search this blog

Monday, March 18, 2019

Open thread: What are the linguistic implications of Olalde et al. 2019?


I was going to write a huge post on the linguistic implications of the latest batch of ancient DNA from Iberia courtesy of Olalde et al. 2019, and then I thought better of it. Admittedly, I don't know enough about the languages of prehistoric Iberia to say anything really useful on the topic. So instead here's an open thread to bounce around a few ideas in the comments.

Just briefly, this is what Olalde et al. say in the abstract of their paper about the relationship between ancestry from the Pontic-Caspian steppe and languages in Iron Age Iberia:

We reveal sporadic contacts between Iberia and North Africa by ~2500 BCE and, by ~2000 BCE, the replacement of 40% of Iberia’s ancestry and nearly 100% of its Y-chromosomes by people with Steppe ancestry. We show that, in the Iron Age, Steppe ancestry had spread not only into Indo-European–speaking regions but also into non Indo-European–speaking ones, and we reveal that present-day Basques are best described as a typical Iron Age population without the admixture events that later affected the rest of Iberia.

However, in the paper it's revealed that "Indo-European regions" actually refers to a Celtic-speaking part of northern Iberia. And it's quite possible that Celts moved into this area from outside of Iberia only during the Iron Age. In other words, the speakers of Indo-European languages here may not have been the descendants of any of the people with steppe ancestry who came to Iberia by ~2000 BCE.

So I'm probably not alone in thinking that the question of the linguistic affinities of these early migrants with steppe ancestry to Iberia (mostly associated with the Bell Beaker culture or BBC) remains open, especially since they evidently had such a profound genetic impact on the later non Indo-European-speaking populations of eastern and southern Iberia. Could they have been the speakers of unattested Indo-European languages, as well as Proto-Iberian and Proto-Basque? If not, why not?

Below is a Principal Component Analysis (PCA) of West Eurasian genetic variation. I highlighted some of the ancient samples from Olalde et al., as well as Basques and other present-day Iberians. The Basques form a tight cluster with most of the Copper, Bronze and Iron Age Iberians, and, unlike the other present-day Iberians, they basically look like an Iberian population from the metal ages. The relevant datasheet is available here.


This is nothing new and very much in line with the results in Olalde et al., but I wanted to emphasize the point that Basques were not just a group that experienced an extreme founder effect in R1b-P312, which is a Beaker-specific Y-chromosome lineage. Rather, they're still very similar to Iberian Beakers in terms of overall genetic structure. So where did they get their language?

See also...

Celtic probably not from the west

Late PIE ground zero now obvious; location of PIE homeland still uncertain, but...

310 comments:

«Oldest   ‹Older   201 – 310 of 310
Aram said...

Zardos

So You think there was a mass killing of males in Iberia but no mass rape? Why so.
Brw what You think from what side the Semitic language came? I mean what haplogogroup?
And do You think that there was a mass killing of haplogroup E in Levant?

Can You show any msss grave in Iberia or in Levant? Any archaeological paper.

Gaska said...

@Them meee- "Is the steppe mtDNA in Iberia mostly due to later Celtic migrations?"

There is no steppe mitochondrial DNA in Iberia at any historical stage, and in BB culture-Central Europe only a dozen mitochondrial haplogroups

@Davidski said- "The prediction made by the detractors of the steppe hypothesis was that R1b-L51 would be found in indigenous Mesolithic, Neolithic and/or Copper Age Iberian samples without any steppe ancestry. That was obviously a bad prediction that has no chance of coming true.

I do not know what other detractors of the Kurgan's theory think, but I think that the theory of Gimbutas as you (kurganists) have interpreted it, has been totally debunked. You say that L51 will appear in Neolithic or Chalcolithic Steppe samples. Okay let's see it, for the moment that's just a fable.

1-I have always said that P312 is western (Franco-Cantabrian region), and that L51 is western (Germany, Alsace, Alps), and I still think the same. Try to prove otherwise and tell me which steppe culture has been found R1b-L51/P311/P312/Df27/L21/U106/L21. Obviously in Iberia more than 100 men of the Neolithic and Chalcolithic have been analyzed and there is no L51, we will see what happens in Italy and France.

2- I have always said that linking a uniparental marker to a certain culture or language does not make any sense because no European culture since the Neolithic is homogeneous enough in genetic terms. Try to prove otherwise.

3. I have always said that the Bb culture originated in Iberia because that's what the C14 data published in 2.104 (Cardoso) and 2.018 (Henriques) say. Prove otherwise

4-I have always said that the non-Indo-European languages ​​of Iberia were proof of the failure of the IE expansion during the chalcolithic in Western Europe, because in the Iron Age more than half of the Iberian peninsula and the south of France spoke Non-IE languages .Time has given me reason, simply the BB culture in Iberia and France DID NOT speak IE

5- For supporters of steppe theory this is a much more serious issue than it may seem, because if it is shown that P312 is Western, not only Iberia and the south of France would not speak IE, but also all those regions where the BB culture had a great incidence, that is, the British islands. Other regions, such as Italy, never spoke IE languages ​​in prehistoric times.

6- This also leaves in bad place your theory that P312 is hidden in the CWC, because that culture is linked to the steppes and IE. If that were the case, P312 would also speak IE, and Iberia has shown that this is not the case. Why? - Because of the obvious genetic continuity of Iberia between the chalcolithic (2500 BC) and the Iron Age. There was no reason to change the language, neither political, nor genetic, not social, nor military nor of any kind. Simply BB culture and P312 did not speak IE

7-In addition to Olalde's studies, in Spain there are more geneticists who work analyzing deposits. Recently, one has been published that together with R1b has discovered G2a in the Bronze culture of the Guadalquivir (related to the Argar culture). Then simply is not true that supposed extermination of which you like to talk so much, and the evident proof is the existence of I2a (5%) and G2 (3%) in the current Spanish population.

Kristiina said...

Correction:

My true opinion, which can of course change at any moment, is that R1b-M269, as well as its major branches R1b-Z2103 and R1b-L51, are from the area between the Balkans and North Caucasus.

Folker said...

@Zardos
I am no linguist, but there are several examples of language split between men and women. Some are consequence of war, with one group eradicating men of the second. Conquering men still use their group original language, and women the conquered group language.
I was thinking of an example in India, but I found something similar with Garifunas in South America.
It is not the same as Ubang or (to a lesser extent) Japanese since the differences between men and women are not related to the result of war.

zardos said...

Such low local survival rates, since some might be even later migrants and not from pre steppe BB, is what to expect from refuges, pockets, survival of slaves and specialists, like smiths, here and there.
Its exactly what to expect after such a large scale conquest with mass killings. If there would have been peaceful interactions, numbers would be much different.
So they hunted them down, because even if the newcomers would have just displaced them from the centers, you would expect concentrations of survivors. But they disappeareds after the conquest and never came back.

Aram said...

If R1b has a 1.5 male offspring in each generation. While I2 man has only 1 male per generation then after 10 generation or 300 years You will have 57 times more R1b than I2.

It is very easy to check. Just do this math.
1.5^10 =~ 57
While 1 ^10 is still equal to 1.

Davidski said...

@Aram

Reality usually isn't that linear.

The situation in Bronze and Iron Age Iberia just seems really weird, because in other parts of Europe various Y-haplogroups bounce back very quickly after they get washed over by the steppe expansions.

zardos said...

I know demographic models which would allow a breeding effect without mass killings explaining the pattern, but those models are completely detached from reality.
Thats not how it works. From the start of the intrusion to the final shift you need so many improbable variables that its like winning in the lottery.
And even then,for many places,time doesnt suffice for this highly unlikely scenario. And this lottery was wonder by the steppe Beakers throughout Western Europe over and over again? Because the calculation must be repeated for every place and settlements which show post BB continuity.
Get real, they used theweapons they got and regularlykilled each other in duels and feuds.
But they must have kept the Iberian locals alive and bred them out peacefully over generations in every place they conquered.
Are you serious?

Gaska said...


@Andrzejewski said..."Kristiina Do you think Basque is a non-IE HG (WHG) language or a non-IE EEF language?" And said "Who knows? Maybe the CHG in the Steppe spoke Proto-Basque and the EHG Steppe spoke Proto-IE?"

This type of questions will pass to the posterity of the genetic debates.

@Zardos-

You have a too passionate, bellicose and apocalyptic vision of what happened with BB culture in Western Europe. There are many arguments to convince you that the story was not as violent as you imagine, it would be a long debate and we do not have to bore others, just tell you one thing, the reproductive success of P312 in Iberia and other regions was not as explosive as you believe, but gradual, and although it is true that in the Bronze Age the Iberian population was mostly R1b, the fact is that I2a and G2a survived the hypothetical massacre. It is true that other groups became extinct like C1a2-H2, but these were always very minor (C1a2-2 samples in the mesolithic) and H2 (5 samples in the Neolithic). However there are more than 70 mesolithic/neolithic/chalcolithic Iberian I/I2 samples, more than 30 G2 samples and these haplogroups are still with us. In the same way there is British Bell Beakers I2a, and I suppose that I2 still exist in the isles no?

