places could satisfy ?
@Matt,
Thank you very much for your work, time and patience!
Since you don’t really have the time to add more populations, Is it possible for me to add those extra pops into the Udmurt data by myself? Or do you know anyone else who could help adding these populations into the plots you made?
Whether a German speaker says "indogermanisch" or "indoeuropäisch" is a matter of taste and style. I personally prefer "indogermanisch", because I feel that "indoeuropäisch" is an unnecessary, exaggerated assimilation to other languages. And I'm doing this without remorse because "indogermanisch" wasn't the expression preferred by the Nazis. They preferred to call it "arisch" after all.
BTW, I got the second dose of the Moderna vaccine last Thursday; I felt medium strong side effects about 15 hours later. I went to sleep and afterwards the worst was already over. I've read about stronger side effects in several cases. And my gf felt like being infected by the flu for 1.5 days. But this all is in no comparison to experiencing long-term damage due to long-covid, or to getting hospitalised because of it, which sometimes happens to younger people too. Of course I've read about the conspiracy theory that the covid vaccine will kill you within 3 years at the most. But this is BS with no evidence supporting it. It's easy to claim stuff, but quite a different thing to prove it.
@Simon_W,
I did not know Germans sometimes call it Indo-German. This is interesting inside information. You can't deny that does give Germanic more central focus than it deserves. It might even give the impression it is a langauge family of only Indian and Germanic.
Then again, why do we include the name Indo in there? It gives India too much focus in IE studies.
I personally wish it was called Euro-Asian. Like how there is the Afro-Asiatic language family. Euro-Asian would be a good counter to it.
If the Iranian languages in northwest Asia didn't disappear, I think the name Euro-Asian would have been chosen.
@ Vinitharya
That is called language shift with interference. i.e. the native phonological characteristics persist during mass adaptation of a new language.
A creole derives from a pidgin which has undergone complexification. naturally, French isnt a creole even in the loosest of sense
Historically, the term "indogermanisch" comes from the fact that the Indians and Germans languages lie at opposite ends of the Indo-European language area.
Not only Lübeck was Slavic and has a Slavic name, but also the name of the entire Hanseatic League comes from the Slavic Vikings - Chąśniks, and the most important Hansanian quarter was the Slavic quarter - Wendish.
Well, East Germans is afraid to admit that they are Germanized Slavs.
@Ramber, if you wanted to get more populations on, I'd write up some notes about how to get the software and run off the stats. You'd need to run off everything as a big batch though rather than just running off different population pair comparisons and adding them in, as the stat is affected by that.
@Norfern
TAF014_scaled,-0.330087,0.060932,-0.063356,-0.117896,0.021235,-0.071117,-0.119855,0.017538,0.203706,-0.008565,0.022734,-0.040764,0.105549,-0.066747,0.097447,-0.039644,0.015385,-0.098817,-0.191438,0.0529,-0.066009,-0.173856,0.109321,0.003133,0.020477
TAF014,-0.029,0.006,-0.0168,-0.0365,0.0069,-0.0255,-0.051,0.0076,0.0996,-0.0047,0.014,-0.0272,0.071,-0.0485,0.0718,-0.0299,0.0118,-0.078,-0.1523,0.0423,-0.0529,-0.1406,0.0887,0.0026,0.0171
So what is this, an ANA ghost?
Looks like a weird version of Taforalt TAF014.
Pretty good, although not ideal. If I'm not wrong, ANA would be even more African shifted than this. I guess I ought to use more reference samples to filter out any true Eurasian input.
Target: MAR_Taforalt:TAF014
Distance: 2.4025% / 0.02402495
72.2 TAF014_scaled
9.6 TUR_Barcin_N
9.4 ITA_Villabruna
6.6 TUR_Pinarbasi_HG
2.2 RUS_Kostenki14
And yeah this is closer to Iberomaurusian's African ancestry.
https://ibb.co/X2VpCqM
@Matt,
Yes I'm very interested in adding more populations on to the Udmurt data. Can you teach me how to get the software and run the stats?
OK Ramber, here's a quick guide - https://pastebin.com/WzXHsKh8
If it's lacking in any detail, come back to me or the Anthrogenica R \ admixtools mob and between us I expect we could solve any problems you have running data.
@Matt
Thanks for the link of the SMBE 2021 video talks. Did you have a take on Isabel Alves presentation about Population structure of Brittany(...)?
