search this blog

Wednesday, July 3, 2019

Evidence of European ancestry in the Philistines


The abstract below has just appeared at the European Nucleotide Archive (see here), so I'm guessing that the relevant paper and accompanying ancient genome-wide data will be published within weeks if not days. Emphasis is mine:

The ancient Mediterranean port-city of Ashkelon, identified as “Philistine” during the Iron Age, underwent a dramatic cultural change between the Late Bronze- and the early Iron- Age. It has been long debated whether this change was driven by a substantial movement of people, possibly linked to a larger migration of the so-called “Sea Peoples”. Here, we report genome-wide data of ten Bronze- and Iron- Age individuals from Ashkelon. We find that the early Iron Age population was genetically distinct due to a European related admixture. Interestingly, this genetic signal is no longer detectible in the later Iron Age population. Our results support that a migration event occurred during the Bronze- to Iron- Age transition in Ashkelon but did not leave a long-lasting genetic signature.

Update 4/7/2019: The paper is now available at Science Advances [LINK]. One of the Ashkelon ancients, who also shows a relatively high level of European ancestry, belongs to Y-Chromosome haplogroup R1 (probably R1b-M269). I've updated my Global25 datasheets with the new samples. Look for the Levant_ISR_Ashkelon prefix. Same links as always...

Global25 datasheet ancient scaled

Global25 pop averages ancient scaled

Global25 datasheet ancient

Global25 pop averages ancient

This is how they cluster in my Principal Component Analysis (PCA) of ancient West Eurasian genetic variation. The relevant datasheet is available here. Based on these results, it's tempting to think that the European ancestry in the Philistines may have been of Greek provenance. But keep in mind that this is just a two dimensional view and a simplification of reality. I'll have more to say about the ancestry of these individuals and the origins of the Philistines in future blog posts.

See also...

Five foot Philistines

How did steppe ancestry spread into the Biblical-era Levant?

202 comments:

1 – 200 of 202   Newer›   Newest»
zardos said...

This would prove the immigration of the Sea People and the biblical, probably genocidal war of the Hebrews against the Philistines.
Will be interesting to see which kind of European admixture came in.

Davidski said...

I'd be surprised if this European-related admixture isn't something from the Aegean. It probably resembles Mycenaean and/or Crete Armenoi ancestry.

Erikl86 said...

Finally, about time. Too bad it's only 10 samples and no more, out of the 150 skeletons they've found.

In any case, the dating here is super crucial.

The later IA samples dating is super-crucial - if its post 586 BC, meaning post early 6th century BC (still considered IA in the Levant, fellas), then this could prove that the remaining Philistines were indeed exiled by the Babylonians.

If it's before that, then it seems Philistines assimilation into local Levantine population was complete and their Aegean admixture was "diluted" over the centuries to basically leave no trace.

I have a hunch EIA Philistines, if the samples taken were already mixed with local Southern Levantines, would plot very similar to contemporary Western Jews or Cypriots. A bit ironic.

Desdichado said...

"I have a hunch EIA Philistines, if the samples taken were already mixed with local Southern Levantines, would plot very similar to contemporary Western Jews or Cypriots. A bit ironic."

Modern political and social considerations are not pleased, I imagine. Probably why we're still seeing nothing on the Tollense DNA, for that matter. It's frustrating sometimes.

EastPole said...

@zardos

“Will be interesting to see which kind of European admixture came in.”

Yes, interesting whether genetics will confirm some linguistic theories:

https://www.scirp.org/journal/PaperInformation.aspx?PaperID=69428

Maybe they were influenced by Thracian/Phrygian people who had some North-East-European origin at least their religion seems to shows some Slavic influence: worship of Sabazios etc.

Ric Hern said...

I wonder if the European bit was unique to Ashkelon ? Was this really a mass migration or rather a Trading Outpost where European traders accumilated ?

Leron said...

The ancient historians already knew of Lydians who set up a colony in Ascalon and brought a cult of the Fish (Itchys in Greek, Dagon in Semitic).

Andrzejewski said...

Some researchers thought that they may have been Anatolian language speakers like Luwian, based on words resembling “tyrannos” and “kupa” etc.

zardos said...

Going by the historical, including biblical accounts, this must have been a mass migration. Not just some traders. These results should finally seal it. The Sea People seem to have been a mixed bunch, but Balkan-IE influence is what I guess for Philistines. Mycenaean-like mixture is the best bet, agree on that.

Michalis Moriopoulos said...

Very exciting news. Of course, I also expect the "European admixture" to be Mycenaean-related. The later Iron Age samples should also be very interesting.

PF said...

Nice. On one hand this is exactly what I expected -- early Sea Peoples (?) / Aegean influence in the Levant. On the other, it's unexpected that it didn't survive locally. This gives more credence to the idea that Western Jews acquired their Aegean-like ancestry somewhere else, and at a different time.

Alternatively, the admixture could have happened then and there, followed by a later migration and replacement/dilution locally. As Erik says the precise dating is important here.

Another option is that this is just one local of several macro events that brought "European" ancestry to the region. There's still a significant shift from Levant_Chl / Jordan_BA --> Sidon_BA (Canaanites). I thought we'd find still more Euro input after Sidon, but it seems whatever transpired may not have lasted, until reappearing again in Classical/Medieval times? Looking forward to the full paper and samples...

Erikl86 said...

@Davidski, it seems the raw data is available - when can we expect to play with these on Global25 ? ;) (or do you prefer to wait for the study to come out before)

John Thomas said...

After my Sunday school lessons I always thought contemptuously of the Philistines as the world's biggest losers. I know now that this is silly, but the old taint is hard to move....

Matt said...

@PF:This gives more credence to the idea that Western Jews acquired their Aegean-like ancestry somewhere else, and at a different time.

I dunno; I don't really get it this idea.

If you reprocess the Global 25 data, the European Jewish groups are between Lebanon Rome and North Italy. They're saliently not at all between Mycenaeans and Lebanon Rome (or Bulgarian IA who are similar to Mycenaeans).
See: https://imgur.com/a/cx8sJnO

So, I mean: We have a populations which are documented as forming in Italy during the Roman period. They're genetically between Italian populations that formed in Roman Italy and the Roman Era population of the Levant (most true for Italian and Spanish Jews and less true for Central European Jews, who have some ancestry probably from German and Polish like populations).

What's the call to imagine they had Aegean-like ancestry other than is mediated through a fusion of Italian-like and Lebanese-like populations?

I suppose you could call up a more complex scenario where they are in those positions because of a complex combination of circumstances between Lebanese-like+Mycenaean-like+Pop X, but it hardly seems like there's any reason for it.

Simon_W said...

@Desdichado

"Modern political and social considerations are not pleased, I imagine. Probably why we're still seeing nothing on the Tollense DNA, for that matter. It's frustrating sometimes."

:D I don't believe it, seriously. You completely overestimate the significance and social impact of ancient DNA research. This stuff is intriguing to some enthusiasts in internet forums, but stays largely unknown and irrelevant to the general public in Germany, be it nationalist and chauvinist or leftist and liberal. So really no team of university scientists has to care about the political impact their research may have. I'm speaking of Germany here, it may perhaps be different in other countries.

Erikl86 said...

@Matt

" We have a populations which are documented as forming in Italy during the Roman period. They're genetically between Italian populations that formed in Roman Italy and the Roman Era population of the Levant (most true for Italian and Spanish Jews and less true for Central European Jews, who have some ancestry probably from German and Polish like populations).

What's the call to imagine they had Aegean-like ancestry other than is mediated through a fusion of Italian-like and Lebanese-like populations?"


The thing is, you also have extensive evidence of significant Hellenistic Jewish communities predating the Jewish community in Rome by centuries, all over the East Mediterranean Aegean basin. You have archaeological as well as historical evidence of Greeks converting to Judaism. And the first Jews to arrive to Rome, spoke Koine Greek, not Aramaic, and bore Hellenistic names.

Western Jews can also be modeled as 35%-40% Levantine, 50-60% Aegean-like, not just half-way between North Italy and the Levant.

Btw, both Aegean Greeks and Sicilians can be modeled as half way between N. Italy and Levant, yet that doesn't make sense.

Also, if you take into account recent evidence that Late Republican , Imperial Era Romans plotted similar to Sicilians and Aegean Greeks, again showing that it's unlikely that Western Jews are the result of N. Italians and Levantines mixing together.

Desdichado said...

"Nice. On one hand this is exactly what I expected -- early Sea Peoples (?) / Aegean influence in the Levant. On the other, it's unexpected that it didn't survive locally. This gives more credence to the idea that Western Jews acquired their Aegean-like ancestry somewhere else, and at a different time. "

Yeah, well the precise dating of the ten specific samples may have something to do with that, too. If they only have a sample or two of that later period where it's no longer present, then it's not a big enough sample to demonstrate anything more interesting than the fact that they did insufficient sampling to make any real conclusions.

Not that that seems to have stopped them, but without sufficient samplings, those conclusions can't possibly be very confident.

Joshua Lipson said...

And yet: Ashkenazim can be modeled as 80-85% Sephardic and 15-20% Northern/Central/Eastern European.

Should be a clear indicator that *either* a binary Levantine-North Italian OR Aegean-Levantine model is missing a significant component. (Perhaps it was overestimated in the past, but it's become underestimated in the lay analyst community.)

rozenblatt said...

Paper is out: https://advances.sciencemag.org/content/5/7/eaax0061

Labayu said...

@zardos

This would prove the immigration of the Sea People and the biblical, probably genocidal war of the Hebrews against the Philistines.

As far as well can tell archeologically, neither Hebrew polity ever conquered the Philistine city states. Philistine Gath appears to have been destroyed by Hazael of Aram-Damascus. Philistine culture was alive and well in Ashkelon until Nebuchadnezzar II sacked the place and reportedly exiled the survivors.

@Leron

The ancient historians already knew of Lydians who set up a colony in Ascalon and brought a cult of the Fish (Itchys in Greek, Dagon in Semitic).

The linguistic connections I know of point to the Luwian languages of western Anatolia, which are not that distinct from Lydian, at least based on our minimal knowledge. The word dagon means “grain” though, dag is “fish”. The fish god hypothesis was a Nineteenth Century idea that stemmed from a misunderstanding of the etymology, which has subsequently been clarified by Ugaritic and Akkadian cognates.

Matt said...

Erik, I'm not sure what you mean by "Aegean-like", but 65:35 Mycenaean:Lebanon_Rome absolutely does not work very well: https://imgur.com/a/xnCyVaX

Too EEF. No can do. Not a goer.

You can, I guess, cancel this out by a scenario of a large scale blending of other sources. But it does not seem simple.

Cretan Greeks seem fine where they are. Roman Era involved a lot of population movement. They're probably not Mycenaean+anything in a 2 way combination but result of much complex population movement between Italy and Greece and other parts of Roman Empire.

If you're talking about some other kind of Greeks that were post-Rome and had a complex ancestry and who were not hugely distinguishable from Italians, perhaps a two way combination with them is then possible, that's possible.

But the Mycenaean and Greek colonist groups we have certainly cannot mix with Roman Lebanese (or any preceding Levantine population) to produce European Jewish groups in a two way mix, or as the dominant or most substantial portion of ancestry.

Arza said...

ASH_IA1 1379–1131 cal BCE mtDNA H92 Y-DNA R1

Best 2-way qpAdm model for IA1:

ASH_LBA Crete_Odigitria_BA 56.9 43.1 NA 19.2 19.2 NA 0.675

So Y-DNA R1 came from Mycenaean Crete?

EastPole said...

Most European ASH_IA1 looks like a mix of Anatolia_IA and less European ASH_IA1:

https://i.postimg.cc/K8Gr4n30/screenshot-504.png

Two Anatolia_IA samples MA2198 and MA2197 come from Damgaard et al. Science 2018. But it is difficult to indentify the archeological context. Are they Phrygians?

https://i.postimg.cc/G2CphjH2/screenshot-505.png

Davidski said...

@EastPole

Two Anatolia_IA samples MA2198 and MA2197 come from Damgaard et al. Science 2018. But it is difficult to indentify the archeological context. Are they Phrygians?

One is probably part Celtic and the other part Cimmerian.

A closer look at a couple of ancients from Hellenistic Anatolia

Drago said...

Are these 3 or 4 late Bronze Age Greeks new data ?

Arza said...

@ Dragos

Nope.

Data file S2
I9033 S_Greece_LBA
I9006 S_Greece_LBA
I9010 S_Greece_LBA
I9041 S_Greece_LBA

Davidski said...

@All

Ashkelon_IA1:ASH067 probably belongs to R1b-M269. But it's difficult to say where that came from exactly, because it seems that R1b-M269 was already present in the Levant during the LBA.

R1b-M269 in the Bronze Age Levant

Andrzejewski said...

Isn’t R1b-M269 the one originating in BB?

Matt said...

Quick and dirty adaptation of the PCAs I generated upthread and discussed to PF and Erik, including the new BA-IA Levant samples: https://imgur.com/a/uhG0qaB

It seems that the samples of Ashkelon IA 1 are, in fact, clinal towards EEF rich Southern European populations of the time (most likely Mycenaean like, but could be others, may take more analysis), in a way that post-Roman European Jewish, Cretan and Southern Italian populations are not.