@Ryan sayd.."So our options are: We are missing data from such a location (maybe somewhere a non-IE culture took over IE lands?)

"Or R1b-L51 is from the steppe but the steppe was multilingual (I think this would be very hard to test for one way or the other"

The steppe was not multilingual, not even the neolithic or chalcolithic European cultures (central and western) was multilingual until the appearance of the CWC and R1a. IF P312 had been in the CWC, would speak IE, and both Iberia and France would have spoken Indo-European languages, and this is not the case.

"Or R1b-L51 isn't from the steppe and the steppe ancestry in Beakers is maternal.Do we have good enough data to check for sex bias with the X-chromosome yet? Like Mathieson did? I would use CWC as a donor population and see if the X chromosome has more or less CWC ancestry than the aDNA as a whole if you can"

Try to do it, and you can verify that a large part of steppe ancestry comes from women. Some mitochondrial haplogroups of the CWC in Germany, Poland and Baltic Countries are identical to the steppes and these later passed to the BB culture. I do not know if that will be enough to justify the 50% approximate steppe signal on the BBC but it is certainly a reasonable explanation. On the other hand, the mitochondrial haplogroups of the different BB regions in Europe are practically identical, the only difference is that steppe markers appear in Central Europe, and not in western Europe.That is why there is a difference regarding the percentage of steppe ancestry between some BBs and others.

Katharós said...

If an invading group has some sort of awareness for an administrative structure and an administrative apparatus is already in place in the conquered area. I think that the ones on top can easily force the administration to speak their language. But their also has to be some common ground within the belief system of different groups that will ease and fasten the adaption of a new language through religion, in the sense of a mediator. Interesting examples are the inscriptions about Jesus on the Dome of the Rock , Jerusalem. For one they are directly talking to a largely Christian population who are naturally familiar with Jesus, but at the same time they serve as an Islamic mediator , in the sense that they reflect Islamic views on Jesus. In addition early Christian writers did not view Islam as an independent religion, but as “a sect” heresy of Christianity. What actually underlines that early Christians viewed Islam as a part of their scriptural tradition. In addition Muslims adopted the administrative structure of the Greeks.

Gaska said...

@Kristiina.."My true opinion, which can of course change at any moment, is that R1b-M269, as well as its major branches R1b-Z2103 and R1b-L51, are from the area between the Balkans and North Caucasus"

You show that you are a prudent woman.

Regarding the origin of Euzkera, I imagine that you will also have some theory. what do you think? because the Neolithic origin seems evident by the root of some words, but I have heard so many theories that every time I have more doubts. The real problem that linguists have in Spain is that despite its evident similarity with the Iberian, modern Basque does not serve to translate the ancient texts written in Iberian. Although there are words that are very similar, for example the old name of Granada is Iliberis-Illiberri, and in Basque new city is Iriberri

zardos said...

Show me a context in which steppe BB being established and there are more than 40 percent local lineages or the like. Where is this gradual change? Where is the integration of local males in Britain or Iberia by steppe BB?
Exceptions, scattered survivors and small pockets, thats what you get after an onslaught.
When things calmed down, the survivors came out of their hideouts and some might have been spared by BB steppe because they were useful.
But again, you dont get that pattern without mass killings.
Show me a proof for the gradual change. But no steppe Beakers avantgarde, single individuals, but integrated local males after one of the main waves of conquest went over the country.
Such low numbers of survivors just prove the point and are no contradiction

Ric Hern said...

The Iberian Peninsula was not a peaceful place before Steppe Related Males arrived. Apparently many fortified Settlements. But overall not as numerous as populations North of the Pyrenees Mountains. A lot of local hostilities could have left Iberia with a small Male Population not big enough to support a majority of females. Maybe R1b Males were invited for dinner.....

Kristiina said...

@ Diego

Do you think that finding roots in Basque that are not shared with IE languages is enough to qualify those roots, and more importantly, the whole Basque language stemming directly from a language spoken by Spanish Cardial farmers?

zardos said...

If it would have been gradual, which still might mean violence and mass killings, just no genocidal targeting of the locals, there would be mixed contexts everywhere. But in all of Britain and Iberia is there a single one? So far it seems that no single steppe BB chief decided to spare larger numbers of local males. Not a single one in the conquered provinces.
And you insist it happened everywhere?
The results would be different if so.

Ric Hern said...

@ Zardos

Let us wait and see what France can deliver. I think a lot of this puzzle lies there...

PF said...

I have to agree with zardos on this one. It seems impossible to explain any way other than a sudden and swift warfare advantage to a specific clan of males. Figure B from the paper is a nice visualization. The decrease in local Y-DNA is immediate around 2500 BC. https://imgur.com/kKV3PbO

If Steppe males and their descendants absorbed women gradually, then Steppe ancestry would have been diluted over the centuries. Instead, it decreased in one fell swoop and then more or less remained constant. Yes, an even small compounding fertility advantage can explain the Y-chromosome imbalance, but it can't explain why the overall Steppe ancestry didn't eventually approach zero.

You'd need absurdly specific and unrealistic scenarios to make the math work. Something like a constant stream of immigrating Steppe males who choose to only mate with unadmixed locals, or absorption of local females in one, maximum two, generations, followed by 500 years of endogamy, etc. These make no sense.

The language question is another challenging paradox... interestingly, the inverse of Anatolia, where you have attested IE languages but little-to-no Steppe ancestry...

O/T, but @Davidski, it would be amazing if you could get the Anatolian_HG genomes published yesterday into G25!

Ric Hern said...

@ PF

If that is the case then the only thing I can think of is a Military Campaign sweeping through, killing as many males as possible and violating as many women as possible on their way to find El Dorado but not settling down and raising their children....

zardos said...

To kill enemy males of all ages, but taking the local females and raising a new generation with them was a completely normal thing throughout human history.
It happened a lot and we all are descendents of such unions regardless of whether it was rape, slavery, concubinat or a regular marriage under the rules and customs of the conquerors.
Human behavioural differences between the sexes can be, in part, even attributed to these fact.
Or why do you think males are more violent, risk takin, less socially adaptive and more xenophobic on average?
They have more to win and more to lose, while females have a better chance of "sitting it out" if they are ready to make a compromise with the environment they meet. Thats why they adapt and try to inluenceand change things slowly. The direct and violent way won't work.

So they accepted their fate and there are thousands of such cases proven for human history.
Usually more local males survived though,so steppe BB are definitely among the most extreme and biggest scale cases of all in human history.

Because the case is so clear its so absurd people argue against it.

Ric Hern said...

@ zardos

I think everybody understand this, just not how Non-Indo-European Languages survived this Steppe Related eradication.

Gaska said...

@Kristiina

First part of the question, Yes, I think it is enough along with other data to qualify a language such as Indo-European or not

Second part of the question, Obviously NO, I think it has 40% of Indo-European influence (Latin, French and Spanish), the distance between the Cardial culture and the current Basque is more than 7.000 years so it is impossible that it has not evolved. In fact, one of the great problems of Basque is that we only have Aquitanian inscriptions of the Iron Age as written references. On the other hand, the high percentage of Palaeolithic and Neolithic ancestry in Basque women, suggests that this obvious genetic continuity could help preserve a very old language. You know that in the Basque Country, women were the ones who transmitted the family heritage (in a certain way, this is still the case), maybe they kept the language too. Of course we will never have proof of it, I hope the linguists solve the problem sometime.

Answering a question with two questions means that you have no opinion about it?

zardos said...

No, a lot do not. Some scientists, journalists and commentators still make up stories of how it happened without mass killings, while this is impossible.
And it is clear by now that the incoming steppe Beakers were no IE speakers.
Now the question remains how that could be and that can be answered by tracing the lineage in question back in time.

Gaska said...

@Ric-"I think everybody understand this, just not how Non-Indo-European Languages survived this Steppe Related eradication"

If Zardos is right, there is no reason to think that R1b-P312 spoke an Indo-European language, they simply spoke Basque or Iberian and imposed it on the native population. It would be amazing to massacre the local population and then adopt their language.It has no sense.
Violence has become the worst enemy of Kurgan's theory

Grey said...

Ryan said...
"I'd add that the star-like phylogeny of R1b-L51 (and R1b-L23 to a lesser extent) suggests the expansion was very rapid and probably uh... "not nice" or not a two way exchange as you'd put it."

the rapidity always seemed significant

the only alternative explanation to violence i've ever come up with is the herders expanded into land the farmers couldn't use or weren't using for some reason

e.g.
farmers 10% of land, HGs still 90% of land, herders arrive and replace HGs
-> farmers 10%, herders 90%

or

farmers 100% land use then some kind of calamity leading to population collapse to 10%, herders arrive during the low point

Kristiina said...