@Matt
Thank you very much for your instructions! I will have a read at first.
@JS, thanks, nothing I don't reckon you wouldn't have already thought of from the presentation. Happy to hear if you've got any ideas you want to run by us tho!
@Ramber, btw, I think you mentioned upthread using the Dodecad K12b to capture differentiation between populations, and I mentioned how I would do this.
Well, this is how I would do it, in case you want to make a reference to this in discussions with anyone (or in case anyone was interested in this bit of the conversation):
1) Transform components the components from Fst distance matrix into a set of dimensional data using the Principal Coordinates Analysis function in PAST. Images show how this is done: https://imgur.com/a/ATQy6K2 (the settings must be exactly the same). This is the output: https://pastebin.com/Mjs8aywK
2) Then multiply the components in each population by the dimensions output from 1). Output: https://pastebin.com/ux4ddipW
3) This information then gives us a nice euclidean distance neighbour joining tree that is similar to what we would get out of Global 25 or any well run PCA: https://imgur.com/a/c6oDVKC. (And we could use Vahaduo on it if we wished as well, or whatever).
4) However, to get the Fst out of it, what we then do is, you run the euclidean distance on the set and then square the output, which turns it back into Fst. Example method: https://imgur.com/a/WGe0IAF. Pastebin of output: https://pastebin.com/nLs7BtdH. (Some examples of some selected outputs, although unfortunately no Udmurts in the datasheet I hold: https://i.imgur.com/zaGBt3k.png).
The reasoning of why I would use the dimensional transformation step is here: https://imgur.com/a/nA0ZV7J . Basically just multiplying by the Fst matrix gets things wrong, but including the dimensional transform and then reversing it gets the right Fsts.
@Rob
1.
’Can you describe your version of the origin (place, time, etc,) of the languages: PIE, Indian (Vedic Sanskrit), Iranian, (Avestan), Anatolian, Hellenic, Armenian, Celto-Italian, Germanic, Baltic, Slavic?’
(…) Let's not waster time with irrelevancies (…)
Do whatever you want, but the truth is that your statement is logically invalid. I remind you that the subject of this post is: The PIE homeland controversy: June 2021 status report.
As it can be observed so far, you have got serious problems not only with logic and common sense, but with memory as well.
I admit that I did not follow what you write, but only after what you have already ‘gafted’ under this entry, it is more than obvious that no sane person should take you and your claims seriously.
I mean by what you wrote that you simply have no idea, how, when and where the above languages formed, etc. It does mean that what you can write is trully irrelevant, and my further ‘discussion’ with you is just waste of my precious time.
Just for the record, I am explaining this intellectual garbage of yours seen below:
2.
''It is unknown who the Avars were, but they could logically be a mixture of Turkic tribes, see:''
(…) No shit , Sherlock (…)
I understand that you confirm my above claim that ‘Avars could logically be a mixture of Turkic tribes.’
It is just a pity my loyal Watson that you did not clearly answer my question, see:
Do you claim that the Avars themselves spoke Turkish like their victorious Turkish ancestors, or that they changed their language (like the Normans from Norwegian to French) and spoke Slavic like the peoples they conquered,.. because it was easier and more convenient for them?
3.
''Koine is a written language created by mixing different dialects of the same language, see:''
Or similar languages, like Balto-Slavic dialects, which is what I said, had you managed to understand what I wrote
If you were at last logically consistent and careful, you would not have made such basic mistakes, and you would not have had to explain them afterwards on July 16, 2021 at 3:58 PM see:
‘Given they spoke related but not identical languages, they would have first underwent koinization (sorry, NOT Kreolization; word slip there).’
Notice that even you stated above that Baltic and Slavic languages are 'related but not identical'.
All IE languages are 'related but not identical', so one of your claims logically excludes the other, so both do not make any sense.
Well,.. unless in your opinion, e.g German and Slavic languages are also 'similar' and there is/was also some unified / koineised form somewhere, (which you can denote in your mind)... :-)
But this is just that nothing, compared to the alleged Avar / Turkish ‘unification’ or formation of the Slavic languages, which Curta logically points to!
@Rob
4.