R1b-M269 is slightly unusual. I'd note that in their highest coverage test, Iberia_BA is comparable in fit to Crete_Odigitria_BA: "To check whether these results are due to the low coverage of ASH_LBA, we repeated this analysis, but this time, we modeled ASH_IA1 as a three-way mixture of each of the candidate populations, Levant_ChL and Iran_ChL. The two latter populations have higher genome coverage and can model ASH_LBA well in combination (table S3). In this analysis, only the models including “Sardinian,” “Crete_Odigitria_BA,” or “Iberia_BA” as the candidate population provided a good fit (χ2P = 0.715, 49.3 ± 8.5%; χ2P = 0.972, 38.0 ± 22.0%; and χ2P = 0.964, 25.8 ± 9.3%, respectively).". Though I doubt that this means Iberia_BA necessarily, there are some suggestions of a these groups representing a broader alliance of peoples along north shores of the Mediterranean stretching.

Matt said...

Does look to me more like Ashkelon IA2 possibly associated with "re-balancing admixture" from some population like Levant_JOR_EBA, rather than necessarily "just" local bounceback.

claravallensis said...

The sea peoples have been connected to Nuragic Sardinians, and the latter in a recent paper have been shown to carry lots of R1b-V88. Could this R1 be in fact V88? Especially given the admixture models they provide.

Drago said...

Odds are it’ll be Z2103; question is whether it nests in Caucasian or European sub-branches

Open Genomes said...

ASH008 & ASH066 J-Z2331
ASH067 R1b-M269
ASH087 L-M20


Ancient Philistine Y-DNA analysis

Andrzejewski said...

@Matt "R1b-M269 is slightly unusual. I'd note that in their highest coverage test, Iberia_BA is comparable in fit to Crete_Odigitria_BA: "To check whether these results are due to the low coverage of ASH_LBA, we repeated this analysis, but this time, we modeled ASH_IA1 as a three-way mixture of each of the candidate populations, Levant_ChL and Iran_ChL. The two latter populations have higher genome coverage and can model ASH_LBA well in combination (table S3). In this analysis, only the models including “Sardinian,” “Crete_Odigitria_BA,” or “Iberia_BA” as the candidate population provided a good fit (χ2P = 0.715, 49.3 ± 8.5%; χ2P = 0.972, 38.0 ± 22.0%; and χ2P = 0.964, 25.8 ± 9.3%, respectively).". Though I doubt that this means Iberia_BA necessarily, there are some suggestions of a these groups representing a broader alliance of peoples along north shores of the Mediterranean stretching."

Might be related to Cardial Pottery?

Andrzejewski said...

"Recent ancient DNA (aDNA) studies have reported a high degree of genetic continuity in the Levant during the late Pleistocene and early Holocene that was followed by increasing population admixtures with Anatolian- and Iranian-related populations at least up to the Middle Bronze Age.."

Perhaps this increasing Anatolian admixture is responsible for the "South Euro" shift in Ashkenazi Jews?

On the same vein, could people who were described as "red haired" have anything to do with this Anatolian admixture? We know Esau was depicted as red-haired and it was said that he may be part-Hurrian (Kura Araxes), and King David was a descendant of a Moabite. Could there be some fire behind the smoke of these mythical legends?

Open Genomes said...

@Josh Lipson

That one Iron Age II Philistine clusters right with Ashkenazi Jews on the PCA.

AWood said...

@Andrezejewski

M269+ wasn't in that part of Europe at the time of Cardial Pottery. You might be getting confused with V88 which was in the Balkans, Italy (because it's in Sardinia), and parts of Spain where Cardial pottery has been found.

Open Genomes said...

An Iron Age II Philistine clusters with Late Bronze Age Southern Greeks (Mycenaeans)

Samuel Andrews said...

If these southeast European migrants aren't Greek then what are they?

Looks like ancient Greeks were a homogeneous population. Hopefully, lefty historians will stop saying being Greek in ancient times was an identity not a race. So, far it looks like the ancient Greeks were a 'race' with common ancestry.

@History experts,

Early Greeks (Myceneans) had been in contact with civilizations in the Near East before classical Greek era (beginning in 800bc) correct? Who else would be a candidate for this *southeast* European introgession into the Near East than Greeks.

Or I guess, there's no documentation of Greek language in ancient Near East (before Alexander the Great) therefore it was a non-IE sea people similar to Greeks.

PF said...

@Matt @Erik

"What's the call to imagine they had Aegean-like ancestry other than is mediated through a fusion of Italian-like and Lebanese-like populations?"

I always doubted the half-Italian theories that have been around for ages for many reasons, but primarily because of one simple fact: there are multiple Jewish groups which cluster together that *never had any history in Italy* nor any significant later admixture with Jews who did. How can that possibly be explained away?

Last year David posted the genomes of Romaniote Jews -- a completely isolated relic population. I made this model for Ashkenazi using Romaniotes and only medieval genomes, which I think will be hard to beat:

[1] "distance%=0.9828"

Ashkenazi_Jew_average

Romaniote_Jew,81.2
Italy_Medieval_Collegno,10
Germany_Medieval,7.2
Nomad_Medieval,1.6

http://eurogenes.blogspot.com/2018/10/ashg-2018-open-thread.html?showComment=1539663555059#c8560389148206335905

So when you say "We have a populations which are documented as forming in Italy during the Roman period" I simply think it's untrue. Ashkenazi may or may not have formed in Italy, but certainly other Jews nearly identical to Ashkenazi (minus exactly what is expected to have been added to them along the way), did not.

The relationship between Italians, Greeks, and East Med populations is super complex and will be really interesting to tease out. I just remembered how in the Canaanite paper they found rare allele sharing between Italy_North and the Canaanites (IIRC)... layers upon layers.

Anyways off to read the paper and check out the genomes. Thanks David for getting them into the G25 so fast!

Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...

M269 was also found in Latvia MN Date: 4251-3976 calBCE

Isn't everything found on the steppe so far including Samara HG downstream at least partially towards Z2103? I remember a paper on that.

Drago said...

“”
M269 was also found in Latvia MN Date: 4251-3976 calBCE”

Are we sure about that ?

Bob Floy said...

Luwians? Something to do with Troy?

Open Genomes said...

Global25 nMonte for the most European of the Philistines, ASH068

GRC_Minoan_Lassithi 25.6%.

Here ASH086 clusters very closely with a Mycenaean:

ASH068 on the Global25 Ward's distance-squared clustering tree

Matt said...

@PF: There may be examples of Jewish groups who are similar to the Ashkenazi, Sephardi and "Italian" Jews without any history in Italy, but the Romaniote Jews don't seem to be this - https://en.wikipedia.org/wiki/Romaniote_Jews

"The Romaniote Rites represent those of the Greek-speaking Jews of the Byzantine (or former Byzantine) empire, ranging from southern Italy (in a narrower sense the Apulian, the Calabrian and the Sicilian Jewish communities) in the west, to much of Turkey in the East, Crete to the south, Crimea (the Krymchaks) to the north and the Jews of the early medieval Balkans and Eastern Europe.

The Sefer Yosippon was written down in the 10th century in Byzantine Southern Italy by the Greek-speaking Jewish community there. Judah Leon ben Moses Mosconi, a Romaniote Jew from Achrida edited and expanded the Sefer Josippon later. This community of Byzantine Jews of southern Italy produced such prominent works like the Sefer Ahimaaz of Ahimaaz ben Paltiel, the Sefer Hachmoni of Shabbethai Donnolo, the Aggadath Bereshit and many Piyyutim. The liturgical writings of these Romaniote Jews, especially the piyyut were eminent for the development of the Ashkenazi Mahzor, as they found their way through Italy to Ashkenaz and are preserved to this day in the most Ashkenazi mahzorim.

The Jews of Southern Italy continued to be Greek-speakers until the 15th century (where they were living together with their Greek-speaking Christian counterparts). When they were expelled and went to different regions of Greece, especially Corfu, Epirus and Thessaloniki, they could continue speak their Greek language, even if this language was somewhat different from that of Greece."


There's more in the wiki article but the point is it's not like there was a frozen community living in Greece which did not receive emigration directly from Italy as above. Or from the Jewish groups living elsewhere that did form in Roman Italy - The article also mentions that Ashkenazi and Sephardi groups "bolstered" the Romaniote community in the late Middle Ages. There would not be reproductive barriers between these Jewish groups from Europe who moved into the Ottoman Empire. Romaniotes are not a "population fossil" that shows a stage preceding the formation of groups who had an origin in Italy!

That said, I'd guess that it's just about possible that Jewish groups formed from a combination of Levantine and "post-Roman" genotype that lived outside Italy, but had substantial Italic ancestry.

But it looks like when it comes to the theory of forming from an "Aegean-like" and Levantine background, where Aegean-like is defined as the Mycenaean-like people who the Empuries colony suggests persisted all the way through the classical period - that's not a goer. (Unlike the AshkelonIA1 set, who look like they do form that way!)

(Updated version of the PCA I produced above including the Romaniote popualation: https://imgur.com/a/6HV5Qgg)

Matt said...

Sam: If these southeast European migrants aren't Greek then what are they?

Possibly Luwian (West Anatolian), since they would have been genomically fairly similar, and the population history of Southern Italy at this time is also not well studied.

Looks like ancient Greeks were a homogeneous population. Hopefully, lefty historians will stop saying being Greek in ancient times was an identity not a race. So, far it looks like the ancient Greeks were a 'race' with common ancestry.

Re how much we can say ancient Greeks (Greek speakers who self identified as Greek) were a single homogenous population probably depends on the exact value of "ancient" - this transect shows ongoing admixture, for'ex.

Greeks within Greece (at least southern and insular Greece) certainly look to have remained relatively similar through the Mycenaean to classical period (hence my comments above to PF and Erik), though this may mask fairly some degree of exchange with genomically similar West Anatolians (the Ionian Greeks from the Western fringe of Anatolia are noted to comprise many of the most creative and interesting classical Greeks).

But as Greek culture spread outside mainland Greece, this may have changed.

So historians who speak of culturally Greek but genetically different people may be accurate so long as they are when talking about a later period within the ancient world and outside Greece.

(As I note to PF and Erik as well, likely Greeks within Greece itself admixed substantially with other Med populations, including Italics speaking groups, under Roman rule).

Erikl86 said...

@Matt

I agree with PF here, and this is one of the most prominent reasons for why it's most likely Western Jews didn't receive their admixture mostly from Northern Italian-like population, but an Aegean-like.

I'll explain.

1. The biggest community in the Italian peninsula was Roman. Recent studies, such as the upcoming Moot et al. study, show that Romans during the Late Republican Era and Imperial Era were not North Italian-like, but clustered much closer to Aegean Greeks and South Italians, most likely as a result of absorption of the Magna Graecia colonies.

2. There is substantial historical as well as archaeological evidence for conversion of Greek people (or Hellenized Greeks) in Asia Minor and the East Mediterranean, predating Jewish presence in Italy by centuries.

3. The Romaniote Jews samples in the Global25 datasheet, which I've provided, are from the Ioannina community, which didn't mix with Ashkenazi or Sephardi Jews. You also going to have a hard time explaining this genetic drift PCA created by Davidski months ago, showing ALL Western Jews - including non-Sephardic North African Jews such as Libyan and Tunisian Jews (and Syrian Jews) - clustering together, pointing to a shared ethnogenesis:

https://2.bp.blogspot.com/-j6Q9nedrve0/W7HB40YRx6I/AAAAAAAAHQY/EvWJh88wWKcgyxEKIu8LPSDDjtGvxiKowCLcBGAs/s1600/Mediterranean_PCA_01-10-2018.png

4. The first Jews to settle in Rome (or Italy, for that matter), spoke Koine Greek and bore Hellenistic names. We know this from the Jewish catacombs both in Rome and the older ones in Venosa. Hellenistic Judaism which was welcoming of proselytes was quite a substantial denomination and we have evidence from The Greek terms that refer to God-fearers (theosebeis) are found in ancient literature (Greek, Roman, and Jewish) and synagogue inscriptions discovered in Aphrodisias, Panticapaeum, Tralles, Sardis, Venosa, Lorium (in Rome), Rhodes, Deliler (Philadelphia) and Miletus.

5. Both contemporary Sicilians and Aegean Greeks can also be modeled as halfway between North Italians and Levantines. Does this mean that Aegean Greeks have substantial North Italian origin? If anything, recent studies have shown that Venetians and other North Italian maritime republics have left basically no genetic impact on the Aegean islands.

6. I do believe some Western Jews have some minor N. Italian admixture - but I believe the dominant component among Western Jews is Aegean-like, whether they got it from Imperial Era Romans and South Italians or in the East Mediterranean. Employing Occam's razor here, I kind of lean on believing they got it in the East Mediterranean.

Matt said...

@erik: 3. The Romaniote Jews samples in the Global25 datasheet, which I've provided, are from the Ioannina community, which didn't mix with Ashkenazi or Sephardi Jews.

Already dropped them into PCA. Above I've referred to emigration of Jewish communities to Greece from Southern Italy, so it seems like this could still provide a substantial part of their origin even in absence of influence from other Jewish groups. But how would you know that this community didn't ever mix with emigrants from Ashkenazi or Sephardi communities, anyway?

"Genetic drift" PCA there shows that the Jewish groups would be modelled with a variable combination of A: some proto population that diverges from Samaritans and B) European populations (the Ashkenazis, Sephardis and Italians seem to prima facie point to Italy). Doesn't tell you that this is due to a shared genetic drift bottleneck, or even if so when that would've happened...