@ Diego

Let’s make a small test with some basic Basque words.
bihotz (Navarrese), bigotz (Baztanese) = heart
ur = water

Let’s take the word «heart» first. The first part «bi» seems not be part of the root because «bi» is often found in different bodyparts. The remaining root is «hotz». The word for «mind» in Basque is gogo. Could «heart» be derived from the same root as «mind»? Two North-West Caucasian words for «heart» are a-gʷǝ́ (Abkhaz) and gʷǝ (Kabardian). In Old Armenian hogi means «soul», «ghost» and «person». Could the Basque words for «mind» and «heart» be derived from the same root as all these forms?

How about the word for water. Does the Basque word for «water» resemble any roots for water in other languages? First of all, we have to take into account that the word for «rain» in Basque is e=uli. Therefore, we have two possibilities : «ur» and «ul». The relevant Eurasian roots seem to be c̣q̇ali (Georgian), c̣q̇ar- (Meghrel), cʰil (Hunza), all meaning «water») and the Nakh verbs meaning «to rain» and «to weep»: =elχ- (Ingush), =ēlχ- (Chechen). On the IE side, we have the Greek form «húdōr» and the Armenian form «dʒur», both meaning water. If the Basque word were related to the IE words, it would be inside the IE development as «r» is not cosidered part of the root. The comparison with Nakh languages looks promising.

IMO, these two exaples with basic roots show that it is not so difficult to find a connection between Basque and the Caucasian languages.

zardos said...

Between Kurgan culture on the steppe and BB with steppe ancestry is a lot of space metaphorically and literally.
Because of the Kurgan -> CWC its still viable. But CWC (proper) is the main thing for the IE prehistory.
As for R1b in Western Europe in general, a lot seems possible by now. Except an Iberian origin of the BB lineages.

Grey said...

zardos said...
"For those which still reach for straws to avoid the explanation by the mass killings of the males in Britain and Iberia for sure, but most likely throughout steppe BB territory:
Matrilocality, status, bought brides etc. explain nothing if the local males completely disappear in such short periods of time."

Iberia seems to have had multiple groups arriving from different directions and they had a lot of shiny metal to fight over so i don't doubt it *ended* in violence - i'm still curious about whether it started that way.

it's an interesting puzzle.

for example, group A arrive in remote low population region looking for copper, marry into local matrilocal culture, develop a military advantage of some kind, expand violently.

iirc DF27 is the key clade in Iberia and (i don't know how accurate those Eupedia maps are) but last i looked (which is a long time ago) the epicenter was the pyrenees near a big neolithic copper mine.

(i may be misremembering lots of the above)

#

speaking of military advantages if it's not bronze swords how old are jinetes? and are they associated with a particular region?


Richard Rocca said...

Davidski said..."Take a look at Figure 2B here...

https://reich.hms.harvard.edu/sites/reich.hms.harvard.edu/files/inline-files/2019_Olalde_Science_IberiaTransect.pdf

The genetic impact of the Bell Beaker and related migrations from the east was very significant from the start, and caused the local Y-chromosome lineages to basically vanish in maybe less than 500 years.

Based on the available data, this process looks like a wavefront that eases somewhat with time, but continues at a significant rate into the Iron Age.

I can't see anything gradual there. It's all very sudden."


That's anything but "very sudden". See Reich's two graphics (Bell Beaker Britain versus Bell Beaker Iberia). As you can see, Bell Beaker in Britain was absolute and sudden (what you are describing). Literally within one or two generations the entire population was replaced, and obviously Indo-European languages as well. Bell Beaker in Iberia was anything but "very sudden", lasting 550 years which is roughly 22 generations. That is more than enough time for partial or total language replacement in the case of Iberia, although I think there is some founder effects at work there... perhaps DF27+ZZ12+ and DF27>>Z220 whereas in plaes like France you have almost 100% Z195 above Z220.

http://www.r1b.org/imgs/David_Reich_Lecture.png

zardos said...

@Grey: That is the modern distribution because of later conquerers and migrants. Basques were best in preserving their Beaker heritage. Its not by chance they kept the language as well.

@Richard: Do I have to remind you on the Roman conquest of Iberia? Its a harsh terrain and more difficult than France or Britain where you get foward easier once you breached the initial defense.
And the pre steppe Iberians seem to have been no easy target.

The question of rapidity must be answered for a specific region after the conquest. If they conquered Basque country and replaced all males there that is rapid, even if they needed another 500 years to come to Gibraltar or the like.
The Spanish reconquista needed centuries as well.

Ric Hern said...

@ Diego

At the end of the day L51s nearest relatives were found on the Steppe almost 3000 years before its descendants entered Iberia. We see U5a1b in Ukraine Eneolithic with a bucket load of R1bs. We see U5a1b and R1b in Rathlin in Ireland with Steppe related ancestry. We see J1c in Ukraine Eneolithic with R1b. We see U5b2a2 in Lithuania, Poland, Germany and Ireland at Rathlin.....So nobody is going to convince me that Steppe related R1b men actually originated in Spain or even Germany.

That said, it is also noteworthy that R1a and R1b was present in the area and time, postulated by most linguists, where and when Proto-Indo-European originated.

With this in mind I think there were less hostile interaction between Neolithic/Chalcolithic peoples in France across the Rhine and some kind of Cultural loopholes were found by R1b Males to integrate into this society and quickly attain high status and privileges associated with it without changing the adopted Cultures Language. E.g.(Like I mentioned Cattle for Brides elevates the status of a females family in some African Cultures.) By moving to stay with the females family the men could cash in and exploit on this elevated status....This did not need to be a slow process as long as Steppe related men had many cattle to use as bride price. And we know Steppe people had massive herds of cattle.

Anyway. I think something like this happened in France and R1b Males who adopted a Frence Neolithic/Chalcolithic Language spread into Iberia from there, maybe welcomed at some places due to a shortage of men and in other places opposed...etc.

Davidski said...

@Richard Rocca

That's anything but "very sudden". See Reich's two graphics (Bell Beaker Britain versus Bell Beaker Iberia). As you can see, Bell Beaker in Britain was absolute and sudden (what you are describing).

Both are very sudden and extreme compared to what generally happened, it's just that the Beaker invasion of the British Isles was even more so, probably due to the lower population densities and less complex societies there than in Iberia.

The Beaker movement into Iberia was something really strange, considering this was southern Europe. It's very difficult to explain why the local Y-chromosomes couldn't bounce back even slightly if this was some sort of political and cultural process.

Andrzejewski said...

@Aram "If Dene Yenisseian theory is true then Yenisseian languages must be an East Asian language not a language of Afontova Gora. Thus the idea that any Q was speaking Yeniseian languages is wrong.
Not surprisingly the most common haplogroup of early Turkish speakers Huns was the Q. The reason for this is simple. Because it was the Q who learned the herding the first in East Eurasian steppe."

You are wrong! Huns spoke Yeniseian. There was a phrase used by the Jie tribe of the Xiongnu Confederacy in Northern China, which only made sense translated from Yeniseian, not Turkic.

Native Americans are Ancient North Eurasians + East Eurasians, which means that they DO have a strong AG3/MA1 affinity.

To me it makes sense that Botai spoke Yeniseian, Okunevo spoke Yeniseian, Huns spoke Yeniseian, etc.

Richard Rocca said...

I guess you and I have a different opinion on what very sudden it then, because 550 years ago the location I'm in now (New York) was still American Indian territory and in a span of 550 years Rome went from a backwater town to the largest empire in the world and 550 years after that, it fragmented into almost nothing. Anyway, like I said, there also seems to be some kind of founder affects going on Iberia and they include only some subclades of DF27, which itself is only a subclade of ZZ11, which itself a subclade of Z40481, which itself is a subclade of P312.

zardos said...

I tell you what the difference to Northern Europe was. In the North the defeated could flee and scatter, even if for example CWC people wanted to exterminate all local males they wont get them all, unless there is a hunt going on for generations. The defeated could move and some groups survive until new Allianzen emerge.

The developed Iberian culture on the other hand needed fertile soils and had fortified centres.

The advance was slower and bloodier with high death toll for the conquerors.
But if they managed to break the centers, they got them all.
Even for those outside of the fortresses, for farmers its much more difficult to survive than for hunters or pastoralists once their defense was broken.
Those which were not killed initially might have just starved to death.
This also explains why the women preferred to live with the newcomers.
Even if they would have used no violence against the women, what I doubt, it was eithergoing with the killers of their kin or dying without food and protection.

Ric Hern said...

What does this mean ?

"replacement of 40% of Iberia’s ancestry"

Does this point to a 20% bigger Female to Male Population ? Or that those males who migrated into Iberia already carried 10% Iberian Related Ancestry?

Gaska said...

@Kristiina

I have read that one of the reasons for relating euzkera with the Caucasian languages ​​is the use of the ergative, but obviously I am not an expert. I don't know if the translation into English is correct.

Gizona etorri da (guisóna etórri da)-The man has come- El hombre ha venido
Gizona ikusi du (guisóna ikúsi du)- Have seen the man- Ha visto al hombre
Gizonak ikusi du (guisónak ikúsi du)- The man has seen- El hombre ha visto

The active subject receives a mark, called ergative or active mark, which is specifically a final -k.

Is that enough to establish a relationship with the Caucasian languages?

Gaska said...