(…) The term "lingua franca" derives from Mediterranean Lingua Franca (also known as Sabir), the pidgin language that people around the Levant and the eastern Mediterranean Sea used as the main language of commerce and diplomacy from''
That's an irrelevant definition. Latin was LF of the Roman Empire, and it was certainly not a pidgin. Most historical LF have been 'natural languages, not pidgins. Pidgins are rare outside modern colonialist scenarios
What is a relationship between what you wrote above, and the Slavic language as the alleged lingua franca of the Avar kaganate?
Of course, you omitted what is at the same time logically the most important and inconvenient for you, see:
‘During that period, a simplified version of mainly Italian in the eastern and Spanish in the western Mediterranean that incorporated many loan words from Greek, the Slavic languages, Arabic, and Turkish came to be widely used as the "lingua franca" of the region, although some scholars claim that the Mediterranean Lingua Franca was just poorly used Italian.’
5.
''How could the koineisation of the Slavic language come about if the Avars spoke Turkish and the Proto-Slavs / Slavs did not use the script, like the Aryans on Z+Altay, etc.?''
LOL are you still going on about Z+Altay ? Yawn
Nope, my dear little genius. Unfortunately, you have not grasped something again. It is both a logical comparison and another question that you have not answered. :-(
My Slavic / IE etymologies for Z+Altay have not been undermined by anyone so far, see above. I bet that you have got neither balls nor any knowledge to undermine it. :-)
(…) Slavic became the koine of the majority of the population, which was European, and by 700 AD, eastern European. Even the core group of 'Asian' Avars could have learned it, but obviously Turkic wasnt relevant in the koinization of Slavic, apart from a few loanwords. This isn't rocket science. (…)
I understand that this unification / koineisation of 'similar languages, like Balto-Slavic dialects' into Proto-Slavic as alleged 'lingua franca Avar khaganate' happened somehow by itself, somewhere in the past, but the Avars had nothing to do with it. And that is what Curta claims?!
Aha. Brafo Maestro! Brafo!
6.
nevermind your thoughts on Curta. I made my own deductions But I will say, your understanding is pretty basic
WoW / L”aL”! This is such logically profound as the highest mountain on the Sun! :-)
You and your scribbles unsupported by anything are both pathetic. It is not going to change, so as I have already stated, I am not going to waste my time on both of you in the future.
@ Skirbha
LOL i don't care about 'your etymologies', your false analogies about Mediterranean sailors, or whose 'boy' / lapp dog you are. Most of use can take on sources selectively and come to our own conclusions. Apparently that is strange to you
And I'm not going to get into a thesis on entire PIE when you cant even understand the 700s.
The premise of LFs in protohistoric Europe is nothing new, in fact the idea has existed for decades, Lunt, Holzer, Nichols have also discussed it w.r.t Slavic. The Turkic Huns spoke Gothic (indeed Atilla might be a Germanic name), so too did Slavic become the prevalent language in the Avar Empire.
This is a no brainer, and your autochthonist pseudo-linguistics don't really change that.
Ramber, also here's the same data as what I did for the Dodecad K12 for Eurogenes K15 if you are interested in that.
Pastebin: Dimensions: https://pastebin.com/WStSPYxs
Predicted Fst: https://pastebin.com/yp56MtMS
Neighbour joining tree: https://imgur.com/a/WE1Gz6V
You can see it captures a lot of structure, although the populations obviously overlap more with G25. Just including for interest that in theory any of these ADMIXTURE runs and spreadsheets can have the same thing (but generally the G25 PCA model is at least as good, particularly for ancient people who often can't fit inside ADMIXTURE components for modern people).
@Ramber, to underline, it is important to *square* the distance after the running euclidean distance on the Principal Coordinates Analysis, otherwise things will be relatively inflated and wrong.
@ Rob
LOL i don't care about 'your etymologies', your false analogies about Mediterranean sailors, or whose 'boy' / lapp dog you are. (…)
Neither do I care about your rude comments and illogical and irrelevant statements that are unsupported by anything, see above and beneath.
You cover up your lack of any arguments and sources with your childish and disrespectful behaviour.
You can have a stupid smile on your face, but everybody that wants to see the truth, can perceive what you and your credibility of statements look like.
(…) Most of use can take on sources selectively and come to our own conclusions. Apparently that is strange to you. (...)