5. Does this mean that Aegean Greeks have substantial North Italian origin? If anything, recent studies have shown that Venetians and other North Italian maritime republics have left basically no genetic impact on the Aegean islands.

Most likely, from the Roman Period. But also the effects of Slavic ancestry.
They're certainly not continuous with Mycenaeans, and are shifted relative to them towards where movements from Anatolia and Italy would place them. Need some admixture. A transect of Greek islands over time will probably show this process happening.

I've opened the possibility above that it may be a population very similar to an Italian Roman population but living in Roman Greece. Just not Mycenaean like.

To clarify, I'm really mainly disagreeing here that the "dominant" component in the European Jewish groups is Mycenaean like, as in 65% Mycenaean, with the balance Levant. Probably 30-35% as most. Consistent with admixture with groups from Italy and Greece who derived substantial ancestry from Mycenaean-like groups, but who were also substantially Italic (as in, pre-Roman Empire Italics). Jewish groups did not form from Mycenaean-like Greek colonists and Levantine populations.

We will see from ancient dna that the Roman Empire was a demic event of receiving Italian emigrants for Greece and Anatolia (and to a lesser extent Spain and to a still lesser extent the other parts of empire), not just a phenomenon of immigration into Roman Italy.

Erikl86 said...

@Matt, I highly disagree with most of what you wrote in your last post.

"Genetic drift" PCA there shows that the Jewish groups would be modelled with a variable combination of A: some proto population that diverges from Samaritans and B) European populations (the Ashkenazis, Sephardis and Italians seem to prima facie point to Italy). Doesn't tell you that this is due to a shared genetic drift bottleneck, or even if so when that would've happened...

It actually does tell me that they are likely descended from the same population, with little further admixture. Considering none of them are +80% Levantine, then this means they all descend from a similar population of Jews post-admixture even with South Europeans. The most historically plausible source for this, would be East Mediterranean population.

Most likely, from the Roman Period. But also the effects of Slavic ancestry.
They're certainly not continuous with Mycenaeans, and are shifted relative to them towards where movements from Anatolia and Italy would place them. Need some admixture. A transect of Greek islands over time will probably show this process happening.


This is wrong based on the recent genetic evidence we have. Aegean Greeks pretty much remained similar to Mycenaeans, with some Near Eastern admixture and very minimal Slavic admixture. This is contrast with mainland Greeks which show no Near Eastern admixture and much more Slavic admixture than the islanders.

I do not believe that Western Jews' Levantine forefathers mixed with Mycenaeans - that would be wrong chronologically. But considering how close Classical Greek sample from Empuries to Mycenaeans, it seems Classical Greeks still were pretty much Mycenaean-like.

The Hellenistic-era and then Roman-era Hellenic population which mixed with Judean Levantine Jews was also, probably, still similar to Mycenaeans on some level (maybe mixed with Near Eastern and Anatolian).

Here, models for Western Jews using the lastest Ashkelon IA2 Levantine samples:

[1] "distance%=1.5458"

Ashkenazi_Jew

Levant_ISR_Ashkelon_IA2,31.2
Iberia_Northeast_Empuries2,26.6
ITA_Collegno_MA,21.8
Baltic_EST_MA,8.2
DEU_MA,6.8
Mozabite,5.4



[1] "distance%=1.8035"

Sephardic_Jew

Levant_ISR_Ashkelon_IA2,32.4
Iberia_Northeast_Empuries2,31.8
ITA_Collegno_MA,28.4
Mozabite,7.4


[1] "distance%=1.6041"

Italian_Jew

Iberia_Northeast_Empuries2,35.8
ITA_Collegno_MA,28.8
Levant_ISR_Ashkelon_IA2,28
Mozabite,7.4

Aram said...

From the 4 y dna we have the J1 are locales.
L-M20 is very unlikely to have a European origin. We don't have any case of L in ancient Europe as far as I remember. At best it can be of Anatolian origin.
Only the R1b can be from Europe.
Imho there was a influx both from Anatolia and Europe.

zardos said...

@Matt: The impact of Roman colonisation should be much bigger in places like Gallia and even Britain than in Greece or Anatolia, simply because of the population densities and impact of the conquest.

Aram said...

It is possible that some shift toward Anatolia started even prior Iron Age. So hard to say did the L-M20 came in IA or before.
This is how Israel_LBA sample (from this paper https://www.tandfonline.com/doi/abs/10.1080/00758914.2017.1368204) looks if Jordan_EBA is used as a reference.

[1] "distance%=1.8746"

Levant_ISR_MLBA

Levant_JOR_EBA,62
Anatolia_Kaman-Kalehoyuk_MLBA,21.6
IRN_Seh_Gabi_C,6.6
BGR_C,5.8
Kura-Araxes_ARM_Kaps,4

zardos said...

@All: Does anyone have a simple overlook for testified Roman veteran colonies?

Davidski said...

@Aram

It seems to me like there's a lot of Caucasus ancestry in these Bronze and Iron Age samples from Israel.

I guess the presence of L-M20 in the Philistine male strengthens the case for this ancestry coming directly from the Caucasus and surrounds.

If so, the Aegean (European) ancestry in the Philistines is largely a different issue, although granted, some of the Aegean and Caucasus ancestries may have moved into the Levant together from Anatolia.

Cpk said...

Do you consider Ashkenazi Jews European or Near Eastern population ethnically?

Davidski said...

@Cpk

Do you consider Ashkenazi Jews European or Near Eastern population ethnically?

European ethnically, but of mixed European/Near Eastern origin genetically.

And I guess this can be easily extended to Southeastern European populations like Sicilians.

Drago said...

Ashkenazi Jews aren’t European; because they don’t identify as such
Their ties are to Israel ; not Poland or Greece

Davidski said...

I wouldn't know about that. My view is based on the fact that their ethnogenesis was in Europe and their own language is Germanic.

Drago said...

My view is based on simply asking around, matter of factly, (of the fairly decent number I know), at least of diasporans

It just shows that identity is complex and multilayered ; and a panoramic genetic view isn’t alone to define European ness

For me; as I said earlier; “Europeans “ formed in the Bronze Age; by and large . Of course; someone who arrived the Roman era and otherwise fully integrated into their local millieu is every bit the same; but the question is less clear for Roma & Jews

Davidski said...

@All

I reckon ASH068, the most northern shifted Philistine, actually has recent western Mediterranean ancestry.

I'm going to look into this more closely.

Erikl86 said...

@Drago, as an Ashkenazi Jew myself, ethnically, Ashkenazi Jewish as a separate ethnic group did form in Europe, as in - Ashkenazi culture was shaped by our ancestors experiences in Europe over a millennia and a half, the severe genetic bottleneck our ancestors experienced happened in Europe, and our albeit minimal Slavic admixture happened in Eastern Europe.

Btw, same goes for Sephardic Jews - ethnically, they formed as a separate sub-ethnic group within the general Western Jewish population in Europe.

However, genetically as well as culturally, we are East Mediterranean population.

A good example are the Roma people - their ethnic identity was changed and shaped in Europe, however genetically they are not European.

Drago said...

And I don’t know even genetically; what 20% exactly means ?

Erikl86 said...

@Drago - what do you mean by 20%?

Western Jews are mostly East Mediterranean population.

Almost all our paternal subclades derive from the Levant ~2000 years ago.

The Northern European component among Ashkenazi Jews is fairly minimal at best.

Davidski said...

Are Cretan Greeks European? Definitely. Their Northern European ancestry is minimal too.

I honestly don't think that this is a simple matter of calculating ancestry proportions.

In any case, these sorts of discussions never end well, so let's move on.

Cpk said...

Then who do you consider Eastern Med. populations?

Davidski said...

@Cpk

Then who do you consider Eastern Med. populations?

Me?

I honestly don't care. If I have to I just draw an arbitrary line where it suits at a given time. There's too much genetic overlapping in that part of the world to fixate about this stuff.

Drago said...

@ Erik

Thanks for your perspective. I'm sure most agree and are aware of the process you outlined, indeed youd know far more about it than us.
Yes, Ashkenazi & Sephardic Jews emerged in Europe, were shaped by it, and in turn helped shape Europe. However, the question is- do you identify as European and are you perceived as such by other Europeans ? This is a complex question and it lays at the core of any question of ethnicity, not just AJs.
And this is not me saying you arent or shouldnt be. As I said, my perspective seems to come from diasporans, who might have come to from 2 or 3 diasporan moves out of Europe, who appeared to me not to have much affinity with Europe.

From an ethnogenetic perspective, you could have a clinal genetic continuum between Sicily and Jews (or whatever it is exactly), but members along it formed at different times, and their ' cultural poles'' and origins myths might be more distinctive & punctuated, and not simply a ''blend''.
For example, the origin myth of Germans is Odins Folk, that of the Slavs (most might not, in fact, know), is Danubian Christendom. What is the origin myth of Ashkenazi Jews ?

(BTW- yes that’s what I mean- 20% European genetically )

Cpk said...

@Davidski I agree with you, too much overlapping in the area. I started the discussion because there is a continuum between S.Europe and N.East and i wondered where the other people would draw the line. Ashkenazi seem to be about half Italian half Levantine but largely Near Eastern farmer descendant population (93% EEF?), and sometimes on different pca plots they are positioned on different places although usully closer to South Italy.

@Drago I don't think how you feel about it / identify matters that much. Your ancestors are who they are.

Drago said...

@ CPK

“ I don't think how you feel about it / identify matters that much. Your ancestors are who they are“

But all that is is admixture.
Not identity

Davidski said...

@Cpk

The 93% EEF figure was never really correct, and nowadays, as far as I know, is only being propagated at Eupedia. My Global25 test shows this...

Ashkenazi_Jew

EEF,29.2
Levant_N,26.6
Yamnaya,24.2
Iran_N,12
Anatolia_N,6
Iberomaurusian,1.2
East_Asian,0.8


zardos said...

Do Ashkenazi really have more Yamnaya than some Southern Europeans? Or is that calculation different from some others made? What are the numbers for Southern Iberia, Sicily, Cypriots and Anatolian Greeks?

zardos said...

I mean they cant be more Yamnaya than Albanians and mainland Greeks. Or can they?

Erikl86 said...

@Davidski

I always have difficulties with these kind of models.

I mean what can we really learn from this? Considering Western Jews were formed much after these populations were no longer around, they really had no part in their ethnogenesis.

Davidski said...

@zardos

Admittedly, I haven't spent too much time thinking about this model, but I'd say it's more or less accurate.

[1] "distance%=2.9721"

Albanian

Yamnaya,35.4
Anatolia_N,32.8
EEF,24.6
Iran_N,3.2
Levant_N,2.2
WHG,1
East_Asian,0.8

[1] "distance%=2.5003"

Greek

Anatolia_N,40.6
Yamnaya,35.2
EEF,12.4
Levant_N,6.6
Iran_N,4
WHG,1.2

[1] "distance%=2.6347"

Spanish_Andalucia

EEF,56.8
Yamnaya,31.6
WHG,7
Iberomaurusian,4.4
Sub-Saharan_African,0.2

[1] "distance%=2.4667"

Sicilian_East

EEF,45.2
Yamnaya,23.8
Levant_N,17.2
Iran_N,10.6
Anatolia_N,2.4
Sub-Saharan_African,0.6
Iberomaurusian,0.2

[1] "distance%=2.7768"

Cypriot

Anatolia_N,38.2
Levant_N,28
Iran_N,21.4
Yamnaya,12.4

[1] "distance%=3.68"

Greek_Central_Anatolia

Anatolia_N,52.2
Iran_N,25.8
Yamnaya,13.8
Levant_N,8.2

zardos said...

@Erik: If AJ have such a high Yamnaya percentage, they have acquired it from a European population and the admixture event must have been quite significant. I think that can be said.
I also wonder about the almost 1 percent East Asian. Probably some Khazars really made it into the genepool of modern AJ.

Davidski said...

@Erikl86

I think distal models can be a useful way of comparing populations, especially when proximate ancestry references aren't available, unknown or very clear.

zardos said...

@David: Thanks. Ok, these numbers make sense relatively. But the absolute numbers are more difficult, because I remember in most models in papers made no European population had more Yamnaya than 60 percent. Rather 55 maximal.
I might be wrong, but in your calculation f.e. Norwegians and Lithuanians must exceed that number by far.
Now its about estimates, but which one is more accurate?

This matters for Philistines and AJ as well, since it changes the necessary European input to create such populations.

Davidski said...

@zardos

The Near Eastern ancestors of Ashkenazi Jews had about 5% Yamnaya ancestry. They were probably similar to this...

[1] "distance%=3.6637"

Samaritan

Levant_N,55.4
Iran_N,25.2
Anatolia_N,15.4
Yamnaya,4

So the relatively high level of Yamnaya ancestry in Ashkenazi Jews does mostly come from their European ancestry, but it isn't from one source, but rather from multiple sources, such as Italy, Germany and Eastern Europe.

Their East Asian admix is probably from contacts via the Silk Road, rather than from the Khazars. That's what their East Asian-specific mtDNA lineages suggest.

Davidski said...