@Richard said- That's anything but "very sudden". See Reich's two graphics (Bell Beaker Britain versus Bell Beaker Iberia). As you can see, Bell Beaker in Britain was absolute and sudden (what you are describing). Literally within one or two generations the entire population was replaced, and obviously Indo-European languages as well. Bell Beaker in Iberia was anything but "very sudden", lasting 550 years which is roughly 22 generations. That is more than enough time for partial or total language replacement in the case of Iberia, although I think there is some founder effects at work there... perhaps DF27+ZZ12+ and DF27>>Z220 whereas in plaes like France you have almost 100% Z195 above Z220.

We finally agree on something.

By the way Olalde has also located Z195 in the Bronze Age Iberia, I guess someone will be checking the SNP calls.

Gaska said...

@Ric

Nobody is trying to convince you of anything. I hope you do not try to convince me, because you will need much more convincing arguments than U5ab1. If I were you, I would look for L51 in the steppes, finding it would end the controversy.

Kristiina said...

@ Diego

No, we have to take account of all elements available to us.

However, the element you mentioned is important and indeed one significant piece in the puzzle.

A said...

Has anyone mentioned the fact that people called Iberians inhabited the Caucasus in Antiquity?

A said...

Caucasian Iberians: https://ipfs.io/ipfs/QmXoypizjW3WknFiJnKLwHCnL72vedxjQkDDP1mXWo6uco/wiki/Caucasian_Iberians.html

Synome said...

Anyone who argues against a language shift for Beakers in Iberia has to then locate that shift elsewhere in Europe, or defend a very early origin among steppe groups.

I don't see how those scenarios are more likely. It looks like Bell Beakers were quite dominant in France and Germany as well. They were at least as steppe admixed as Iberian beakers and had almost universal R1b.

Putting aside cultural and geographic arguments, based only on parsimony, I think it still makes sense to locate the language shifts in Iberia--especially since we have to account for multiple different non IE languages.

Here's another theory: non IE languages in Iberia were adopted after the Beaker expansion from the area of the Mediterranean. Think of the Basque-Sardinian hypothesis. But this scenario has its own problems. Where's the IE substrate in the paleo-Hispanic languages and Basque? I think it's at least worth considering, though.

Andrzejewski said...

The fact that a supposed language shift took place in or around Iberia but nowhere else is problematic.

Also, I find it hard to understand or believe how come anywhere else in Europe the Yamnaya mixed with the EEF population to reduce the pure Kurgan aDNA to a mere 50% (or 20%-30% in Southern Europe), but in England or Spain the BB exterminates all the male lines (or even the entire population as is in England).

Hard to believe how come the Anglo-Saxon didn't wipe out all the Welsh Britons whereas the BB in BA England effectively did just that (towards the Stonehenge builders).

Grey said...

JuanRivera said...
"Horsemen are likely to be just some decades younger than domestic horses."

yes i meant jinetes in the sense of the specific form that was a big deal in the middle ages: swarms of javelin armed light cavalry.

however a quick google seems to suggest they were a reaction to Berber cavalry.

on the other hand transhumance can lead to superior light cavalry

https://en.wikipedia.org/wiki/Transhumance#Spain

Mem said...

This is not true.Most linguist translated hunnic phrase with Turkic explaination.Only two guys explained it with Yeniseian.Yeniseian theory is most weakest in at all, also Xiugnu remains were not genetically different from Gokturk and Early Oghuz remains.

zardos said...

The newcomers might have been disadvantaged as fathers genetically even if Iberia, and the Basque country in particular, was predominantely Rh- at their arrival.
Did anyone check the Rh- SNPs for a comparison? If they can be retrieved, it would be interesting to see which group had a higher frequency and whether that really was a thing at that time or a later founder effect or some other recent development.

Matt said...

@Ryan: "Do we have good enough data to check for sex bias with the X-chromosome yet? Like Mathieson did? I would use CWC as a donor population and see if the X chromosome has more or less CWC ancestry than the aDNA as a whole if you can."

Ryan, one other thing you may be interested in (Davidski you also may be interested), and I forgot to mention this earlier (and bugged me all day), is don't count on the X to show a clear signal, because it can be confounded by complex sex bias processes.

For example, let's say we have a mixture between 100% Corded Ware males, 100% Globular Amphora females, to generate Pop B. Extreme but imagine it. Autosome is 50:50 Corded Ware:GAC, X is 33:66 Corded Ware:GAC, and that's the limit of skew achievable in one generation. Y dna will be 100% Corded Ware and mtdna 100% GAC.

Now let's say that Pop B mixes *back* with Corded Ware, with an equally strong male bias. 100% Pop B males, 100% Corded Ware females. Autosome is then 75:25 Corded Ware:GAC, while X is 77:23 Corded Ware:GAC (or even level with autosome). But y-dna remains 100% Corded Ware, and mtdna is 100% Corded Ware. (If you iterate this process, it's imaginable then to end up with populations with 100% CW mtdna, 100% CW y-dna, and quite low and equal amounts of X and auto CW ancestry).

This is a pretty extreme scenario, but can give some idea of the kind of reversals that are possible on the X. This kind of extreme scenario is not so likely, but in more moderate form it can still confound comparison of X:Autosome ratio, which often has some measurement error that can end up large compared to small differences in X:Autosome proportion.

We probably have to bear in mind that warfare would probably be common between the hybrid cultures that formed and their parents, and was probably conducted by kin linked male groups. The sort of dynamics that froze sex bias in place in for example Spanish America - consciousness of race and language and religion, and hierarchical society built around them - may not hold true in the late Copper Age and early Bronze Age.

This sort of thing is possibly also significant in Iberia, where post-Bronze Age migration from North African, Greece and Italy probably has some degree of male sex bias as well, from males with less steppe ancestry than the Iberian BA and different sex bias admixture patterns.

Heraus said...

The problem to be solved is simple : are Iron Age Iberians from zones known to be IE-speaking in Roman times characterized by a greater influx of "Central European" admixture ?

If so, then it means that Celtiberian areas were subject to one more migration from Central Europe which enabled a linguistic shift, from a pre-IE language (let's name it "Basco-Iberian") to an IE language (a modality of Celtic ? the Cantabrian language seems to be enigmatic though indisputably IE : a pidgin ? Cantabria and Asturias must be tested).

The issue is that samples from La Hoya, presumed to be Celtiberian, are quite modern Basque-like : we might need more samples. Above all we realize that we are forced into thinking about what Celtiberia was about : was it really massively IE-speaking ? It might just have been the elite, the one who wrote and let us epigraphic testimonies.

I am quite tempted to believe that the situation was identical in southern France : Aquitania which would lately be known as Novempopulania was undoubtedly pre-IE-speaking (modern-day Gascony i.e "Vasconia") but what about areas close to it ? Pre-IE remnants in Occitan placenames in the whole of southern France and to put it clearly the "facies" of autochtonous southern French people make me think that celtization was quite superficial south of a La Rochelle-Montélimar line.

Davidski said...

@Heraus

The Celtiberians are Basque-like, but they're actually shifted east compared to Basques and cluster near Hallstatt samples from Czechia.

Davidski said...

@Synome

Anyone who argues against a language shift for Beakers in Iberia has to then locate that shift elsewhere in Europe, or defend a very early origin among steppe groups.

I don't see how those scenarios are more likely.


That's actually a secondary issue. The main issue is that the direct descendants of Iberian Beakers, with practically 100% frequencies of Beaker paternal lineages, spoke Iberian and still speak Basque.

You first have to prove somehow that the predecessors of these languages weren't their native languages before moving on to models of how and where they acquired them.

zardos said...

The problem with steppe BB and Iberian R1b being non-IE is no big one. Britain and France, no biggy, but it gets big if its about other, related R1b lineages. Because later Celtic and Italic tribes seem to have been R1b dominated as well. So the most plausible explanation is that there was an ethnolinguistic split of R1b lineages at some point, with one group staying or turning to non-IE, the other becoming or staying IE.
Any suggestions on that issue? Or is the IEisation by Celts and of Italia even thinkable without R1b lineages being the main, original carriers?

Heraus said...

@Davidski : You are right. So I presume the conclusions are quite obvious : a subsequent migration enabled the linguistic shift from pre-IE Vasconic to IE Celtic everywhere in Iberia where IE-languages are known to have been spoken.

Whatever happened to those people BTW ? Let's remember La Hoya is in present-day Alava, modern-day "Basque Country" but this part of the Ebro valley never was mainly Basque-speaking in recent times.

The latest study about modern Iberian genetics clearly identifies a peculiar genetic "universe" around what was La Hoya, in modern-day La Rioja and Rioja alavesa (white circles on the map). Presumably the local melting pot survived in one way or another, different enough from "pure" Basque people in their mountains some kilometres in the North.

https://www.nature.com/articles/s41467-018-08272-w

Above all, for those interested in Iberia, we just need new samples. What are the precise genetic borders of the Basque world for instance ? Are Basque placenames in La Rioja (around Ezcaray) and old Castille the product of medieval migrations or a "miraculous" long-term isolation ? What about modern-day Catalan Pyrenees where half of the villages have undisputably Basque names ? I see that the 6th century sample from Catalonia shows individuals who are still Basque-like (while others areshifted towards the East due to East Mediterranean migrations : were they coming from the neighbouring mountains ?