Anyone can form an opinion based on the analysis of what we write and how we argue. Well, unless someone is as equally ‘courteous’ and ‘non-biased’ against logic and facts as you are, then he or she will 100% like what you write. :-(
(…) And I'm not going to get into a thesis on entire PIE when you cant even understand the 700s. (…)
The topic of this post is about 'The PIE homeland controversy: June 2021 status report' and only this really interests me. Everything else about IE relates to this.
I have already understood that you have nothing reasonable to write about this or anything else, so do not go even more crazy and better calm down now. Finally stop trolling and and litter this thread.
(…) The premise of LFs in protohistoric Europe is nothing new, in fact the idea has existed for decades, Lunt, Holzer, Nichols have also discussed it w.r.t Slavic. (...)
Quote from you: Sorry who's Lunt, Holzer, Nichols and why should their view be (a) Gospel?
(…) The Turkic Huns spoke Gothic (indeed Atilla might be a Germanic name), so too did Slavic become the prevalent language in the Avar Empire. (...)
So logically, because Turkic Huns (like the Turkish or Ugro-Finnish Atilla, which itself according to you might be a Germanic name), allegedly spoke Gothic / Germanic,.. that is why Slavic allegedly become the prevalent language in the Avar khaganate...
WoW / L”aL”! It is a true Masterpiece of your intellect! Applause again! :-)
(...) This is a no brainer, and your autochthonist pseudo-linguistics don't really change that. (...)
You already proved what you know and understand, so I do not care about your ‘logic’ or germanocentric-pseudo-science, etc. I bet you will not be able to 'prove' your baloney this time as well,.. by other way than by denoting it in your mind… Am I wrong? :-)
Finally, I do not think you can improve logic of your argumentations, but at least try to improve your English. It is hard to read this gibberish full of silly mistakes.
Of course there has been the theory that the PIE homeland was located in northern Europe. This was just one out of a large number of theories, and it cannot be said that this was particularly popular among German scholars after 1945...
Here's Mallory's map showing the theories on the PIE homeland published after 1960:
https://justpaste.it/657at
The last supporter of a PIE homeland in northern Europe that I know of, was Carl-Heinz Boettcher who published his theory in the year 2000. Boettcher was a studied sociologist and economist, and only a hobby archaeologist. He didn't believe in his theory for racist, neo-Nazist reasons, but because he believed modern archaeology would prove it.
As you may know, there was that immobilist, anti-migrationist paradigm in archaeology that was very fashionable until quite recently. Until the ancient DNA revolution.
The German archaeologist Alexander Häusler for instance was deeply convinced that the steppe theory can't be right, because Yamnaya didn't spread widely to the west, and where it spread, like in Hungary east of the Tisza, it didn't last, but was quickly assimilated. And he believed Corded Ware was too different from Yamnaya to have the same origin.
Considering this, and the large number of theories put forth by linguists and archaeologists, it was a magnificent breakthrough when ancient DNA came into play, especially with the pioneering works of the Reich lab. This has refuted any possibility that the PIE homeland was in the TRB in northern Europe.
And yeah, sure there had been the theory that the PIE were the blond and blue-eyed Nordic race of Scandinavia and surrounds, that they should be called "Aryans", and that the Germanic people are the purest Aryans. Obviously this has been the pet theory of the Nazis. I would call it the "Aryan" theory, or, to avoid any confusion with the real Aryans, the "Aryan theory of Nordicism". It hasn't had any academic significance after 1945. And it has been refuted by population genetics, but continues to be popular among neo-Nazis around the world, who don't read anything published after 1945.
All autochthonists are boring & wrong: ultra-Noridicists, the 'we wuz Illyrians' crowd on Anthrogenica; those with imaginary continuities in toponyms and pollen; Koch , Collis & their fanboys; so forth
@Matt
Thank you very much for your work on Dodecad K12b and Eurogenes K15. Really appreciated that.
I will let you know if I have problems with running the stats on Udmurts.
@Ramber, no problem, it was interesting to do.
As a last thing to round out the set of using this method on what I consider the 'big 3' of genome bloggers and their ADMIXTURE projects, here's a set of the same thing for Harappa Project, which was a really good ADMIXTURE as well that was deliberately less Europe focused.
Dimensionalized Harappa Project ADMIXTURE spreadsheet: https://pastebin.com/k3n3HqAd
Trees of Dimensionalized Harappa ADMIXTURE : https://imgur.com/a/2ME66cy
(There are multiple of the same population on trees because Harappa separates out populations from different papers / datasets. These could be merged of course but I didn't have time.)