@zardos

[1] "distance%=4.4674"

Norwegian

Yamnaya,51.8
EEF,35.6
WHG,12.6

[1] "distance%=8.3249"

Lithuanian

Yamnaya,53.4
EEF,26.6
WHG,20

Davidski said...

Balts and Scandinavians have the highest Yamnaya ratios in my analyses, and this is in line with scientific literature.

But to get better fits I need to add more easterly hunter-gatherers and even Siberians, and this does drop their Yamnaya ratios a bit.

zardos said...

Which silk road population do you have in mind for the EA admixture in AJ according to their mtDNA? So you propose instead of Khazar converts, Jewish traders might have taken some Central Asian wives and brought them to Central Europe?

zardos said...

Or a mixed trading post community migrated back and was integrated into the main population. Probably more likely.

Erikl86 said...

@Davidski

"I think distal models can be a useful way of comparing populations, especially when proximate ancestry references aren't available, unknown or very clear."

I see your point and partially agree, however as you yourself just shown, without the historical context this could be highly misleading, which is why I always prefer to use more relevant populations that extremely ancient ones.

As for Yamnaya admixture in AJ, it makes perfect sense because it appears most of the East European admixture in AJs is from Polish-Lithuania, populations with very high Yamnaya ancestry which even 10-15% admixture from it will surely increase Yamnaya ancestry significantly, alongside the South European ancestry Ashkenazi Jews have.

Andrzejewski said...

"
Greek

Anatolia_N,40.6
Yamnaya,35.2
EEF,12.4
Levant_N,6.6
Iran_N,4
WHG,1.2"

Why are EEF and Anatolia_N under two separate categories?

epoch said...

@natsunoame

"because we are the ancestors of the oldest culture and civilisation in Europe."

How can you be ancestor to the oldest cultures, and yet they are long, long dead and you are alive?

I am only half joking here.

natsunoame said...

Obviously the right word is successors here. I can assure you that culture is not dead yet.

Cpk said...

If Ashkenazi are 25% Yamnaya this would mean they are about 40-50% Norwegian-like. Not sure about that.

natsunoame said...

You all seem so confused in your cogitation...Thats what the big lies about european history done in finale. Now, these manipulations some individuals trying to transfer to DNA and archeology discoveries. Just skip half of the truth and the dirty job is done.
There are no mysteries here, everything is already written and we know it even before to see it in black and white... because we are the heirs of the oldest culture and civilisation in Europe.
A quote :
"... Those peoples who, after the end of the so-called" The "great migration of the peoples" are called "Slavs", they are nothing but the remnants of the peoples who had previously inhabited the respective lands and were named by Greeks and Romans in long lists: Thracians, Gethys, Misi, Dacian, Agatharis, etc. . However, these sources of evidence are neglected by our Germanic and Slavicists, or they are ridiculed simply because they do not fit within the pandemist and pan-Slavist tendencies. However, these historians are forced to "wipe out" (along with the flow of the "Great Migration of the Nation") to represent "missing" / almost all the old peoples of the earth and then to make them "appear" newcomers to look for their ancestors and, of course, having put painstaking work without meaning, do not get any results. "- Wilhelm Obermüller, Vienna 1872

No need to mention /because I already did and gave that info/about all the plethora of classics claiming the same. Slavic people are locals but Greeks are not, not at all. Even jews admitted already the fact they are part from their tribes/ the video down below/. Their language now is something like a part of Indo-European languages /may be thats why is staying so lonely there on that tree :)))/ only because of their long cohabitation with the autochthonous tribes. Same with Armenians, but "why" not with their neighbours Georgians. That is only because of the same contact Armenians had with Balkan people their language has changed.
Now about the language in short. The line of development of one language is from synthetic to analytic form. The synthetic language start to change only in contact with the analytic one. And those changes we are talking about are not measured in hundreds but in thousands of years. Only an old and analytic language can be your PIE. About so-called Slavic group of languages I am pretty sure only one is so old and analytic such as structure and grammar. Exactly the one which gave to the others in this group their alphabet, exactly the same gave the first books to them, including on Christian themes. Still staying in French, German, Russian, English....museums but not in where they belong.
Obviously when you mention greek culture or DNA you meant Thracian one. Thats what even the names of the famous cities like Corinth, Perinthus, Mycenae, Athens and others reveal to us, because their etymology is not greek and they can not be explained by this language. Same with their Gods and culture of course. Not to mention about tablets with Linear B containing again not greek names and very old aristocratic generic names leaving no doubt the origin of the same.
About the Roman era you have to look at history sources again, but you have to choose wisely of course. I will mention just one thing: some of the greatest roman emperors were of Thracian origin. One of them gave to his people autonomous church. To the first Christians in Europe actually. That is admitted and written by a Pope, Innocent III / Innocentius III; 1160 or 1161 – 16 July 1216/ in official correspondence.

https://www.youtube.com/watch?v=XObk07uabLI&has_verified=1

Matt said...

@Erik, verbally you may be disagreeing with me, but I don't think your models disagree with me.

Let's say we assumed a model where Medieval_Collegno is about 30% Mycenaean/Empuries2 and the balance is "Iron Age North Italian", then you would arrive at estimates where:

Ashkenazi - 34%, Sephardic - 41%, Italian_J - 45.8%.

That seems closer to my estimate of 35% than a 60:40 two way model of Mycenaean:Levant.

Though I think Collegno is actually not a group to use the average of in these models.(Assuming you are using the rows labelled as such in the latest datasheet).

Consider that many of the samples are rather North European. See individuals and averages on this G25 meta PCA - https://imgur.com/a/IzgYxM6

A more realistic "Iron Age North Italian" is probably more likely to be between HRV_early_IA and Iberia_North_IA. Using Collegno, although understandable, probably underweights "Iron Age North Italian" ancestry.

It probably also means my quick calculation above overweights total Mycenaean like ancestry!

Also consider that Greeks in Crete and the Peloponnese plot mostly in a similar place and show a cline towards Northern Balkan Slavic populations, and both they and their cline aren't intermediate between Mycenaeans and Northern Balkan populations at all.

That must mean a mixture between a population much like Greek Islanders today and Northern Balkan Slavic populations, not that Greece was Mycenaean-like all through the the Roman period until Slavic migrations in Late Antiquity....

This change must have happened between the classical period (if Empuries2 is representative) and late Antiquity (when Slavic migrations seem to begin to affect the Balkans and Greece), which only leaves the Roman Imperial period (which just happens to provide the right components of ancestry for this too).

It also means that the ancestry proportions of total Mycenaean (or classical Greek) ancestry in Greek Crete and Peloponnese can't be too different than you models show for Jewish populations (which is hard for me to call "Aegean Greeks pretty much remained similar to Mycenaeans").

@zardos, well, I'm proposing an admixture between two large populations here (Italic and Greek), so pop density and size issues are not an issue in that case.

Re admixture into Northern Europe many people would argue that population density would mean that many examples of greater admixture into larger long term populations couldn't happen, including the Slavic related one above and we've seen examples of the same argument often applied on this blog in Bronze Age South Asia and Copper Age Western Europe by some partisans ("IVC/Chalcolithic Western European population was too big for any substantial population change!"). And yet here we are.

natsunoame said...

Epoch
Now you can comment substantially not to catch for one word, right?

Andrzejewski said...

They lived in Eastern Europe for over 1,000 years so it is possible

Samuel Andrews said...

@Davidski,

Using more recent sources in G25 I get different EEF/Yamnaya/etc numbers for southeast Europeans.

zardos said...

@Matt: The difference is Romans didnt conquer to destroy the conquered and take all their women, settle their land on a big scale. Especially in the Eastern Mediterranean it was about overtaking existing Hellenistic states. They mainly exchanged the administration.
In Gallia on the other hand they fought tribes in long wars with genocidal effects and a lot of the Roman settlements were created by Italian settlers. The impact must have been much larger.
Without the demographic collaps, the impact would have been much larger because of the urban concentration.

Samuel Andrews said...

Slightly more EEF than Myceneans.

3.9713"

Levant_ISR_Ashkelon_IA1:ASH068

Mycenaean,49.9
Minoan_Lasithi,29.1
Balkans_IA,16.4
Italian_Bergamo,2.8
Tepe_Hissar_ChL,1.8

A few other of the Philitine samples are a Levant/south European mix. Looks like the same kind of south European as this person was. Imo, it would make sense to say we are dealing with a single population from southern Europe possible Crete going to ancient Canaan instead of something more complex like multiple different populations being involved. Also, I doubt Luwains or any population from Anatolia is involved as they probably has significantly more CHG/IranNeo than this sample.

Andrzejewski said...

Why did they separate EEF and Anatolia_N into 2 distinct categories?

Samuel Andrews said...

@Davidski,

I estimate Ashkenazi Jews have 15% Yamnaya ancestry, Romanite_Jews 9%, Italian Jews 14%, Sephardic Jews 11%.

Davidski said...

Yeah, my estimate was probably inflated for whatever reason. Oh well.

Matt said...

@zardos, though, you it seems could say the same about Spain, I'd imagine, and change was estimated at about 26%, which to me possibly seems less than the change from flow between Italy, Greece and Anatolia.

I can't imagine urban collapse would've mattered *too* much to change the picture, though the cities may well have been different.

Reasoning: urbanization is about 25-30% at max in Roman period, and that's an average of much more in Italy, Greece, Western Anatolia and the Levantine and African parts of the empire, and less in France, Spain, Britain, so perhaps something like 15% in France. Even if urban population in France was totally "Roman", then that's only about 15% of the population there. And cities were in theory demographic sinks repopulated from the country...

I doubt we'll ever have a transect of France to make estimates from, or maybe even Gaulish dna!

@sam, I'd suggest to try modelling ASH067. In my reprocessed G25 PCA it sits close to Anatolia_Kamen-Kalehoykuk_MLBA, slightly tilted towards Minoan Lasithi.

Not at all intermediate between Ashkelon_LBA and Mycenaeans/ASH067. Way too CHG and not significantly Levant_N for that. Looks like an Anatolian.

Matt said...

ASH067 vs ASH068 vs ASH066 visualised in the first *3* dimensions of the West Eurasia PCA, not just the first two: https://imgur.com/a/JA0xapc

ASH067 (Mr R1b-M269) has a tilt in dimensions towards CHG and a relative absence of Natufian/Levant influence that places him as like a MLBA Anatolia, whereas ASH068, she looks more like the Mycenaeans (but she's very slightly richer in WHG I'd guess).

The samples here are Central Anatolians so it wouldn't be too surprising for Western Anatolians to be closer to the Aegean populations like Minoans.

Davidski said...

So these Philistines might be of Luwian origin, at least in part?

Matt said...

I'm being brash here, but I would say it seems to like the sample's G25 and West Eurasia PCA position strengthens the possibility.

I'd note that in slight support the original paper's Fig 3, which isn't great because of its choice to model using LevantN+WHG+IranChl (there are reasons they might want to do this, but it's suboptimal), ASH067 gets scads of Iran_Chl and obviously isn't intermediate in ancestry proportions between ASH066 (the hap J boy with the typical Levant profile) and ASH068.

It's a big shame they couldn't get more dna out of the full 108 set of samples they looked at, other than these four buried children in the IA1 set and the six others. That could tell us more about if ASH067/ASH068 were outliers or not. (The presumption is not, but we can't know I guess).

A said...

Apparently the Philistines used some kind of Minoan/Cypriot derived script.

Drago said...

Just tried some plots of my own. 2 things that stand out are

- IA - Anatolia M2197 - East Pole is probably correct - looks Phrygian. Nothing particularly Celtic about it.

- Agree that Asjk-IA1 is pulled towaerd Aegean, genetically, rather than central Anatolian. Again, identity & linguistics is a more complex issue. Not sure on what basis we can be sure he is ''Luwian''. He could even be LBA Sicilian

Drago said...

Plot

Matt said...

NJ tree using pop averages and treating the IA Ashkelon 1 four as individuals: https://imgur.com/a/Ohu5SB3

ASH068 NJ joins with Mycenaean (roughly), but ASH067 NJ joins with BA Anatolian populations. ASH2-3 is on a fairly long branch not close to any population mean today. ASH066 is pretty squarely sitting with Levant populations.

Same data as in NJ tree above used to generate PCA. In dimensions 1 and 2 a simple look could make you think that ASH067 is intermediate Levant populations and Mycenaeans, though he is closest to Anatolian BA - https://imgur.com/a/NHm9tKL (the peril of only looking at two dimensions).

But look in the next dimension up and its clear that he is offset towards CHG and intermediate West Asian populations and Mycenaeans, and not on intermediate the Levantine populations and Mycenaeans - https://i.imgur.com/nGyIAlS.png

Drago said...

back to first principles

Levant_ISR_Ashkelon_IA1
Barcin_N 46.6%
Levant_N 25.8%
CHG 24.6%
EHG 3%
WHG 0%
Iron_Gates_HG 0%

Distance 3.4274%


With a 2: 1 ratio of Barcin / CHG, it is clear this this guy came from somewhere rich in ANF
These ratios are ~1: 1 in the so far data set from central Anatolia E-MLBA

So, at face value; we could be looking at western Anatolia - the Aegean; however given his time frame (IA); this guy is exceedingly western

Open Genomes said...

ASH067 R1b-M269 on the Global25 Ward's distance-squared clustering tree

You can see here that ASH067 who is R1b-M269 clusters right with the Bronze Age Anatolians. It's very likely that he's of Luwian descent, so he could be from the Lukka (Lycians) or the Tjeker (Teucrians / Trojans).