BTW what about the well-known African contribution in West Iberia ? It now seems rather obvious it is fairly recent since Middle-Ages Iberians show high North African admixture and ancient Iberians lack it. Admixed people in eastern Iberia (the Valencian kingdom and Aragon) were expelled to North Africa and such areas were subject to massive peopling from the North (and southern France as well) during the Reconquista and during the XVIth century.



Tuga in the City said...

Napoleon couldn’t succeed in invading Iberia, Hitler didn’t even dare to try it.. but some tribes thousands of years ago succeeded and wiped out every male... makes sense..

Heraus said...

@Tuga : Napoleon did conquer Iberia and actually had his brother become king of Spain. The Pyrenees are all but a border, actually, similar genetic people live on both sides of the mountains (from Basque people in the west to Catalan people in the east).

That Iberia is a mountainous landscape is obvious, it explains why local genetic micro-universes were possible before a strong phase of homogenization during the Reconquista (and then the XIXth century with many internal migrations within Spain).

As for R1b people, let's not forget that some Y-haplogroups curiously possess a reproductive advantage. Biology is complex. Male lineage replacement might not have been 100% cultural.

epoch said...

@Davidski

"You first have to prove somehow that the predecessors of these languages weren't their native languages before moving on to models of how and where they acquired them."

The Beaker paternal lineages can't have originated in Iberia. So the way to go is check if we can find any sign of a Basque or Iberian languages in hydronyms and toponyms. The Vennemann obviously proposed a Vasconian substrate but that has been rejected by many linguists (IIRC his theory required Basque to hardly have changed over the years). An earlier proposal by Hans Krahe gave IE etymologies for a lot of hydronyms.

https://en.wikipedia.org/wiki/Old_European_hydronymy

If R1b didn't originate in or near Iberia, and Celtic expanded in the Iron Age, Hallstatt is generally considered as origin of Celtic La Tene and these were R1b. Maybe not the same subclade but at least we know that R1b was responsible for Celtic, came from an area with absence of Vasconic toponyms and was foreign to Iberia.

Gaska said...


Davidski is right and also everything seems to indicate that Df27 is practically Iberian, we have Df27 in a Bell Beaker burial in La Rioja, another in a site of the Cogotas culture (cueva de los Lagos, la Rioja), that is the continuation of BB culture in Iberia, and in the Iron Age (Edetania, Iberian Culture-Mediterranean coast). On the other hand we have Df27-Z195 in the Valencian Bronze Culture, with what the genetic continuity is more than evident. On the other hand, although we do not have samples from the Basque territory south of the Pyrenees, Basques are exactly the same as the Iberians of the Iron Age and they do not speak IE either. In addition, the Tartessians-Turdetans are also R1b. This simplifies things a lot to Spanish historians, and means among other things that the Basques are Iberians that were isolated in the mountains of the Pyrenees due to historical circumstances, where they developed some subclades typical of Df27.

This makes it very difficult for the Lusitanians to be Indo-European despite their relationship with other Celtic peoples of the peninsula, because the Bronze Age in Portugal is also absolutely R1b-P312, and leaves the Celtiberians who, as Davidski says, are similar to the Czechs of Hallstatt as the only Iberian candidates to speak IE. These Celtiberians occupied a relatively small territory of the North Plateau and also the only sample we have is I2a. However, I suppose that when more deposits are analyzed they will also be R1b, because the main focus of R1b-P312 in the peninsula is in the provinces of Burgos and La Rioja, bordering with the current Basque territory and Navarre.

The chances of P312, grandfather of Df27 speaking IE are Zero



Heraus said...

@Diego Arroyo de Lagasca Encinas :

Let's add that Strabo said Aquitanians did physically look like Iberians. He did not write that they looked like Vascones (in modern-day Navarre and Upper Aragon).

Actually, the ancestors of modern-day Basque-speaking people are never really differentiated from other Iberians by Greek and Latin geographers. And now it is pretty logical as we have just realized that most Spaniards did diverge from Basque people mainly in historical times (which is something few people expected to be honest) if we except some IE dominated lands which had suffered from Celtic invasions a bit before Roman conquest.

One day we will wonder why a very sane theory such as basco-iberism was abandoned for decades before being resurrected recently (in the linguistic field as well with Basque being used to decypher Iberian first names).

Gaska said...

Now what will be interesting is to check with respect to E1b, what percentage of the current proportion of 7% in Spain comes from the Roman conquest, from the Visigoths and from the Moors. The rest of the story is easy to explain because the Reconquista caused R1b to recover lost territory, in fact 70% of men in Granada are currently r1b, a higher percentage than in Galicia.

I do not think Napoleon was very satisfied with his experience in Spain

Drago said...

I bet Etruscans come back full of R1b-L51

Hehhee

Heraus said...

@Diego Arroyo de Lagasca Encinas :

IMO this is a major aspect of Olalde's study : recent historical migrations have had a massive impact in Iberia. Too many peope denied the Reconquista was also a major demographic phenomenon : it was and one had just to have a look at what happened later on in the Canary Islands then the Americas. It was all about male Iberians looking for new opportunities.

Lately, we have learnt that the Balearic Islands indeed lost its autochtonous population, that in Al Andalus a half-North African half-Iberian population was presumably dominant (are their descendants in Maghreb ?), that Roman migrations had a deep impact on the emerging of Iberian nationalities, that North African migrations reached areas as far north as pre-Pyrenean Catalonia between the 6th and the 12th century (and I suppose the whole of West Iberia), ...

This is really fantastic. Now only minor aspects still need some explainations, such as knowing more about the genetic landscape of the border between Basque-speaking lands and what would become Castille, where Castillian was born. I'm also quite curious about knowing who ancient Cantabrians were as their language seems to be IE yet they allied with Aquitanians who were Vasconic-speaking during the Aquitanian war.

zardos said...

"Tuga in the City said...

Napoleon couldn’t succeed in invading Iberia, Hitler didn’t even dare to try it.. but some tribes thousands of years ago succeeded and wiped out every male... makes sense.. "

That's no argument at all, because Napoleon and Hitler failed miserably in Russia, yet you can make a long list of steppe people conquering it.

Also, you should never forget that modern guerilla wars were almost never total in their character. Which means the conquerors let the villagers survive, let them live on and tried to get the active combattants primarily. Attacks on civilians took place, villages here and there were burned to the ground, but not on a large scale and in a genocidal manner.
That's different for ancient times, where, if a people did more than one uprising, the result was quite often a true war of annihilation, which means no field, no animal, no human they met was spared. All were killed or taken away, so the basis of the guerilla war was gone after some time. Especially farmers can't survive that kind of warfare for too long, if the attackers have a solid base and supply.

Iberia was conquered more than once, at least large parts of it. So no argument at all. Its not like Napoleon couldn't have conquered Iberia, its just that he had a lot of other places to care for too. And as gruesome as it was, it was no war of annihilation.

Davidski said...

@Zardos & epoch

Celtic developed in the post-Beaker societies of Central Europe where several Y-chromosome lineages, including R1b-L51, were already present at high frequencies for a very long time.

So I'm not seeing any strong parallels with what happened in Iberia, where you'd need R1b-L51 to quickly move across some plausibly hard linguistic borders to achieve frequencies of ~100% in different non-Indo-European populations at the opposite (north and south) ends of Iberia.

Davidski said...

@Dragos

I bet Etruscans come back full of R1b-L51.

They might, but that won't prove anything either way, because if R1b-L51 is found at extreme frequencies in linguistically very distant populations across Western Europe, then it might just mean that Indo-European-speaking Beaker-derived groups were opportunists who switched languages whenever necessary to fit in.

truth said...

Heraus said:

", that in Al Andalus a half-North African half-Iberian population was presumably dominant (are their descendants in Maghreb ?), "

Based on ? This is not what the Olade paper shows.

Heraus said...

@truth :
Gene flow from North Africa continued into the Muslim period, as is clear from Muslim burials with elevated North African and sub-Saharan African ancestry (Fig. 2D, fig. S4, and table S22) and from uniparental markers typical of North Africa not present among pre-Islamic individuals (Fig. 2D and fig. S11). Present-day populations from southern Iberia harbor less North African ancestry (25) than the ancient Muslim burials, plausibly reflecting expulsion of moriscos (former Muslims converted to Christianity) and repopulation from the north, as supported by historical sources and genetic analysis of present-day groups (25).

truth said...

OKay, but the average ancestry of those samples is about 75% Iberian and 25% North-African (excluding outliers with SSA ancestry or Levantine), but Not 50%/50% as you said. Yet, we don't know if that's representative of the whole Al-Andalus, or just the SE Iberia area, still, this reflects the Reconquista had a huge impact, as modern populations from the same area are nowhere near those levels today (specially important in the case of Valencians, which at most today have 3-4% NA), also the expulsion of the MOriscos in the 17th century had an effect

Simon_W said...