(Harappa Predicted Fst sheet is too big to go on Pastebin, but the method above will show how to generate it from the the dimensionalized ADMIXTURE - simply run Euclidean distance on the set and then square it).
I still think it has some superior qualities when it comes to looking at South Asian groups and still some insights to give that aren't as available elsewhere. It didn't quite seem to do what Zack Ajmal expected in terms of finding South Asia specific components, but that's probably just due to South Asian population history and how the history is a combination of admixture from other continental groups *plus* either founder effects or admixture that remained hyper localized due to endogamy.
This exercise was also useful in making me think a bit more about some possible uses of ADMIXTURE with ancient dna - typically people don't use it because the damage and missingness with ancient dna makes components form that just separate adna from modern dna, limiting informativeness... but maybe if we take account of this using the Fst dimensionalization method, this isn't such a big issue... Will need more thought tho! So thank you for a discussion that prompted that.
@Rob
@ Skribha
That ok; your pride is wounded because I said you’re irrelevant & boring
You can add hysterical too
July 21, 2021 at 12:09 AM
Sorry for answering so late.
Well. it is exactly the opposite. You have described above yourself and the quality of what you write. You can add to this a lack of any respect for logic and people, such as comparing me to a ‘lapp dog’, etc., or the inability to write in correct English. English sentences must contain verbs and SVO, but you do not use that and prefer your garbage slang. These are your standards.
Anyone reading what you write may have already found out that it is not worth taking seriously, either you or anything you write.
Finally, you deleted this post above, just to write even more schizophrenic nonsense which is irrelevant and not related to the topic of this post, see:
All autochthonists are boring & wrong: ultra-Noridicists, the 'we wuz Illyrians' crowd on Anthrogenica; those with imaginary continuities in toponyms and pollen; Koch , Collis & their fanboys; so forth
To sum up you and your joyful creativity:
I do not know what you drink, smoke or take, but you had better stop it immediately because it does not work for you well.
@ Skribha
Yeah respect is earned, and you've struck out. You don't make any sense and you've come here with a bee in your bonnet about your etymology of Altay (as if it's relevant) & Curta. You've demonstrated that you don't understand basic linguistics; nor the idea that LFs have been around for decades (and no that's not gospel, but merely precedent historiography which you are ignorant of)
'' such as comparing me to a ‘lapp dog’''
That's your own self description (''I am a Schmidt and MaÅ„czak’s boy'') reflected back.
Is this the Manczak who imagines that Slavs come from the Vistula-Oder region on the basis of his 'word comparisons'? How deep
And my quotation about autochthonists is pretty accurate summation. Too bad you lack the intelligence & sense of humour to get it
Given that your ideological paradigm is collapsing, it is little wonder that you are running around like a headless chicken
Dear Davidski
Recently Blogger has not published my three-part comment again.
In it, I briefly described, but with data and sources, my version of the formation of the PIE and other Indo-European languages.
Probably due to the large number of links, these 3 comments ended up in spam.
Please pay attention to this as I am posting them again following this comment.
I hope you will publish them this time and that it will contribute to the resumption of the factual discussion about the formation of the PIE, etc.
I associate the PIE with the appearance of the first burial mounds / kurgans in the archaeological cultures of Varna and Suvorovo.
If you think otherwise, I would love to hear your arguments that it could not be as I claim.
If you disagree with what I wrote, destroy it in an honest discussion or let others do it. Please treat me honestly and fairly.
Note that unlike some, I always go to the point and provide sources for what I claim.
Additionally, I am well mannered and polite, unlike some of whose rude and illogical comments about me you post, see the one above.
Best ragards
SKRiBHa
@EastPole
As I have mentioned it above, I associate PIE with the appearance of the first burial mounds / kurgans in the archaeological cultures of Varna and Suvorovo.
As I promised, I have described the formation of PIE and other IE languages, but unfortunately I could not publish it here again, see above, so I posted it on my blog.
Can you take a look at it and write me what you think, please?
Here is a version without any links:
@All (except ‘Rob’)
Finally, let's do something constructive and summarise this thread.
Again I propose to establish a discrepancy protocol.
Can you describe your version of the origin (place, time, etc,) of the IE languages?