Is this the first evidence of R1b-M269 in Bronze Age Anatolia?

Drago said...

Unless he’s literally straight off the Boat; his non-Levantine ancestry is far too western to be from Anatolia BA.
He’s got ~ 25% ancestry from across the Sea

Drago said...

Indeed; as per the authors themselves
“Similarly, genetic clustering assigns ASH_IA1 with an average of 14% contribution from a cluster maximized in the Mesolithic European hunter-gatherers labeled “WHG””

No chance this guy came from central Anatolia


Romulus the I2a L233+ Proto Balto-Slav, layer of Corded Ware Women said...

So if the IA Philistines cluster with Bronze Age Anatolians, and Bronze Age Anatolians have no steppe, is it the case that they have WHG but no steppe?

Davidski said...

ASH067 does have steppe ancestry, he just wasn't tested for it in the paper, except, of course, on his Y-chromosome.

Bob Floy said...

I've suspected for a long time that the Philistines and "Sea people" in general were Luwian or Luwian-adjacent. If the Philistines were Greek, wouldn't the authors of the bible have been aware of that? They certainly were aware of Greeks. It was never going to be that simple.

Drago said...

That’s true; it’s not that simple. So the idea that Sea Peoples = luwians is simplistic
They were probably a diverse group of people cashing in on the Bronze Age collapse
But this guy doesn’t look Luwian
In fact; if OGs parsing is getting hits for L51 . Did I read it right ?

Matt said...

Modelling a simulated pop with those model proportions above, then dropping onto Davidski's West Eurasia PCA, and another one with slightly enriched Anatolia:CHG ratio (A: 1.85:1 vs B: 1:1) - https://imgur.com/a/qlVTUs5

(Overall only 11% more CHG and 11% less Anatolian, while leaving the EHG and Levant proportions unchanged, so the two sets should have a hugely differnt position).

Visually, the population with richer CHG seems to fit the majority of the Anatolia_EBA+MLBA samples less well than the model proportions Dragos gave for Ashkelon_IA ASH067.

Both are much closer to Anatolia_EBA+MLBA and ASH067 than either are to the sampled Mycenaeans or Minoans. Both ancestry configurations, the one that Dragos gave which also fits well Anatolia_MLBA as well, or a slightly richer level of CHG, fit ASH067 better than any of the samples from populations "over the sea" we have so far. But samples somewhat from the west within Anatolia of where Kamen Katalhoyuk was would likely be even better.

Drago said...

this guy has extra western ancestry; I did not say the entirety of his ancestry is from west of Anatolia
He is getting hits for R1b-L51 even

Bob Floy said...

@drago
"the idea that Sea Peoples = luwians is simplistic "

Yes, that's why I said "Luwian or Luwian-adjacent".
In other words, probably IE speakers from the Anatolian west coast. That's not simplistic.

Davidski said...

Modeling ASH067 with qpAdm is a pain in the ass. He's coming back 100% Anatolia_Kaman-Kalehoyuk_MLBA every time.

I might have to take back my claim that he has genome-wide steppe ancestry, unless Anatolia_Kaman-Kalehoyuk_MLBA also has it, which is not certain of course. Anyway, I'll keep trying and might do a blog post in a couple of days on both ASH067 and ASH068.

Off topic, but I saw a peregrine falcon today, in and around what appears to be a nesting site. Very impressive thing, not so much when just sitting there on the ledge, but flying around and trying to catch other birds.

And a few hundred meters from that falcon site is a goshawk nest in a large tree. Again, very impressive animals to watch when they're prowling around in the air.

Matt said...

Dragos, I would guess ASH067 was probably "fresh off the boat" or the child had at most the equivalent of one grandparent who was local.

Anything like 40-50% requires his other ancestors to be outside the range of present day populations (extremely rich in CHG relative to Iran Chl and Iran N beyond any populations that are sampled and exist at that time). See: https://imgur.com/a/paKNLgj

If ASH067 has only anywhere from 25-30% local ancestry, then his other ancestors have to be pretty close to most of the Anatolian MLBA samples, closer than they are to Minoans or Mycenaeans.

@David, sounds like a cool sight.

Katharós said...

The Hebrew Bible is interesting, because it does not mention language barriers between the Philistines and the Hebrews. Concluding that the story of the Philistines was written down much later and the name Philistine reflected a well-known and defined regional meaning that was already established. I think these Jewish scribes interpreted Philistines (or more precisely the Philistine-coast) in a theological sense, as people who were non-Jews/Goy. The history of Gaza also reflects this aspect, as non-Jewish inhabitants in a regional and theological sense. Interestingly the people of Gaza worshiped their own city God called Zeus-Marnas, an incarnation of Dagon till the Late antiquity. Of course, until they were eventually converted to Christianity, one way or the other. And to the Luwians , I don't necessarily see Philistines as some sort of Luwians, but rather the Luwians as rulers of the Philistines, in sense of a Indo-European elite ruling over a Minoan like population in the Aegean-Sea.

Bob Floy said...

@David

Not too long ago there was a Peregrine hanging out in a city park near my house, in a tall tree. They're awesome.

Aram said...

Here what Damgaard supplements say about Hellenistic period samples were MA2197 and MA2198 were found
---
Context Stratum IIa1–2 (Hellenistic Period Samples)
The Iron Age levels at Kaman-Kalehöyük—including the Hellenistic period—can be divided into
4 architectural substrata from IIa (youngest) to IId (oldest). Substratum IIa can be divided into 5 chronological units based on ceramics. From youngest to oldest these are IIa1–2, IIa3–5, IIab–IIc1, IIc2–3, and IId1–3. In unit IIa1–2 (Hellenistic Period) both human and animal skeletons were found in pits. These fall into 3 different burial types: some containing only animal skeletons, others containing only human skeletons, and some with mixed human and animal skeletons.
Pit P1156 in the North Sector XV: a human skeleton was buried in a flexed position. Human and
animal bones were apparently deposited together deliberately. Such burial features appear only in stratum IIa1–2 and may be correlated with a population change as well as possibly linked to incoming Galatians like at Gordion.

Drago said...

Matt
I guess you’re looking at it from a different angle

The reason why i don't think that ASH067 is from Bronze Age Anatolia is precisely because he ''plots amongst'' them. Taken at face value, one might conclude that he came from there.
However, this fails a basic sanity check- if he is the product of BA Anatolia & Local ancient Palestinian groups, then he'd plot well to the south-east of where he actually lies (red arrow). But for him to be partially Levantine, his other ancestry must be even more western Population X

This can be outlined with some figures


Levant_ISR_Ashkelon_IA2
Levant_N 70.6%
CHG 28.6%
EHG 0.8%
WHG 0%
Iron_Gates_HG 0%
Barcin_N 0%

Anatolia_MLBA
Barcin_N 39.6%
CHG 33.6%
Levant_N 26.8%


Average ANF (Anatolia MBA & IA2 ) = 20%
CHG 30%
Levant = 43 %


But actually,

Levant_ISR_Ashkelon_IA1
Barcin_N 46.6%
Levant_N 25.8%
CHG 24.6%
EHG 3%
WHG 0%
Iron_Gates_HG 0%

This would be a constraint even if he had only 1/4 local ancestry

Davidski said...

@Matt & Bob

The really interesting thing was that there were pigeons sitting and flying around near the falcons, sometimes only something like 10 meters away.

And there was a puddle down below that the pigeons drank from occasionally. But whenever they went down to the puddle, they went as a group, and came back as a group.

It looked very deliberate. I should have filmed it and put it on Animal Planet, or whatever that channel is called, along with David Attenborough's commentary.

Aram said...

Davidski

It is hard to say do Levant_BA has a genuine Caucasian ancestry or maybe Iranian or even non of them. I tried to see what Levant_EBA wants if it has a choice between Iran or Caucasus. It definitively takes Iran, but the fit is not good. .Anatolia was also present there.
The presence of Levant_Chl seriously decrease the level of this Iran input. That is why it is even possible that the apparition of haplogroup J in Levant during Bronze Age was not a result of massive migration from Iran/Caucasus but rather a diffusion from North Mesopotamia. From Hassuna and Halafian complexes. The same can be said about haplogroup L.

[1] "distance%=2.6997"
Levant_JOR_EBA

Levant_ISR_C,70.4
Levant_PPNB,11.6
IRN_Seh_Gabi_C,8.2
IRN_Hajji_Firuz_C,7.6
Kura-Araxes_ARM_Kaps,2.2

Drago said...

Arame
Ok thanks for clarification ; although the individual was not carbon dated (multi level site should be)
But what we have is multiple evidences of groups moving from Europe to Anatolia; and apparently the Levantine coast from the LBA

Bob Floy said...

@David

Pigeons are really smart and very well organized, they're some of the best parents in the animal kingdom, among other things. I don't know why they get so much hate.

Davidski said...

OK, I finally got a hang of modeling these Philistines with qpAdm. Very interesting stuff. Those models in the paper are way too basic.

But it might take me a while to write this up.

zardos said...

Pigeons are just fine as long as there are not too many and they dont come too close. They spread dirt and diseases otherwise.
Also they have a much too small distance to humans in the city. Sometimes they collide with people. The wild ones keep a much larger distance of several meters. That was their urban adaptation.
At least they dont attack people like some crows and raven.
Peregrine falcons are impressive indeed.

Drago said...

Davidski
Where abouts was the habitat. ?

Davidski said...

Just outside a country town in outback central Australia.

Joshua Lipson said...

@Drago

"The reason why i don't think that ASH067 is from Bronze Age Anatolia is precisely because he ''plots amongst'' them. Taken at face value, one might conclude that he came from there.
However, this fails a basic sanity check- if he is the product of BA Anatolia & Local ancient Palestinian groups, then he'd plot well to the south-east of where he actually lies (red arrow)."

Yes, exactly. "He's an unadmixed individual from whatever place his PCA plot location suggests" is a very strange null hypothesis.

Andrzejewski said...

@Katharós " And to the Luwians , I don't necessarily see Philistines as some sort of Luwians, but rather the Luwians as rulers of the Philistines, in sense of a Indo-European elite ruling over a Minoan like population in the Aegean-Sea."

Historians tended to regard the Philistines as migrants from Crete, probably related to the Minoan (Linear B) culture.

I on the other hand have ALWAYS thought of them as Trojan refugees who settled on the coast of Southern Israel following the destruction of the City of Troy, as described in the Iliad: for once, the time frame is really contemporary (Iron Age, 1200 BC, etc), and the most important thing is cognate words such as "Goliath" which are identical with "Alliathes", a Luwian/Phrygian king from the 8th century BC.

Leron said...

If Steppe men could travel to the Levant I don’t see why any Anatolians couldn’t do it either. The Hittites had a naval base to their south and the Lycians were prolific pirates. The headgear of the Peleset/Philistines were an obvious giveaway that they hailed from around southwest Anatolia, like the Carians.

Andrzejewski said...

@Leron like I explained to @Karharós, Philistines may be Trojan Luwian refugees and cognate words like “seren” = tyrannos, “Goliath”= Aliathes and “kupa” = (cap, helmet, from a PIE root) may get me vindicated.

epoch said...

@A

"Apparently the Philistines used some kind of Minoan/Cypriot derived script."

So did the Mycenaeans on Crete.

natsunoame said...

Drago
What do you think about some connection between Tutankhamun's R-M269 and these samples with "probably R1b-M269"?
R.C.Scholz from iGENEA told me :..."the haplogroup R-M269 is more common in western Europe today, but it originated in the Black Sea region about 9.000 years ago. Thats why it is much more likely Tutankhamuns lineage moved from this region to Egypt, not from Western Europe." To add here and the cultural and historical traces is important if we don't want to unnecessary speculate with datas only from DNA samples.

Ben Osland said...

@Zardos

This paper has more on the far-eastern genetical contribution into Ashkenazi.

https://www.ncbi.nlm.nih.gov/pubmed/25669617

TLT said...

Somewhat unrelated, but oh well.

@Davidski

>Balts and Scandinavians have the highest Yamnaya ratios in my analyses, and this is in line with scientific literature.

>But to get better fits I need to add more easterly hunter-gatherers and even Siberians, and this does drop their Yamnaya ratios a bit.

Yeah they have the highest amount of Steppe eneolithic DNA, but what would really happen if you were to add in a Baltic HG component? I have seen some people do it and it doesn't significantly affect the steppe eneolithic ancestry in northwest Europeans like the English for example (since they seem to have very little Baltic HG ancestry). But it reduces the steppe eneolithic component in Baltics and Scandinavians from 50s% to somewhere in the 40s%. Is this true or did they make a mistake in their modeling?

Also, regarding some other populations, I have heard talks about the IE ancestry in populations like the Kalash and Tajiks as being extremely exaggerated, with their real steppe eneolithic ancestry barely amounting to 30% at most (when accounting for an separating away some components closely related to steppe eneolithic). Is this true or false?

Sofia Aurora said...

@Davidski

SPLENDID NEWS GUYS!!!!!

Another Late Upper Palaeolithic individual had her aDNA sequenced!!
I AM TALKING ABOUT THE FAVIGNANA WOMAN FROM AN ISLAND NEAR SICILY!!!

The burial is from the cave "Oriente C" and belongs to the Epigravettian culture.