@ Davidski

"But which part of Iberia was definitely IE before the Celts came? I'm aware of Lusitanian as a possibility in SW Iberia, but that's not a clear case and still disputed."

Lusitanian was actually rather located in Western Iberia:
https://justpaste.it/7j487

@ Matt

"there's no attestation of any Basque like languages outside Iberia"

I have no doubts that Aquitanian was Basque-like, and it was chiefly located north of the Pyrenees:
https://en.wikipedia.org/wiki/Aquitanian_language

Ryan said...

@Synome - "Anyone who argues against a language shift for Beakers in Iberia has to then locate that shift elsewhere in Europe, or defend a very early origin among steppe groups."

It's easier to have a weird founder effect before an expansion rather than after though.

Samuel Andrews said...

Important to remember, the Muslim samples are not representative of all people who lived under Muslim rule in Spain. They came from one location: Granda. Also, pre-Muslim, Granda samples surprisingly have the same level of North African ancestry.

Samuel Andrews said...

...Therefore, Granda Muslim samples cannot be used to measure population replacement caused by the Christian Reconquest.

Samuel Andrews said...

Olalde 2019, refutes idea of Visigoth ancestry in Spain. Anyways, it doesn't locate where the "East Med" (Sicilian, Crete-like) admixture comes from. It isn't from the Greek colonies because the Greek samples cluster with Myceneans. Some of the Medieval samples have it. The Granda samples, including pre-Muslim, have it.

Andrzejewski said...

@Samuel "Olalde 2019, refutes idea of Visigoth ancestry in Spain."

What happened to them??? Got swamped with the Moors during the Reconquista and kicked to North Africa? Did all Visigoths migrate to America with Cortez and other conquistadores? I'm poking fun here because I find the idea ridiculous.

BTW, speaking of Moors, there were only 20,000 invaders out of a total Iberian population of close to 3M. So less than <1% of the total. Most of the Mozarabs and the Muladis were either Iberian converts from Christianity to Islam or Christians living under Muslim rule.

Speaking of "Eastern Med" admixture (which was up to 25% during Roman times), did they too get confused with Arabs and get deported to Morocco? It's all just hard for me to believe that of all the post-BB invasions: Romans, Germanic, Arab, Berber, etc, almost NONE left almost zero trace in modern populations. Let alone the Visigoths, who largely replaced Roman soldiers and administrators as the ruling elite, and who mostly led the Reconquista.

Andrzejewski said...

@All but especially @Samuel Andrews

Speaking of Germanic and language switch, don't you find it odd (similar to BB switching to Basque), that 40% of France north of the Loire is inhabited by Germanic Franks, Burgunds, Alemany etc, but that outside Alsace-Lorraine they switched to Vulgar Latin -> French?

It would have made sense that the bulk of population settling in Northern France would've switched the local's language to some dialect of German.

Samuel Andrews said...

Nature 2019.
Late Pleistocene human genome suggests a local origin for the first farmers of central Anatolia
https://www.nature.com/articles/s41467-019-09209-7

Heraus said...

@Simon_W :
Indeed Aquitanian is Basque-like, and quite probably the ancient form of modern-day Basque as spoken on French territory. It is also a well-known fact that a Romance language was born on those former Bascoid-speaking lands named Gascon which happens to share phonetic features with Castillian and Basque (a repulsion for f or initial r).

From an epigraphic point of view, Aquitanian appears to be the very same language as the one found in written ancient remnants in modern-day Spanish Navarre.

Let's add that Sardinian non-Romance placenames are also undoubtedly Basque-like : one just has to have a look at a map of Barbagia, the inland mountainous area in Sardinia. But I suppose it's the other way round : Basque is Sardinian-like.


@Andrzejewski :

Linguistic borders evolve. For instance, Dutch used to be spoken in a greater area of northern France in what should be called the "French Netherlands".

https://fr.wikipedia.org/wiki/Flamand_de_France#/media/File:Noord-Frankryk_Nederlands.jpg

As for the French language, it is so influenced by Germanic stressing and so different from what a standard "Gallic" Romance language should be (see Occitan for that, halfway between Catalan and Northern Italian dialects) that it is enough to understand Germanic influence was heavy. Just that Northern Gaul was already too much populated. Or that Latin culture was too strong and desirable.

Yet I am quite inclined to believe that Germanic genetic impulse was rather massive north of the Loire river. When one is French one knows that people in the South and people in the North are not really the same.

epoch said...

@Davidski

"
Celtic developed in the post-Beaker societies of Central Europe where several Y-chromosome lineages, including R1b-L51, were already present at high frequencies for a very long time.
"

But isn't that exactly the reason why we can assume that the original language of R1b "tribes" is an IE dialect? The only other possibility is that the "tribes" that later became Central European Celts picked their IE dialect up moving through IE speaking area's. But the almost complete absence of non-IE toponyms and hydronyms in the areas of NW and NE France and the Low Countries makes it harder to see a non-IE language for the original BB's, presuming that area is where they originated.

So, that almost complete absence, plus the R1b in Hallstatt would make an IE dialect as original language for BB parsimonious.

J.S. said...

@ Andrzejewski

"40% of France north of the Loire is inhabited by Germanic Franks, Burgunds, Alemany etc, "

Are you trolling?

Matt said...

@Simon W, when I say "there's no attestation of any Basque like languages outside Iberia", I mean in the wider sense that supports Basque languages ever being in Central Europe or Eastern Europe or the steppe, or anything that provides positive evidence for their presence at this date far from Iberia.

Gaska said...


@Davidski-"You first have to prove somehow that the predecessors of these languages weren't their native languages before moving on to models of how and where they acquired them"

Exactly and the genetic continuity is so evident between the chalcolithic and the iron age that it is materially impossible to do so. Now we are talking about conclusive evidence, not speculation, written Iberian texts that show that it is a non-Indo-European language and continuity of the uniparental markers.

I do not know what is the origin of Basque, Iberian, Aquitanian, Etruscan and Tartessian (Caucasus ?, Anatolia ?, Sardinia ?, European Neolithic? Western hunter-gatherers?), But it is evident that there could be a Pan-European language linked to the Neolithic farmers. Then the solution may be in Germany or France, taking into account the Non-Indo-European issue of Aquitaine, and all the north of the Pyrenees.

It is increasingly evident that the third wave of Gimbutas related to the Yamnaya culture only reached Hungary and Poland where Z2103 was found, then if R1b-M269 has its origin in the steppes, its arrival in Central Europe had to be very old, 2,000 or 3,000 years before the BB culture, enough time for some clan to change their language to that of the Neolithic farmers.

Davidski said...

@epoch

But isn't that exactly the reason why we can assume that the original language of R1b "tribes" is an IE dialect?

Yes, you mean R1b-M269 tribes of course. That's indeed a very plausible assumption, but that's all it is at present. Whether it's confirmed beyond reasonable doubt, or totally debunked, depends on what the ancient data will show in the near future.

But I'm not saying anything new. After all, that's why a lot of us are still here taking part in these discussions. Nothing's cut and dried yet, especially not the details.

But the almost complete absence of non-IE toponyms and hydronyms in the areas of NW and NE France and the Low Countries makes it harder to see a non-IE language for the original BB's, presuming that area is where they originated.

My feeling about toponyms and hydronyms has always been that they're not a natural feature of the landscape. In other words, they obviously have to be passed on by someone, and we have to consider the possibility that sometimes they're not.

So, that almost complete absence, plus the R1b in Hallstatt would make an IE dialect as original language for BB parsimonious.

Actually, I haven't yet really questioned the idea that the early Bell Beakers with steppe ancestry were originally and by and large Indo-European speakers. I think they were, but I now suspect that many Beaker groups eventually switched languages to whatever was useful, including non-Indo-European languages.

But since you ask, I have to reiterate my comment above that nothing is cut and dried, and it's not a given that any Bell Beakers were Indo-European speakers.

More data from Urnfield and Hallstatt groups might play an important role in eventually deciding this issue. I suspect that if these groups show significant genetic substructures that include non-Beaker ancestry which links them to attested Indo-Europeans in the east, then that might screw things up for the association between Beakers and Indo-Europeans.

In fact, as you probably know, one of the Hallstatt samples does show that type of ancestry from the east. So if this sort of stuff can be linked directly to the expansions of Indo-Europeans into Western Europe, then things might get even more interesting than they are now.

Aram said...

Hallstat's heartland was in Alpe were later a Raetian is attested. Kind a strange that in supposedly proto Celtic homeland a non IE language could survive.

Urnfield expanded into East Iberia and Ertruria (Villanova) were later again two non IE languages are attested.

So why Urnfield, Hallstat can be multiethnic places while BB cannot? It must be at all cost 100% Basque even if no one have seen any Basque influence in Italy for example.

This logic feeds trolls like Carlos imho.






Davidski said...

@Aram

Hallstat's heartland was in Alpe were later a Raetian is attested. Kind a strange that in supposedly proto Celtic homeland a non IE language could survive.