Below you will find my version of the formation of the PIE and Anatolian (Hittite), S/Hellenic, Armenian, Celto-Italic, Germanic, Indian (Vedic Sanskrit), Iranian (Avestan), Baltic, Slavic, Tocharian.
Note! I am not saying that there are no logical contradictions there, especially regarding the EEMH, ANE, WHG, EHG, WSH, SHG, CHG, EEF as responsible for a formation of the languages.
The possibilities of the PIE’s origin are as follows:
1. EEMH,
2. ANE > EHG> WHG > SHG,
3. CHG > WSH,
4. EEF.
It looks like ANE is everywhere, see:
WHG = ??%EEMH + ??%EHG/75%ANE
EHG (Haak 2015) = 75%ANE + 25%WHG (??%EEMH + ??%EHG/75%ANE)
EHG (Wang 2018) = 9%ANE + 91%NoWHG???
SHG = ??%WHG + ??%EHG
WSH = ??%EHG + ??%CHG
EEF = ??%WHG-??? + ??%CHG-???
CHG = ??%WHG (??%EEMH + ??%EHG/36% ANE) + 64%Dzudzuana???
PIE
My current version of formation of the PIE is logically related to the appearance of the first burial mounds / barrows / kurgans / Go’R+Ka / hill in Varna / Suvorovo, see:
Conflict or Coexistence: Steppe and Agricultural Societies in the Early Copper Age of the Northwest Black Sea Area
Blagoje Govedarica
According to this assumption, PIE would have been a language of the first builders of the hill barrows / kurgans from Varna / Suvrovo,... and even earlier a language of Miners / Górników /, Metallurgists / Hutników and Blacksmiths / Kowali from the Balkans from Vinca, Rudna Glava, or Ai Bunar, etc.
Logically PIE would have been the original language of the WHG European Forest Hunters I2 i.e. Old Europeans, not the EHG, WSH Herders R1b, R1a from the steppe / forest-steppe, or the CHG / EEF Farmers G2, H, T, J, C1a2, E1b1.
The PIE stems/roots would have been simultaneously alternated (‘satem’ and ‘centum/kentum’) and PIE itself would have been a synthetic / inflectional language.
However, PIE could have been a mixture of languages and vocabulary of:
• WHG Proto-PIE Hunters / Miners / Metallurgists / Blacksmiths, i.e. Old Europeans from the Eastern Balkans, e.g. vehicle / wóz / Wo'Z, wheel / koÅ‚o / KoL”o, Kółko / Ko'L”+Ko, gold / zÅ‚oty / ZL"oTy > (Z)+aLTay,
• EEF Farmers, e.g. ???,
• EHG, WSH Herders, e.g. horse / equus / Kuc / Ko’C, KÅ‚usak / K+L"oS+aK, Klacz / K+LaC".
This mixture could have formed Varna and Suvorovo cultures.
Post-PIE
Then, along with the technology of mining, smelting and processing yellow / gold and brown metals such as gold / zÅ‚oty / ZL”oTy > Z+Altay, copper / Miedź / MieDz’ and bronze / BrÄ…z / BRa”Z , this technological elite would have transferred its language to the steppe / forest-steppe, including Sredny Stog, CWC, Yamna and late Cucuteni– Trypillia cultures.
The map below describes the Post-Yamna state, but it has some errors, see CWC, Fatyanovo, Balanovo, Abashevo, Sintashta, Andronovo allegedly arising from Yamna, etc.
/wikipedia/commons/b/b1/Indo-European_migrations.jpg
The tradition of barrows / mounds / kurgans building may indicate the so-called dialect continuum from Varna, Suvorowo, Sredny Stog, Yamna, CWC, Fatyanovo, Balanovo, Abashevo, Sintashta to Andronovo.
Note! There is no NIE substrate listed in Proto-Slavic and Proto-Baltic!
Anatolian (Hittite)
EEF + WHG + EHG, WSH / Post-Yamna + Post-Sredny Stog + Post-CWC + Hurrian substrate.
Anatolian (Hittite) already was secondarily distorted and devoiced in comparison to CWC language (+H / alleged so-called laryngeals). Hittite sounds written as h/+H was a description of a secondary devoicing generated by Hurrian NIE substrate.
S/Hellenic
EEF + WHG + EHG, WSH / Post-Yamna + Post-Sredny Stog + Post-CWC + Post-BMAC? + Minoan / Anatolian NIE substrates?