She had mtDNA haplogroup U2 and share a lot of East European Hunter-gatherer ancestry with part from WHG.

The study has also morphological analyses and it is a prelude to a bigger paper by Fontana about the genetic landscape of the Epigravettian complex that engulfs Sicily and generally the Mediterranean to the eastern part of the Balkans, SW Anatolia and up to the border if the Black Sea.
The age of the individual is 14200 BP and there is also an analysis of other specimen with it.

Below is the link:

https://www.biorxiv.org/content/10.1101/692871v1

Bob Floy said...

@natsunoame

"What do you think about some connection between Tutankhamun's R-M269 and these samples with "probably R1b-M269"?

Was it ever confirmed that Tut had M269? I don't think it was.

Matt said...

Josh Lipson: "Yes, exactly. "He's an unadmixed individual from whatever place his PCA plot location suggests" is a very strange null hypothesis."

Joshua, do you make this same argument for ASH067, that it is wrong to describe her as likely she is an unadmixed, Mycenaean-like migrant from the Aegean, and that she is likely instead the admixture of a vastly more northern population with an Ashkelon_LBA like population?

Is that she is an unadmixed Mycenaean-like individual a "very strange null hypothesis"?

I suspect that you don't, and the reason you don't is that it is implausible to suppose an unsampled population well beyond the range of sampled variation in the period in question (and in fact likely beyond any sampled variation) to admix with Ashkelon_LBA to produce ASH068.

In my opinion, and supported by the plots I've produced upthread, the same is true of ASH067, when you look at the affinities of CHG vs Iran_N in the plots. No existing population at the time can mix with Ashkelon_LBA to produce a position with the right distances to CHG relative to IranN and IranChl.

(and this is an odd thing about Dragos post which you're quoting: "Levant_ISR_Ashkelon_IA2 - Levant_N 70.6%, CHG 28.6%".... but Levantine post-Neolithic populations are in all the academic models admixed between IranN and LevantN... not CHG. It's an example of conflating the CHG and Iran_N, exactly what most posters on this blog vociferously complain about when it comes to examining the Yamnaya!)

I'll be interested to see if Davidski's qpadm models support my contention here, or reject it....

Samuel Andrews said...

@Sofia,

Wow, thanks for sharing! This confirms all of Italy was WHG territory.

A possibility based on mtDNA is......
Its mHG U2'3'4'7'8'9 is native to Italy going back 30,000 years. It isn't found in any mainland European WHGs but is found in Paleolithic Italy. This would indicate although the Sicily WHG is basically identical to mainland WHGs, it also has local ancestry it doesn't share with mainland WHGs.

No dots are connecting so far. But, I suspect Italy was the launching pad for Loschbour-like WHGs who took over western Europe after 15,000 years ago.

natsunoame said...

https://www.igenea.com/en/tutankhamun
If you have more wide view on the theme including cultural, historical connections, artifacts with strong parallel as for sample ceramic but these from the Balkans are 2000 years older. Similar rituals (Asia minor, Crete, Perperikon) from the time of Akhenaten as well as artificial cranial deformation.
On the ceramics of the first three dynasties were found signs whose origin was not yet established. It is clear that they are not hieroglyphics, they are thought to be like markings of pottery workshops or craftsmen made crates.

Interesting in this case is that a large number of these signs are present around 2000 years earlier in the Balkans. This is the so-called Danube or Old-European script. Even more interesting is the fact that a number of signs have quite clear parallel with the Old Bulgarian runes from Pliska and Preslav.While in Egypt mysterious signs disappear from use after the First Transitional Period - the end of the third millennium BC (2170-2040 BC), at Balkans the tradition goes back to the Middle Ages.
There are still evidences of the influence of old Balkans on the population of Egypt. It is about common or close words. Since the languages ​​are of a completely different origin, the parallels can not be explained by some common root of the distant past.
In addition to Robert Boval, there are other scientists who believe that in the distant past there have been migrations of Europeans to North Africa, and more specifically to Egypt. Prof. Ian Best and Sibila von Reden, and others. believe that people from the so-called Kurgan culture have even reached Egypt. As I can see iGenea scientists are on the same page like it or not.
And you should tell me indication for what is the genetic cocktail of the Copts. What about the Coptic alphabet?!
Let me save you that last check which will bring you to the wrong path again about the ancient’Greeks ...You can check alone who is Michael Everson. So, by chance, in his sphere of knowledge fell the following languages: the Slavic, the Ancient Greek and the Coptic.
working with apps that are based on the Coptic-Greek language link, translates the same caption from Coptic to Greek and Cyrillic.
The result:
In Greek the text is unreadable, in Cyrillic it is readable.
The conclusion of the author:
"A Coptic reader would have found the Cyrillic text far away - far easier to read than the Greek one." I do not propose Coptic to unite with the Cyrillic alphabet, but I suggest that the link with the Greek be revoked.“

Marko said...

@Davidski and all
Have you seen this paper yet? It appears that proto bulgars were very similar to Alans and left a substantial genetic impact in modern Bulgarians. Also some mention of "bronze age thracians" but I think this might refer to Bulgarian Yamnaya

https://www.biorxiv.org/content/early/2019/07/04/687384.full.pdf

"We suggest that contemporary Bulgarians are an admixture of ancestral Slavonic groups, rich on locally absorbed EEF DNA and Proto Bulgarians, rich on Caucasian DNA and genetically related to the bearers of the Saltovo Mayaki Culture from 6 to 8 century AD"

"Bronze Age proto Thracians are genetically closer to early medieval Slavs (represented here by Croatian samples) than to contemporary Bulgarians and their influence on Bulgarian population genomics is not direct, but is probably mediated by early Slavs"

"Unlike BA Thracians and the early Slavs, SMC carry substantial Caucasus admixture, related to the tribes from Bronze Age Armenian plateau and seems to have transmitted this admixture to the contemporary Bulgarians... Both Scythian samples from the Hungarian steppe and the Alans from Saltovo-Mayaki culture bear strong genetic resemblance to the Bronze Age Caucasian samples, which is missing in central Asian nomads but is presented in the contemporary Bulgarians"

"However, contemporary Bulgarians have received their Minoan component mostly through population exchange with Byzantium and their Bronze age Thracian component trough admixture/population exchange with early medieval Slavs and Croats. The signal that distinguished contemporary Bulgarians from the other Balkan nations is the unique signature of SM-Alan people, who appear amongst the direct precursors of contemporary Bulgarians"

Simon_W said...

@Desdichado
"Modern political and social considerations are not pleased, I imagine. Probably why we're still seeing nothing on the Tollense DNA, for that matter. It's frustrating sometimes."

A little add-on on this: In Eastern Germany you see right-wing riots if a German is killed by a Middle Eastern refugee. In Western Germany we've just seen a green riot against the exploitation of brown coal. But riots because of a scientific paper? No way! :D And anyway, it's a well known, established fact that the Germans of Mecklenburg-Vorpommern are derived from Medieval migrants from western Germany who settled in Slavic territory. Does this make them feel foreign there? Of course it doesn't. So how could evidence about a non-Germanic affiliation of some Tollense samples change this at all?

Davidski said...

@Marko

I think this quote shows the value of that paper...

Bronze Age proto Thracians are genetically closer to early medieval Slavs (represented here by Croatian samples) than to contemporary Bulgarians and their influence on Bulgarian population genomics is not direct, but is probably mediated by early Slavs.

Croatians = early medieval Slavs? Nope.

natsunoame said...

Marko
Is this misunderstanding and contradicting statements are copy/paste from a different articles or from one only?

Davidski said...

@natsunoame

https://www.igenea.com/en/tutankhamun

This is BS. There's no evidence that Tutankhamun belonged to M269.

Marko said...

@Davidski
They are loose with their terminology, in another part of the paper they mention Croatians are a stand in for medieval Bulgarian slavs which makes the choice less peculiar. Do you find fault with the general hypothesis here that there is an extra layer of Caucuses like ancestry within modern Bulgarians that was mediated from an Alan like population during the iron age?

@natsunoame
They're all quotes from the same paper that is linked. Which statements did you find contradictory? They aren't immediately apparent to me

Simon_W said...

As for the Europeanness of Ashkenazi Jews... I suppose in the 19th century most of the assimilated, non-orthodox Jews loved Europe and considered themselves to be Europeans, but they were not accepted because of anti-Semitism, which eventually gave rise to the Zionist movement which culminated in the creation of Israel, and now many regard this as their home, even if they're still in the diaspora.

The question if Ashkenazi and other western Jews are European or Near Eastern (or the exact percentage of their Near Eastern ancestry) is regarded by many people as a highly important question in the Israel-Palestinian conflict, though in fact and in practice it's completely irrelevant, because either way, the killing or expulsion of Israeli Jews is quite obviously a no-go and the voluntary dismantling of Israel is not a realistic option.

Funny though how some well-meaning, politically hypercorrect non-Jewish people assert that being Jewish is only about a religion, and that there is no Jewish people. (Which is also the position of the Palestinians, I read.) Often you also hear the phrasing "he's the son of Jewish parents" to avoid the supposedly offensive wording "he's a Jew". These views have been thoroughly refuted by DNA research.


Davidski said...

@Marko

Do you find fault with the general hypothesis here that there is an extra layer of Caucuses like ancestry within modern Bulgarians that was mediated from an Alan like population during the iron age?

I do, because it'd be pretty easy to tell if Bulgarians experienced an extra pulse of Caucasus ancestry compared to their neighbors, and I'm not seeing that.

All Southeastern Europeans have more Caucasus ancestry than other Europeans, but this didn't come from the steppe, but rather via Anatolia in several waves from the Late Neolithic to the Middle Ages.

Simon_W said...

@Zardos

"In Gallia on the other hand they fought tribes in long wars with genocidal effects and a lot of the Roman settlements were created by Italian settlers. The impact must have been much larger."

There were the colonies where Roman war veterans were settled. But even there the Gaulish locals were in the majority, the veterans were placed there to Romanize them. I've seen estimates on the percentage of the Roman element in the colony of Augusta Raurica. I don't remember the exact number, it must have been something like 10-20% or so. But ancient DNA evidence would be welcome to confirm or refute this.

Andrzejewski said...

Simon_W "As for the Europeanness of Ashkenazi Jews... I suppose in the 19th century most of the assimilated, non-orthodox Jews loved Europe and considered themselves to be Europeans, but they were not accepted because of anti-Semitism, which eventually gave rise to the Zionist movement which culminated in the creation of Israel, and now many regard this as their home, even if they're still in the diaspora."

Right. And lots of non-Orthodox Jews in Germany, Hungary, France, Czechia but much less in Poland, Ukraine, Romania, Russia etc. intermarried with their "gentile" neighbors, with pre-WW II German Jews having a non-Jewish parent, spouse or offspring at a rate of at least 50%. These people did not necessarily regard or consider themselves "Jewish" and nor did the Orthodox Rabbi but Hitler and his ilk did, so therefore many German Jews with their mixed heritage and mixed religion spouses packed up and move up to Israel, back known as "British mandate Palestine".

Likewise, in the Soviet Union Jews were being severely discriminated against and oppressed, and all religions in general were banished from the public sphere. Therefore many Jews intermarried with their non-Jewish neighbors, and if the Soviet Union didn't collapse I doubt there would be one Jew left by 2030. Alas, that event propelled many Jewish Russian to immigrate to Israel, where the existence of one grandparent is enough to be eligible, they packed up and took their non-Jewish spouses with them and their offsprings may have very few Jewish blood in them.

Why am I bringing it up? Because it explains what @Davidski published, that Ashkenazi Jews have 25% Yamnaya. Makes lots of sense.

Drago said...

Matt

Your plots haven't demonstrated anything, nor have you articulate anything of value. That I only used CHG doesn't change anything. Iran N can be added it shows the same thing
It shows IR2 is more or less a continuity of LBA, and Ash-IR1 lies West of Anatolian BA


You are also ignoring the fact that this is a time of attested movements from Europa to the Near East (or perhaps you simply don't know & have put no real effort in understanding).

If you wish to support your position- produce isotopic evidence . I won’t hold my breath.

Drago said...

@ Marko
Agree with Davidski on that
The conclusions arent fully sound. Seems like another paper from Bulgaria trying to disattach Bulgars having anything to do with Turkic

natsunoame said...

Drago
Where do you see any attachment between Turks and BulgARIANS, Respectively from BulgARIA. You are free to check the oldest name of Thracia. On what level you see those connections, linguistically, anthropology, DNA, cultural, Religion? There is only historical bloody one of 500 years slavery under Ottomans, that’s why your remarks are close to delusion...
https://www.researchgate.net/publication/265289655_Y-Chromosomal_haplogroups_in_Bulgarians

Samuel Andrews said...

@Andre,

The Ashkanzi reference group used in DNA tests are pure blood. About half of their ancestry is from ancient isreal. Their European ancestry is mostly southern European not recent mixture with eastern Europeans & Germans. Overall, they are 15% Yamnaya.

Samuel Andrews said...

@TLT,
"Also, regarding some other populations, I have heard talks about the IE ancestry in populations like the Kalash and Tajiks as being extremely exaggerated, with their real steppe eneolithic ancestry barely amounting to 30% at most (when accounting for an separating away some components closely related to steppe eneolithic). Is this true or false?"