Urnfield expanded into East Iberia and Ertruria (Villanova) were later again two non IE languages are attested.


Right, well there you go, I just basically suggested that Urnfield and Hallstatt might have been Indo-Europeanized from the east. So that would explain your observation that they were multi-lingual. Haha.

But seriously, we need more data, especially from that critical period when these ancient archeological cultures morph into groups that are attested as speakers of Indo-European and non-Indo-European languages.

Heraus said...

No one has stated that the BB world was 100% "Basque", just that Basque appears to be a remnant of a language family that one could name Sardo-Vasconian the many lexical roots of which still name many hydronyms and oronyms in Western Europe (the most famous one is *kar/kal "rock").

As for pre-IE placenames in Italy, if we except Sardinia which is its own case, it is not difficult to find Ligurian placenames in NW Italy ending in -asco or -asca (just like in neighbouring Provence). Nobody really knows what ancient Ligurian language was about : labelling it Indo-European is rather daring. Strabo specifically said Ligurians were not Celtic initially.

epoch said...

@Davidski

Food for thought.

One thing about hydronyms though. You are right that one cannot say why a certain hydronym changed with a change of language or not. America has kept Indian hydronyms and toponyms where the Cape has not kept original Bushmen hydronyms (apart from one, I just googled).

But there are certain details which are quite interesting. Take for instance the name of the river Meuse: This french name ultimately came from proto-Celtic "Mosa". However, sound laws require that the Dutch name for the river, "Maas", is retreived from a proto-Germanic "Maso". Still it is undeniable that both these names have the same origin, but the name went through Celtic and Germanic sound changes independantly. That means that the river must have carried its name from before the Iron Age. As its name is explainable from a PIE root this means we can attest PIE in the Bronze age. A similar thing goes for "Rhine".

Desdichado said...

Razib Khan adds his two cents: https://www.gnxp.com/WordPress/2019/03/20/the-population-turnover-in-westernmost-europe-over-the-last-8000-years/

Ebizur said...

"No one has stated that the BB world was 100% 'Basque', just that Basque appears to be a remnant of a language family that one could name Sardo-Vasconian the many lexical roots of which still name many hydronyms and oronyms in Western Europe (the most famous one is *kar/kal 'rock')."

cf. Ancient Greek χᾰ́λῐξ */kʰá.liks/ → /ˈxa.liks/ "small stone, pebble; gravel, rubble (used in building and concrete making)"

Latin calx */kalks/ "limestone; chalk; the finish line"

Armenian քար [kʰɑɾ] "stone, rock"

Mongolian ᠬᠠᠳᠠ ‎qada > хад had "boulder, rock, crag; cliff, ravine"

Manchu hada "boulder, rock, crag"

Evenk кадага kadaɣa "rock, cliff"

Yakut хайа xaya "mountain"

Uzbek qoya ‎"rock, cliff"

Turkish kaya "huge rock, large rock mass; rock cliff, rock precipice, palisade"

Tundra Nenets хой xoy "hill, tundra"

Hungarian hegy (sounds like the English word "hedge") "mountain; point (of pencil, knife), tip (of finger, tongue, nose)"

I do not see why toponyms containing a sequence /kar/ or /kal/ should be presupposed to be related to Basque harri "stone" in the absence of corroborating evidence.

Ric Hern said...

@ epoch

Kareiga. Gamtoos, Keiskamma, Kei, Tarka.....mostly Eastern Cape.

Ric Hern said...

@ epoch

The problem with preserving Khoi-San Hydronyms is/was basically the Clicks in those Languages which most Europeans find very difficult to pronouns....it is not just about making any click sound, there are variations....

epoch said...

@Ric Hern?

I probably should have said Western Cape, but already saw Gamka there as well. Anyway, the gist of my post is that in the case of European hydronyms absence of proof is certainly not proof of absence but that we actually can say something from some names: Their IE origin must be earlier than proto-Celtic and proto-Germanic.

Ned said...

Have I got this right? I can see three realistic scenarios (but let me know if they're not):
a) R1b I-E herdsmen entered Iberia and lived in a symbiotic relationship with native farmers in which time they adopted the farmer's tongue. Then after 550 years some calamity (famine, plague, massacres) happened and the farmers all died out.
b) Sredny Stog was R1A I-E and developed into Corded Ware and Khvalynsk R1b pre-Basque developed into Bell Beaker after enormous slow migration. R1A elite gained control of key Beaker clans somehow and Germanic (R1a-664) and Celtic became I-E through elite dominance. Those R1b who stayed at home somehow became Bashkir speakers.
c) Basque is a highly aberrant I-E language.

Ric Hern said...

@ epoch

Yes I agree, and I pointed out why Hydronyms could have changed. People just couldn't pronouns some of the native peoples words.

Kristiina said...

@ Ebizur

The root "kar" in the meaning rock, stone is very widespread indeed. It is found everywhere in Eurasia, and more so if we combine it with the root "kal".
The closest forms to the Basque form are the following:
Armenian kʰɑr rock
Breton karreg stone
Welsh carreg stone
Kabardian q'ər stony ground, rock;
Ingush qera/qiera stone, rock
Agul qIarqI rock
Urartuan qar=w rock

On a somewhat different note, the arabic word for cave is "ġār" and also comes close.

Considering that modern populations all share genetic components (e.g. most Europeans share WHG, CHG, Basal and EHG), I think that it makes sense to believe that certain macro-roots such as "kar" in the meaning of "stone", "rock" are ultimately derived from common roots.

Kristiina said...

What I argue soon is somewhat speculative, but probably of interest to many.

The distribution of the root kar ‘stone’, ’rock’, including the Basque word 'harri', could easily be linked with CHG, while the Proto-Slavic form *gora (mountain), Lithuanian gùras (ledge) and the Uralic forms kúrək (Mari) and gureź (Udmurt) (mountain) to CHG part in EHG.

If we presume that roots spoken by Basal-rich populations should be found in particular among the Afro-Asiatic speaking language groups, it is of interest to look for roots meaning ‘stone’ in Afro-Asiatic languages. This is the result:
Romanized Arabic ħájar
Romanized Hebrew ʾeven
Romanized Syriac kefā, ʾavnā, šoʿā
Romanized Akkadian abnu
Romanized Ge'ez ʾəbn
Maltese ġebla, ħaġra
Romanized Coptic p-ōne
Tarifiyt Berber az’r’u
Hausa dúutsèe
Oromo (Wallagga) d'agaa
Somali dhagax, shiid

Only the Arabic and Maltese forms bear some resemblance to the form ‘kar’. On the basis of this small analysis, I would link the Basque word harri ‘stone’ with CHG and not with Basal/EEF.

capra internetensis said...

The arrival of Steppe and Iranian related ancestry in the islands of the Western Mediterranean

https://www.biorxiv.org/content/10.1101/584714v1

It's like the party was winding down and someone shows up with a fresh keg

Gaska said...

Thanks in Spain we are celebrating, we just need our French friends to cheer up to celebrate with us

Gaska said...


D. Fernandes-

In Sicily, Steppe pastoralist ancestry arrived by ~2200 BCE and likely came at least in part from Spain as it was associated with Iberian-specific Y chromosomes.

Now you will understand why we said that the markers related to the BB culture in Western Mediterranean (Liguria, Balearic Islands, Morocco, Sicily and Sardinia) came from Iberia.

Gaska said...

In Iberia, most people buried with artifacts of the Beaker complex had little if any Steppe pastoralist-related ancestry (from here on denoted “Steppe ancestry”), but Beaker cultural practices were adopted by people in Central Europe were in part descended from Steppe pastoralists and then spread this material culture along with Steppe ancestry to northwestern Europe. In Iberia, Steppe ancestry began to appear in outlier individuals by ~2500 BCE4, and became fully mixed into the Iberian population by 2000 BCE5.

Gaska said...

In SicilySteppe ancestry arrived by ~2200 BCE, and likely came at least in part from the west as it was associated with the Iberian-specific Y haplogroup R1b1a1a2a1a2a1 (Z195), thus documenting how Iberia was not just a destination of east-to-west human movement in Europe, but also an important source for west-to-east Steppe ancestry reflux. In Sardinia, we find no convincing evidence of Steppe ancestry in the Bronze Age, but we detect it by ~200-700 CE.

Grey said...

Samuel Andrews said...
"Also, pre-Muslim, Granda samples surprisingly have the same level of North African ancestry."

Granada covers a lot of the same territory that was colonized by Carthage.

https://cdn.britannica.com/42/1042-004-25A2A76C.jpg

http://www.islamicspain.tv/For-Teachers/maps/Spain%20Granada%20Color.jpg

Grey said...

Andrzejewski said...
"It's all just hard for me to believe that of all the post-BB invasions: Romans, Germanic, Arab, Berber, etc, almost NONE left almost zero trace in modern populations. Let alone the Visigoths..."

one possibility might be you have an original population A who live in the countryside and strategic towns and invader B comes along who massacres and replaces the A population in the strategic towns but not the countryside and then invader C comes along and massacres and replaces the B population in the strategic towns but not the countryside and then invader D comes... etc.

maybe to transform a population genetically an invader needs to take over the productive countryside - especially if it's true that urban centers were population sinks?