S/Hellenic dialects already were secondarily distorted and devoiced in comparison to CWC language (-W, S>H>?, +H, B/P>PH).
Armenian
EEF + WHG + CHG + EHG, WSH / Post-Yamna + Post-BMAC? Proto-Kartvelian or/and Urartian substrates?
The Armienian language was secondarily very distorted and devoiced in comparison to CWC language (P>H, G>J, D>T, T>D, S>H, -W/P/?).
Celto-Italian
EEF + WHG + EHG, WSH / Post-Yamna + Post-CWC + Post-Bell Beakers. Vasconic / kartvelian like substrates? + (semitic like adstrate ???).
Celto-Italian languages were secondarily very distorted and devoiced in comparison to CWC language, (P>KW/L”, D>F, itp.)
Germanic
EEF + WHG + SHG + EHG, WSH / Post-Yamna + Post-CWC + Post-Bell Beakers + GAC/TRB? Vasconic / kartvelian like substrate? + (semitic like adstrate ???) = ‘lingua franca’ of ‘north-eastern Roman borderland’?
Germanic languages/dialects were secondarily very distorted and devoiced in comparison to CWC language, (D>T, T>D, S>H, K>H/G, G>K, B>P, B/P>F).
Indian (Vedic Sanskrit)
EEF? + WHG? + EHG, WSH / Post-Yamna? + Post-Sredny Stog + Post-CWC + Post-BMAC? + Harappan substrate.
Slavic, not Indo-Iranian etymology for Z+Altay proves that the Proto-Indo-Iranian could not have developed earlier in Fatianovo, Balanovo, Abashevo, Sintashta, but only later in Andronovo or only in BMAC, see:
‘It has nothing to do with the Iranian languages, see L>R, Proto-Indo-Iranian *È·́Ê°r̥Hanyam (“gold”), Proto-Indo-Aryan *źʰárHiá¹£, from Proto-Indo-Iranian *È·́ʰárHiÅ¡, from Proto-Indo-European *ǵʰelh₃- (“to shine”), cognate with Avestan (zairi) Sanskrit हिरण्य (hiraṇya), Avestan (zaraÅ„iia, “gold”), etc.’
The Vedic Sanskrit was secondarily distorted and devoiced in comparison to CWC language (L>R, +H, etc.).
Note! Compare Polish / Slavic simultaneous ‘satem’ and ‘centum/kentum’ alternations, which I am going to publish soon in the following posts.
Iranian (Avestan)
EEF? + WHG? + EHG, WSH / Post-Yamna? + Post-Sredny Stog + Post-CWC + Post-BMAC/Yaz + Elamite substrates?
Iranian (Avestan) is a logical reverse of Vedic Sanskrit, see e.g. deva, etc. It is even more distorted and devoiced than Vedic Sanskrit, i.e. it must have been formed much later, somewhere below BMAC / Yaz or already in the Iranian Plateau, as another distortion of the already partially distorted form of Vedic Sanskrit (L> R, S>H, +H, +F, P/B>F, G/Z>J, etc.).
Possible BMAC/Yaz or/and Elamite substrates in Proto-Iranian.
Baltic
EEF? + WHG + EHG, WSH / Post-CWC + Ugro-Finnic N1c / Post-Seima Turbino adstrate ???
The majority of Slavic sufixes and NIE substrate do not exist in Baltic!?
Slavic
EEF? + WHG? + EHG, WSH, Post-Suvorovo, Post-Sredny Stog, Post-CWC + GAC/TRB?
There is not NIE substrate in Proto-Slavic or Polish!?
There are not any alleged Iranian / Scythian-Sarmatian-Alanian-Ossetian genetic and linguistic borrowings in Proto-Slavic or Polish!!!
Tocharian
EEF? + WHG? + EHG, WSH?, Post-Yamna / Afanasievo? + Post-CWC? + local Asiatic substrates / adstrates?
Sporadically preserved ‘satem’ forms!
Questions:
1.
What about NIE Basques 100% Post-Yamna + Post-Bell Beakers + ???
2.
What about Aria R1b Yamna from Kutuluk kurgan?
3.
What about secondary devoicings, e.g. +H, S>H, B>P, B/P>PH/F, -W, D>T, T>D, K>H/G, G>K, etc.?
Post a Comment