That's correct. Steppe/IE ancestry in Kalash & Tajik was exaggerated before we had ancient DNA from south/central Asia. It first seemed they had as much as north/central Europeans (so 50%). The reason it was exaggerated was because Kalash & Tajik have some ANE ancestry from 'Neolithic' central Asia.

Now, it looks like Kalash have about 30% Andronovo, Tajick have about 45% Andronovo. In comparison most Indians have 15% Andronovo, Indian Brahmin have 30%.

In circa 400ad (before Turkic expansions), most of central Asia was inhabited by Iranian-speakers with 30-50% Andronovo ancestry. Tajicks today are an isolate today but there used to be lots of people in central Asia similar to them.

Samuel Andrews said...

@TLT, Andronovo is 70% Yamnaya. So, do the math, you see the central Asians with the highest Yamnaya ancestry have as much as the Europeans with the least Yamnaya ancestry.

Samuel Andrews said...

Prendergast 2019. July 5, 2019.
Ancient DNA reveals a multistep spread of the first herders into sub-Saharan Africa
https://science.sciencemag.org/content/365/6448/eaaw6275/

TLT said...

@ Samuel Andrews

>That's correct. Steppe/IE ancestry in Kalash & Tajik was exaggerated before we had ancient DNA from south/central Asia. It first seemed they had as much as north/central Europeans (so 50%). The reason it was exaggerated was because Kalash & Tajik have some ANE ancestry from 'Neolithic' central Asia.

>Now, it looks like Kalash have about 30% Andronovo

30 x 0.7 = 21% Yamnaya. That was quite a lot of over-estimation in the older research. I assume that it has something to do with them just using ANE as a proxy for steppe ancestry, right? Is it true that both EHG and EHG descendants are not nearly as ANE-rich as expected either?

Andrzejewski said...

@Sam "That's correct. Steppe/IE ancestry in Kalash & Tajik was exaggerated before we had ancient DNA from south/central Asia. It first seemed they had as much as north/central Europeans (so 50%). The reason it was exaggerated was because Kalash & Tajik have some ANE ancestry from 'Neolithic' central Asia."

What "ANE ancestry from 'Neolithic' Central Asia" are you talking about?

The only one I could think about is via Yamnaya and Sintashta.

Davidski said...

@TLT

Early farmer ancestry from Anatolia and the South Caspian region mixing with Botai-related ancestry in Central Asia (probably from Kelteminar and similar groups) produced a very similar effect to Yamnaya and Sintashta ancestry, because at a very basic level most of the ancient components were the same.

The only way to get around these sorts of issues is to sample the most proximate ancient sources of gene flow as possible, but even then there will be some ambiguity in the results due to the fine scale of differences involved.

TLT said...

@Davidski

Thank you for the information. So these Botai guys are pretty much similar to the western Siberian HGs, correct? Compared to EHG, how much ANE does WSHG have exactly?

Bob Floy said...

@Andre

Iran_Neo is like 40 or 45% ANE.

Samuel Andrews said...

@TLT,
"30 x 0.7 = 21% Yamnaya. That was quite a lot of over-estimation in the older research. I assume that it has something to do with them just using ANE as a proxy for steppe ancestry, right? Is it true that both EHG and EHG descendants are not nearly as ANE-rich as expected either?"

In a previous paper (before there was ancient DNA from southcentral Asia), Yamnaya and IranNeo and Onge were used to model southcentral Asians. The reason their Yamnaya score was exaggerated was Yamnaya was forced to swallow up the ANE inherited from central Asian hunter gatherers.

According to Davidski's G25 PCA....
EHG is 70% ANE.
West Siberia HG is 84.7% ANE
Botai is 80.5% ANE.
Kazakh_steppe HG is 57% ANE
Dali HG (also in Kazakhstan) Is 68% ANE

Neolithic farmers from Tajickstan had about 20% extra ANE from unsampled hunter gatherers from central Asia. BMAC farmers had about 10-15% extra ANE. They had no east Asian ancestry however unlike Kazkah hunter gatherers who had signifcant East Asian ancestry.

It was mostly a mix between SC Asian farmers like that who mixed with Andronovo which created modern Tajik and Kalash and Pathan. Kalash & Pathan also have significant South Asian ancestry.

Davidski said...

@TLT

So these Botai guys are pretty much similar to the western Siberian HGs, correct? Compared to EHG, how much ANE does WSHG have exactly?

Yes, the Botai people appear to have been derived from a population of West Siberian foragers who moved south into the Kazakh steppe, and this type of ancestry was at one stage found at significant levels throughout Central Asia, and even all the way from the North Caucasus to Mongolia and south into India.

But I don't know how much ANE it has. I haven't looked at deep Eurasian ancestry components for a long time. I don't find these sorts of topics as interesting as I used to, mainly because they're so abstract.

This post and comment thread might be useful to check out...

Some myths die hard

Andrzejewski said...

@Sam "Neolithic farmers from Tajickstan had about 20% extra ANE from unsampled hunter gatherers from central Asia. BMAC farmers had about 10-15% extra ANE. They had no east Asian ancestry however unlike Kazkah hunter gatherers who had signifcant East Asian ancestry.

It was mostly a mix between SC Asian farmers like that who mixed with Andronovo which created modern Tajik and Kalash and Pathan. Kalash & Pathan also have significant South Asian ancestry."

Are you talking about Kelteminner and Botai when you talk about these unsampled HG from Central Asia? And aren't BMAC mostly Iran_N not unlike the Elamites? Floy said that Iran_N was 45% ANE so maybe the extra ANE did NOT come from Botai or Keltemminer?

Bob Floy said...

@Andre

I think it varies, but somewhere around 40% IIRC.

@David

How's that analysis of the Philistines coming along?
This is exciting.

Matt said...

@Drago, when we've been discussing

a) specifically sample Ashkelon IA1 ASH067,

b) my contention is that higher dimensions that capture differences between Iran_N and CHG and between Barcin and LevantN, and models sensitive to these distinctions, will show that ASH067 is not admixed between Aegean (Mycenaean) and Levant LBA

then showing a two dimensional plot with an average of all the heteregenous Ashkelon IA1 samples is not exactly going to move the argument forward at all. (for some reason in the same plot as plotting individual samples for other populations which are far less hetereogenous?)

Buried infants are also unlikely to show isotopic differences from locals, even if all their ancestors were foreign.

and you must know this, so it seems hard to think you're arguing in good faith at this point.

Let's see what Davidski does with models that are more sensitive to the distinctions between Barcin and LevantN and between CHG and IranN, than merely glancing at the first two dimensions of PCA, or present LevantLBA as admixed between LevantN and CHG.

Matt said...

E.g. whether Davidski's qpadm models still show anything like "(ASH067) coming back 100% Anatolia_Kaman-Kalehoyuk_MLBA every time" or anything fairly close to that and closer than a two way Aegean+Levant LBA mix.

Davidski said...

This might take a while. I'm in the back of a car and will be for the rest of the evening.

Drago said...

Davidski
Which of the 4 pink crosses are the IA -1?

Drive safe !

Shaikorth said...

@Andrzejewski

Sarazm_eneolithic & related populations pretty clearly had West_Siberia_N or Botai-type ancestry that can't be explained with something from earlier Iran, or by EHG, WHG etc. since these were used in their distal modeling.

This is the qpAdm model listed for Sarazm, unlike with many other neolithic populations from Iran and eastwards there were no alternatives:
P-value 0.192
Iran_N 75% Anatolia_N 3% West_Siberia_N 23% (adds to 101% due to rounding of fractions)

Tepe_Hissar_CA (from Iran where WHG-related and Anatolia_N-related ancestry was more common) was the only one of that bunch which could be modeled with Karelia_HG instead of WSHG but that fit was worse than the one with WSHG, in other words ANE increases with distance from Iran which is the opposite of what we'd expect if BMAC's ANE is from ancient Iranian populations.

Matt said...

@Dragos: It's reasonable to suggest that the sample has very slightly more Barcin related ancestry than the Kaman_Kalehoyuk samples, which are not from the west of Anatolia, of course (central).

But accounting for that, it is still closer to EBA and MLBA Anatolians that are sampled than any of the Greek samples, just as the neighbour joining analysis shows.

Even taking into account any local Israel Ashkelon LBA will not greatly change this, because it must be relatively small. (As above there is no particular reason to suspect a simply a larger admixture, and to model with large Ashkelon LBA admixture would seem to require another population beyond the plausible range in IranN vs CHG and BarcinN vs LevantN affinity).

My main contention here is that ASH067 is not intermediate Mycenaeans and Lebanon LBA, and cannot be modeled as admixed between them, nor really as having any great Lebanon LBA admixture at all. Accepting that, we have to look for other reasons for the sample's positions in PCA (in all dimensions).

Beyond that, as I've mentioned above, I am not saying for sure where this sample's ancestors originated, just that from all the evidence, western Anatolia looks most consistent to me. People can readily suggest that Anatolian_MLBA like people (with very, very slightly more Barcin related ancestry) like the sample lived elsewhere, but I don't see any strong evidence for this (how does LBA Sicily or eastern mainland Greece or whatever place within "Europa" seem more likely to have the right mix of ancestral components?).

(@All, replying to a deleted comment here)/

Drago said...

Matt
Was just waiting for first Davidski to reply
Anyhow, my contention was not that IA-1 is “intermediate Mycenaeans and Lebanon LBA, and cannot be modeled as admixed between them”, (his ancestry must be more complex than that) -
but simply that he’s more western than the body of anatolians which existed in MBA Anatolia

Davidski said...

@Drago

All of the pink or rather red crosses are Ashkelon_IA1. But I'm not sure which one is ASH067.

The datasheet is here tough if anyone wants to check.

https://drive.google.com/open?id=1II2FpgoOsRqVG2qWS1LgqlC5RwknKtdd

This can be done with PAST 3.

https://folk.uio.no/ohammer/past/

By the way, you're probably right that across multiple dimensions ASH067 is more western than anyone in BA Anatolia that we have available.

However, as far as I can tell, what Matt is saying is that as yet unavailable samples from western BA Anatolia might be more like ASH067.

I do have some very interesting qpAdm results in the context of this discussion, and I'll post them as soon as I can. But I'm not sure if they'll solve anything one way or the other, because they're showing a more complex picture IMO.

Drago said...

Dave
Thanks for clarifying; that’s what I thought ; and your plot is revealing. Heck, one of the IA1 is even more northern than Mycenaeans (!)

“what Matt is saying is that as yet unavailable samples from western BA Anatolia might be more like ASH067”

I have no doubt that there will be late bronze & Iron Age individuals from western Anatolia; we already know this , thanks to MA -2197; who sits with Bronze Age Balkans.
I also conclude that Barcin chalcolithic (3800 BC) is not along that cline . So we have at least some time constraints

Drago said...

During the LBA / IA transition, people from as far as Central Europe were coming to the Mediterranean ; but we also already know that from the Italian abstract

Matt said...

Davidski: By the way, you're probably right that across multiple dimensions ASH067 is more western than anyone in BA Anatolia that we have available.

Than populations averages, yes (individuals, less sure - MA2212 from Anatolia_Ovaoren_EBA looks about as "western farmer" and they or MA2213 from the same population look like ASH067's closest neighbour).

However, as far as I can tell, what Matt is saying is that as yet unavailable samples from western BA Anatolia might be more like ASH067.

Yes, with the caveat that the Barcin shift of ASH067 compared to Anatolia_MLBA samples isn't really very large (it's not like ASH067 is near Minoans and is way south of Mycenaeans).

ASH067 seems to me he can't be explained by having an an ancestor proxied by Aegeans (whether Mycenaean or Minoan) and significant admixture from Ashkelon LBA, because that looks to break affinities to CHG relative to IranN and LevantN.

It doesn't seem to me it can be explained by a ghost population combining with Ashkelon LBA either, because the required specific CHG affinity relative to IranN and LevantN affinities for that ghost would be outside of the clines seen and samples known in the West Eurasia PCA and Global 25 (and are probably not possible from our understanding of the phylogentic tree structures of how CHG, Barcin, IranN, Levant came to be).

So a Kaman Katalhoyuk Anatolia_MLBA like set of ancestors with slightly more Barcin N ancestry than their average seems like the best for this sample. I suggest that a reasonable place and time for them to come together is Anatolia west of where Kaman Katalhoyuk is (assuming a cline between Kaman Katalhoyuk and Minoan Lasithi, which seems a fairly reasonable assumption); it is possible this set of ancestors could coalesce somewhere else, it just seem less likely/obvious.

Matt said...

Here's a visually simpler implementation of what I am expressing here, plots based on West Eurasia PCA 1-3, and I've done nothing to it other than flip PC1 so "north" is more positive: https://imgur.com/a/GUo6GiQ

PC1 vs PC2: You can see the individual position of each IA1 sample; ASH068 is like Mycenaeans but slightly more "northern" (or rather slightly more shifted along the Minoan->Mycenaean path), ASH066 sits with Ashkelon LBA and IA2, and the other two including ASH067 look intermediate.

PC1 vs PC3: On PC3, which specifically tracks CHG+Barcin affinity vs WHG+IranN+LevantN affinity, the impression of ASH067 looking intermediate ASH068 and ASH066 vanishes. You may be just about able to model ASH067 using Minoans and the most CHG shifted Ashkelon LBA, but its a stretch (edit: the proportions for Minoan+mostCHGAshkelonLBA don't look viable to me actually - ASH067 is about 50:50 between Minoan and the most CHG shifted Ashkelon LBA in 1v2, but only 66:33 in 1v3).

zardos said...