Simon_W said...

@ Heraus

"As for pre-IE placenames in Italy, if we except Sardinia which is its own case, it is not difficult to find Ligurian placenames in NW Italy ending in -asco or -asca (just like in neighbouring Provence). Nobody really knows what ancient Ligurian language was about : labelling it Indo-European is rather daring. Strabo specifically said Ligurians were not Celtic initially."

Ha! Of course the Romans who noted a difference between Celts and Ligurians were quite right, but not being Celtic doesn't equal being non-IE. And a non-IE toponymic substrate doesn't prove that ancient Ligurian wasn't IE either, because it's the nature of a substrate to stem from older times. During the last decade the Italian linguist Adolfo Zavaroni has done a number of studies on Ligurian inscriptions from the Tosco-Emilian Apennine (i.e. from the area of the Friniates) and from Liguria (from Beverino, La Spezia and from Campocatino in the Apuan Alps). They all used the same script and the same language, and his findings prove beyond doubt that it was an IE language of the Italo-Celtic branch. See for instance:

https://www.academia.edu/23414809/LINGUA_SCRITTURA_DIVINITÀ_E_ARTE_RUPESTRE_DEGLI_ANTICHI_LIGURI_5._L_ISCRIZIONE_DI_BEVERINO_LA_SPEZIA_2016

https://www.academia.edu/24851735/La_lingua_degli_antichi_Liguri._Iscrizioni_e_figure_sacre_su_due_rocce_di_Campocatino_Alpi_Apuane_

Simon_W said...

As for the similarity between Empuries2 and the Mycenaeans, I find it curious that nMonte likes to use Empuries2 abundantly to model my own coords, but is completely reluctant to use Mycenaeans as a substitute:

https://justpaste.it/60os3

There must be some subtle difference causing this, but what? I doubt it's due to Iberian admixture in Empuries2.

Bogdan said...

I believe over the course of humankind until today, the spread of dominant languages is mostly attributable to trade and trade network dominance. It is not unreasonable to assume that prior to establishment of more sophisticated trade networks, one could walk 100km in any direction of Europe and encounter people that spoke an entirely different dialect and/or language...

As stated by Davidski and others previously, “genes don’t speak languages, people do”.....

Bogdan said...

Kristina said:

“IMO, these two exaples with basic roots show that it is not so difficult to find a connection between Basque and the Caucasian languages”

Funny, but one of the observations my spouse (Russian native) stated in our first trip to Basque Country was how the local language intuitively SEEMED to be similar to Georgian and/or Caucuses languages, yet obviously very different....

Ever asked yourself why there are, even to this day, so many language variations that exist in the Caucuses? Why do they persist?

Why were the Iberians writing in a Phoenician / Greek tablet alphabet circa 800-1000 BCE? (If not much earlier) And why were the Celts hauling ass down to Huelva around the same time?

bellbeakerblogger said...

I'm late to this discussion, but see my comments at
https://bellbeakerblogger.blogspot.com/2019/03/beaker-language-and-badass.html

They were a little too long for the comments space.

epoch said...

@MomOfZoha

Sorry, missed your response, so my answer is late.

"On an independent note, I just cannot comprehend why it is so difficult for people to understand the multi-lingual conglomeration approach to "steppe" explanations. Given that more recent historic variations of steppe nomads are known to have been multi-lingual (and multi-ethnic) conglomerates -- something that most likely proved to their advantage -- why in the hell could that not have been the case in earlier iterations of "steppe people"..."

There should be a massive exchange of loanwords would the be the case. We don't see much - not any at all, actually - evidence of Basque in any IE language apart from in SW Europe.

Take e.g. Hungarian, which has a massive amount of Turkic loanwords. Or proto-Uralic, which sees IE loanwords from all phases of IE presence on the steppe.

andrew said...

@Davidski

"The genetic impact of the Bell Beaker and related migrations from the east was very significant from the start, and caused the local Y-chromosome lineages to basically vanish in maybe less than 500 years.

Based on the available data, this process looks like a wavefront that eases somewhat with time, but continues at a significant rate into the Iron Age.

I can't see anything gradual there. It's all very sudden."

Steppe ancestry per Olade starts to show up ca. 2400 BCE. The Bell Beaker phenomena starts to show up ca. 2800 BCE. The data is sparse enough that it is easy to miss developments that happen in the four hundred year period before the deluge when BB culture is formative. We have very little data in this critical formative time period during which BB ethnogenesis takes place.

In culture and language, there is a huge premium to being first, and you don't have to be first by all that many generations to have a pervasive impact on later arrivals.

For example, while a very large percentage of white Americans trace their origins to late 19th century and early 20th century immigration from Southern Europe (especially Italy), Germany, Ireland and Scandinavia, the linguistic impact of the later migrants on the languages spoken in the U.S. was meager. (Cf. in another illustration of the disconnect between the population genetic share of migrants and their linguistic impact, Yiddish has has as much or more impact on the English language than Italian, 19th century or later German influence, Irish, or the Scandinavian languages, despite the number of migrants from those places being much greater. Likewise, migrants to the U.S. from Japan and China in the 19th century had virtually no linguistic impact.

On the other hand, while French migration to North America has been minimal for centuries and makes up a pretty small percentage of the U.S. population's ancestry, these people have substantial impact linguistically that remains important in Louisiana (e.g. French Creole) and Quebec, and has had even greater impact of North American toponyms. (Similarly, the Dutch whose political role was gone in the 1600s, have had a significant ongoing impact regarding the phonemes used in the NYC dialects of American English and word usage and toponyms in upstate NYS.

To have linguistic impact, the newcomers have to have more socio-economic prestige than the people before them. But, if there was already a steppe ancestry elite in BB culture that developed in the time period from 2800 BCE to 2400 BCE which doesn't show up in our limited data set, the BB culture may have had enough prestige to prevent later steppe migrants from displacing the EEF language adopted by the early steppe migrants.

Is this the right answer? I don't know.

Is this a feasible model which isn't ruled out by the evidence including Olathe? Yes.

Urki said...

This is a recent fact. As far as we know France has always been the demographic core of Europe, at least until the end of middle ages. Until then population density was lower eastward and westward

olga said...

Some hypothesis about the replacement of male haplogroups in Spain.
1.-There was no replacement because they were natives, but there are not enough samples to prove it.
2.They were replaced by mass killing of men. There is no archeological proof of this circunstance.
3.- They were replaced by reducing them to a sort of slavery to the bottom of society, driving women to take partners among the most succesful foreigners.
4.- Men sterilizing diseases like Mumps could have been brought from East, diminishing the rate of reproduction of native men.
5 The foreigners had a similar authosomal genoma than the native iberians, but with some more steppe and the Haplogroups R1b related.
6.- The foreigners spoke a sort of vasconic pidgin that enriched the iberian vasconic language with new words.

7.-The iberian vasconic language is the product of different layers of paleolithic natives, neolithic coming from the Aegan Sea hopping through the mediterranean islands, or crossing the Balkan and Italic península and the Atlantic Megalithic Culture and was spoken in the Iron Age south of the Garonne, diminishing because of the roman acculturation and the settling of visigoths in their lands turning the owners to serfdom. This last situation could have force people to run and cross the Pyrenees looking for shelter in the Saltus Vasconum or Basque Mountains, the actual Spanish Basque Provinces. This land was poor enough to keep the romans at the distance, keeping the language and the old ways.
8.-The following words or roots to form new words are found in many places.
Urarte: Between rivers Ur= water arte=between Ur and Uruk were located between two rivers and the región was called Urartu, Mesopotamia in Greek.
Go/ga materialized in Gora: High, altitude goia, gana ara all referred to the top of the mountains. We find that Word from the Caucasus to the Balkans. Montenegro=Czarna Gora
Be= down, at the feet Gabe, Garve , Algarve
Iri /ili/uri are references to urban places. Ilion= Troya Ur, Uruk cities in Mesopotamia
Berri =new Illiberris= New village
Sardinia refers to a bank of fish
Ibai, iber refers to a river, or the Banks of a river. Caucasus, Aegan Balkans, Italy ,Spain. Iber, Tiber
Zaldi = horse zeldo=cart in Hittita coming from de Luwian according to Parpola
Ill= death Moon= Illargi= Light of the deaths related with english ill?
Argi= light the root of Argentum?
Urumea (river in Euskadi) Urmia (lake in Azerbaijan
Itz refers to water. Itsaso= sea
Otz = cold The root in many hydronimes and places actually non basques in Spain. Oja, oca

Bihotz= heart Bios=life
Seme=son
The direction goes to the East, but not to the North of the Black Sea, rather to the South.
I am an amateur of the subjet, so the spanish guys may correct me.

DAMIANGOYS said...

just look at iberian autosomal test results and you will see the clues all in there about the Iberian back to late bronze early iron age , it is pretty clear, that they come from central Europe Balkans , DNA DOES NOT LIE

«Oldest ‹Older   201 – 310 of 310   Newer› Newest»