@Ben: Thank you. I wonder how reliable mtDNA is for determining that it was direct Chinese admixture.

@Sam: I think most underestimate the Italian influence from Roman times in the West. However, only extensive testing will solve this and we need sufficient data from all participating historical groups.

Bulan Goldstein said...

@natsunoame oh, give us a break. Not just the founders and rulers of the First Bulgarian Empire but the rest of the dynasties until Ottomans: Asen, Terter and Shishman dynasties were Turkic (Cuman) in origin.
Would you like the explain Tangra worshipping of original Bulgars or Cuman Turkic names in Asen, Terter and Shishman dynasties?

TLT said...

@Davidski
Thank you for the link.

@Samuel Andrews
>"In a previous paper (before there was ancient DNA from southcentral Asia), Yamnaya and IranNeo and Onge were used to model southcentral Asians. The reason their Yamnaya score was exaggerated was Yamnaya was forced to swallow up the ANE inherited from central Asian hunter gatherers. "

So basically half of south Asian ANE comes from IE people and the other half comes from these WSHG-farmer mixed people, right? I have read that certain south Asian populations have elevated frequencies of haplogroup Q compared to most other south Asians, perhaps this could be because of that WSHG impact.

Open Genomes said...

@David @Matt

Compare ASH068, the most "Western" of the Iron Age 1 Philistines, with ASH067, the R1b-M269 Iron Age 1 Philistine:

Global25 nMonte ancestry composition of sample: ASH068 Population: Levant_ISR_Ashkelon_IA1 Iron Age Near East

Global25 nMonte ancestry composition of sample: ASH067 Population: Levant_ISR_Ashkelon_IA1 Iron Age Near East

ASH068 has zero Levantine ancestry, and is 25.6% Late Bronze Age Minoan Lasithi. Interestingly, ASH068 is 3.0% Kubano-Tersk. Here is your Steppe ancestry among the Mycenaeans.


ASH068 on the other hand has 8.8% Ashkelon Late Bronze Age ancestry, so he is a bit admixed with the Bronze Age locals. However, he has 45.0% Peloponnese Neolithic, 21.2% Iran Seh Gabi Chalcolithic, and 9.4% Kura-Araxes.

ASH067 on the Global25 Ward's distance-squared clustring tree

Since it's invalid to compare an individual with those in the same cluster that they are in, the breakdown of non-Late Bronze Age Ashkelon components is good for a Bronze Age Anatolian. As has been pointed out, the very high Iran Seh Gabi Chalcolithic doesn't fit Minoans, Mycenaeans, or Levantines. The very high Peloponnese Neolithic is a proxy for the Western Anatolian Neolithic.

The "problem" for people here seems to be that he's R1b-M269 and doesn't show obvious signs of Steppe ancestry. A possible explanation is that this Steppe ancestry is very early and did not come down through the Balkans, unlike what it seems for ASH068, but rather could have come down through the Caucasus into the Kura-Araxes culture around the 4.2 Kiloyear Event (2200 BCE) or earlier where it got diluted on the way westward from Northwest Iran to Western Anatolia.
The Indo-European Anatolian language (Luwian?) would have been imposed by elite dominance across Anatolia, as happened later with Greek and Turkish. (Even the Cappadocian Greeks spoke and wrote Turkish, albeit with the Greek alphabet.) Even today there is a relatively high percentage of R1b-M269 across Anatolia.

Remember too that the Late Bronze Age - Early Iron Age Levantine I2062 is R1b-M269 and is ancestral for R1b-L23. (This makes him extremely likely to be R1b1a1a2b-PF7562.)

Global25 nMonte ancestry composition of sample: I2062 Population: Levant_ISR_MLBA Bronze Age Near East

I don't see direct Steppe ancestry for him, but while he's heavily admixed with the Chalcolithic and Bronze Age Levantines at 63.8%, he has Darkveti-Meshoko at 17.8% which also points to North Caucasus ancestry, mixed with 14.4% Iran Neolithic and Chalcolithic ancestry. While much of this is common to Middle and Late Bronze Age Levantines, the result of the Amorite migrations from the north, it seems that I2062 has an excess of North Caucasus ancestry compared to the other Middle and Late Bronze Age Levantines. Regardless, he still clusters with Late Bronze Age and later Levantines. If his ancestors had Steppe ancestry, it was completely diluted by the time of the Late Bronze Age / Early Iron Age.

a said...

Is ASH067 connected to the sample found around Firuzz Teppe and F38?The Iron age sample F38 branches from Iran has interesting connections
https://www.yfull.com/tree/R-Y19434/

Andrzejewski said...

@Samuel Andrews @TLT "So basically half of south Asian ANE comes from IE people and the other half comes from these WSHG-farmer mixed people, right? I have read that certain south Asian populations have elevated frequencies of haplogroup Q compared to most other south Asians, perhaps this could be because of that WSHG impact."

The Burushu, speakers of a language isolate in Northern India are a mixture of BMAC and Botai with some Caucasus and Iran_N: they have y-haplogroup L, Q, E1b1b and H. Their language does not seem related to either Elamite or Dravidian so it's hard to tell if it came directly via Iran_N farmers. On the other hand, BMAC has a lot of Iran_N blood so it could be the descendant language of BMAC?

Leron said...

I don’t see the need to shoehorn some kind of deep Steppe ancestry: “they had it originally, it just got very diluted.” Occams Razor says they didn’t have Steppe in the first place.

Erikl86 said...

@Andrzejewski,

You said:

"Why am I bringing it up? Because it explains what @Davidski published, that Ashkenazi Jews have 25% Yamnaya. Makes lots of sense."

First of all most German Jews perished, and did not make it to Israel, and the ones that didn't perish mostly migrated early enough to the United States. Soviet Jews didn't only migrate to Israel, they also went to the US.

The ones which left the USSR back in the 1970s, which included my father, mostly chose Israel and didn't intermarry with non-Jews. The ones that left after the collapse of the USSR in the 1990s did have some intermarriage, in fact out of the 1.2 million Soviet Jews which have arrived to Israel, about 250,000 are intermarried to some degree (from 50% intermarried to having only one Jewish grandparent, but the latter is rare). So the intermarriage weren't that common, and most that did extensively intermarried most likely forgot their Jewish identity.

In any case, if I try to model other Western Jewish populations with the same populations Davidski modeled Ashkenazi Jews here, this is what I get:

Davidski model:

Ashkenazi_Jew

EEF,29.2
Levant_N,26.6
Yamnaya,24.2
Iran_N,12
Anatolia_N,6
Iberomaurusian,1.2
East_Asian,0.8

My runs:

Romaniote Jews

EEF,21.4
Levant_N,29
Yamnaya,15.6
Iran_N,11.6
Anatolia_N,21.4
Iberomaurusian,1
East_Asia,0.4


Sephardic Jews

EEF,26.8
Levant_N,20.6
Yamnaya,17.2
Iran_N,11
Anatolia_N,21
Iberomaurusian,3.4


Italian Jews

EEF,29.8
Levant_N,25.6
Yamnaya,19.2
Iran_N,8.4
Anatolia_N,15
Iberomaurusian,2

^^ As can be seen, the Yamnaya admixture in Italian Jews is 19.2 vs. 24.2 in Ashkenazi Jews.

Considering historically Italian Jews are considered ancestral to Ashkenazi Jews, that means that Ashkenazi Jews merely gained 5% more Yamnaya admixture (of their total ancestry of course). Descendants of Italian Jews which gained additional 10-15% Polish-Lithuanian admixture in Eastern Europe, would certainly account for an increase from 19.2 to 24.2 Yamnaya admixture.

Andrzejewski said...

Why would Sephardic (Middle Eastern) Jews have 17% Yamnaya?

Erikl86 said...

@Andrzejewski,

Most likely from the same source the mutual ancestor of all Western Jews gained it. Even Syrian Jews seem to have it:

Syrian Jews:

EEF,17.8
Levant_N,30.4
Yamnaya,13.8
Iran_N,18.8
Anatolia_N,17.8
Iberomaurusian,1.4

Erikl86 said...

Also @@Andrzejewski, just to add that Sephardic Jews aren't any more Middle Eastern than Ashkenazi Jews (well, maybe in the sense of having less Northern European admixture), and their ethnogenesis is in Iberia, Western Mediterranean, but they most likely mostly descend from Roman Jews also and the first Jewish communities in Iberia are documented in Northeastern Iberia, so most likely they've migrated to Spain from Northern Italy (similar to Ashkenazi Jews) only they've then migrated West via S. France rather than North to Northern France and later German Rhineland.

Sephardic Jews have arrived to the Middle East only after 1492, when they were expelled from the Iberian peninsula by the infamous Alhambra decree and were largely invited by the Ottomans to settle in Turkey, the Balkans and the Levant.

Others mostly fled to North Africa were they married and largely absorbed pre-existing non-Sephardic Jewish communities in Morocco and Algeria.

Davidski said...

@All

Back on topic. No more discussions about the history of European Jews.

Davidski said...

@Leron

I don’t see the need to shoehorn some kind of deep Steppe ancestry: “they had it originally, it just got very diluted.” Occams Razor says they didn’t have Steppe in the first place.

They did, because they came from Southern Europe where steppe ancestry was widespread since the Bronze Age.

Matt said...

@leron: The most parsimonious explanation probably depends on if an M269 subclade can be assigned.

If his M269 sits on, or close to the branch that expanded suddenly at around 5kya and where all populations with that marker who are sampled at roughly this time (earlier than ASH067) have a steppe ancestry background, then OG's explanation seems most parsionious. That is, it's more parsimonious for one sample, later in time, to have lost "steppe" autosomal from the original ancestor than for a large set of earlier samples to have gained it.

If it's some kind of M269 deep branch at 15kya, then it's probably some deeply diverged group that was never really on a steppe ancestry background (perhaps some component that later formed part of the steppe ancestry background, much later, at most).

Davidski said...

ASH067 does have steppe ancestry. I'll post the qpAdm models later today.

Davidski said...

@natsunoame

You sound crazy. One more post like that and you're banned.

PF said...

@Matt @Erik

Apologies for the late response -- holiday weekend here in the US -- but anyways Erik already pretty much voiced my thoughts on the matter better than I could have. Whatever contact with "Italians" there could have been just doesn't seem enough to explain a Mediterranean-wide common ancestry that's clear to me exists in all western Jews. Also, I think we should pause this part of the convo a bit until we finally get the upcoming Italian samples and see how they come out...

To be clear I don't think western Jews are exactly a simple 2-way mix of Levantines and Mycenaeans. Nor do I think the handful of Mycenaean samples we have are perfectly representative of the entire ancient Aegean world. Matt, I do of course agree when looking at a PCA that the Mediterranean Jewish cluster doesn't fall on a straight line between Levant_BA and Mycenaeans -- it looks slightly shifted towards the direction of CHG.

Take this fwiw but I'm having trouble finding a much better 3-way fit... that entire cluster just wants seemingly wants something extra CHG-specific beyond what exists in Mycenaeans.

[1] "distance%=2.3576"

Romaniote_Jew

GRC_Mycenaean,46.8
Levant_ISR_Ashkelon_LBA,37
ARM_LBA,16.2

I've always thought that Anatolia is relevant to this discussion, and my very loose hunch is that ultimately the Aegean source is from its eastern coast, that is, western Anatolia. Perhaps there may find a population not altogether too dissimilar from Minoans and Mycenaeans but more CHG-shifted. Certainly we see CHG input as far back as Tepecik, so I think it's very possible that there were populations in western Anatolia and eventually further west that had varying degrees of CHG on top of a similar ANF baseline. This is of course if we're looking for a single perfect population to mix into BA Levantines when in reality it could be and probably was more complicated.

PF said...

Quick PCA of East Med / Levant with the new samples from Ashkelon included. https://imgur.com/a/6RLefsM



Matt said...

@PF, yeah, the main shift of real Jewish populations vs Mycenaean+Ashkelon_LBA/IA2/Lebanon_Rome would be in a greater CHG+Steppe direction, so there is the possibility of using West Asian greater Caucasus region populations to pick up the extra needed. ARM_LBA is somewhat more steppe+CHG shifted (a bit more like a present day N Caucasus pop), so that would work relatively well.

I'm not totally sure if this is sound in the historical sense though, or not.

I still think there are some slight offsets in relatedness that favour post-Roman and present day Italian/Greek+Ashkelon_IA/Lebanon_Rome (basically there is a slight but indicative offset of real WHG in all populations which can't be explained without western mediterranean ancestry - you can't "make" any of them with contributions from the eastern mediterranean region alone). But ARM_LBA+Mycenaean+Ashkelon_LBA/IA2/Lebanon_Rome gets a lot closer than two way models with Mycenaean.

(In terms of two way models, funnily enough a two-way Palestinian+Mycenaean model would also come close, as two models go, but historically that doesn't make sense because the Palestinians only have the right position because of all the extra Ottoman era geneflow!).

Davidski: Apologies if this hits on your prohibition about talking on origins of European Jewish populations, I felt this was OK in spirit to reply and conclude(?) the thread of discussion in earlier comments.

«Oldest ‹Older   1 – 200 of 202   Newer› Newest»