search this blog

Thursday, February 23, 2017

Kurgan waves as far as the Atlantic


Paleogenetista Carles Lalueza-Fox is currently collaborating with Harvard on a major ancient DNA paper on the population history of the Iberian Peninsula, which is scheduled to be published next year. He was recently interviewed by Spanish Newspaper LaVanguardia about the project. Here are a few interesting quotes (pardon the translation):

And the mixing of genes continued...

Yes, 4,000 years ago came the Kurgan people, who domesticated the horse on the Pontic-Caspian Steppe and spoke Proto-Indo-European, the ancestral language to Celtic, Latin, Greek...their impact on our DNA was high.

What does that mean?

Their numbers grew at the expense of the previous population: today the Kurgan genetic footprint makes up 40% of the Western European genome!

Also in the Iberian peninsula?

Fifty per cent of our ancestry is derived from Neolithic farmers, 30% from the Kurgan people and 20% from hunter-gatherers.

Note the date: 4,000 years ago. That's not a Bell Beaker date, it's an Atlantic cist tradition date. What the hell is the Atlantic cist tradition, you're probably asking. Have a look at the video here, or, if you're too lazy, this screen cap.

See also...

Yamnaya-related admixture in Bronze Age northern Iberia

137 comments:

Samuel Andrews said...

I still think the U5a1b1 in Spanish Bell Beaker is of Steppe origin. The fact he didn't mention anything about North Africans could mean North African ancestry in modern Iberians is of recent(Moorish) origin.

Davidski said...

The fact he didn't mention anything about North Africans could mean North African ancestry in modern Iberians is of recent (Moorish) origin.

He said that in another recent interview about the project. But it's obvious that the vast majority is recent, without the Paleogenetista telling us that.

Unknown said...


new study:
http://www.pnas.org/content/early/2017/02/17/1616392114.abstract

Helgenes50 said...

With Global and nMonte
we get the same results

[1] "distance%=0.3144 / distance=0.003144"

Spanish_Cataluna

Iberia_CHMN 42.20
Steppe 31.45
Lengyel 23.30
Levant_Neolithic 3.05

Gioiello said...

"Yes, 4,000 years ago came the Kurgan people, who domesticated the horse on the Pontic-Caspian Steppe and spoke Proto-Indo-European, the ancestral language to Celtic, Latin, Greek...their impact on our DNA was high" (Lalueza lalala).


Of course this explanation doesn't fit with the data of the differentiation of the R1b haplogroups in Europe, and the YFull data are underestimated at least for an 1.26 factor (if A00 separated more than 300000 years ago as to Shi Huang and his colleagues) or 1.17, if Poznik is right (275000 years ago). Now I understand why the paper on BB is late to be published: these geniuses of Harvard are stil making their calculations to manage the autosome.

Davidski said...

@Gioiello

Human Y-chromosomes are portable, including the mutations they carry. Think about it.

Karl_K said...

These 4,000 year old people will be very similar to the Irish Bell Beakers from 4,000 years ago. Weren't they all R1b1a2a1a2c-L21 ? Which is still amazingly high in both Ireland and Basque country to the present day? Weren't they also U5?

Karl_K said...

So the key point would be that the early Iberian Bell Beakers did not have steppe ancestry, but there was a massive migration of men and women to the Atlantic coast from somewhere with high Yamnaya-like ancestry. These immigrants fused with the original Bell Beaker locals and then their expansion really took off.

What reasonable options are there besides boats and beer. They immediately hit the coastline, rivers, and the islands. Especially around copper rich areas.

André de Vasconcelos said...

@Sammy

"The fact he didn't mention anything about North Africans could mean North African ancestry in modern Iberians is of recent(Moorish) origin."

Maybe, maybe not, it could mean a lot of things or nothing at all. It wasn't relevant for the interview in which he was just stating the general large impacts in Europe and a lot of small things about certain individuals. Also that all WHGs had blue eyes.

The Moorish period thing doesn't look very compatible with its historical and geographic distribution - more in the West, less in the East. Why do Galicians, Leonese and Asturians score more NA ancestry than Valencians, when the former were only occupied for decades - or less than that - and the latter up to 1238 (ignoring the few years in which it was controlled by El Cid)?
Can also mention Aragonese and Catalans, the scenario is the same. Tarragona retaken in 1117, Zaragoza 1118, Tirwal(Teruel) 1171, Tortosa 1148, Huesca 1096 (briefly Carolingian by early 800s).

Not saying the Moors didn't leave anything, but I feel an earlier origin is of significant importance, more so than the early middle ages.

Al Bundy said...

In a 2015 interview Fox says he doesn't think the Anatolian hypothesis has been killed,based on the Haak paper and the Chang linguistic paper.Seems like he's changed his mind.

Alberto said...

I wouldn't be surprised if early Iberian Bell Beakers were like MNChl samples we have. But it would be more surprising if during the full Bell Beaker period (c. 2200-2000 BC) where they share the same cultural package as Central and NW European ones they're not R1b and have steppe ancestry. It would be a strange case of cultural diffusion without gene flow.

I'm not sure about this Atlantic Cist Tradition and how it relates to the later and better known Atlantic Bronze Age, but one of the points that Koch makes about it, is that it might have spread Celtic languages along the Atlantic, but this left the Basque Country untouched. So it doesn't fit well for R1b (not only about Basques, but in general all of Iberia, where there is no gradient in R1b frequencies from the Atlantic to the Mediterranean). Besides, the most important and dominant culture from this time in Iberia is El Argar (SE Iberia), which is not derived from Bell Beaker and has more links to the Eastern Mediterranean, so again this would make a difference in the genetics of Mediterranean Iberians and Atlantic ones.

I hope the paper doesn't take to long to come out and can shed some light on all these questions.

Matt said...

Compare to nMonte result from merge of Days of High Adventure and Globe10 PCA together:

Spanish: Europe_EN - 56.95, Iberia_Mesolithic - 23.55, Kotias - 13.25, Steppe_EMBA - 6.25 (all of Latvia_LN1:ZVEJ28, Ukraine_N1:StPet12, Loschbour, Hungary_HG, Motala_HG - 0)

Slight preference for more local HG than Lalueza-Fox's (formal stat based?) estimate and a more Kotias rich introduction.

General results across Europe: http://i.imgur.com/tza6Laz.png
General result breaking Europe EN into 91% Anatolian, 8% Loschbour (how nMonte fits it on these data in accordance with the formal stats): http://i.imgur.com/DE88psZ.png / http://i.imgur.com/nkl5WNC.png

alobrix said...

@Helgenes50

If the Atlantic Cyst people was the vector to steppe genes in Iberia, then you should find higher steppe percentage genes in Galicia. In fact, there are cultural affinities between Ireland and Galicia even nowadays. But I doubt it.
In the Iberian Peninsula there are genetic structure, with the main differences between atlantic and mediterranean peoples.
http://www.ashg.org/2014meeting/abstracts/fulltext/f140122918.htm

Could you try Global and nMonte in Galicia and Portugal?

Heraus said...

As for NA admixture, we don't know anything about what Iberian Neolithic people were like in the whole peninsula : as far as I know, we only have samples around Atapuerca and the Ebro valley where "Sardinian-like" people gradually become "Basque-like" through admixture (either local and residual WHG admixture or early steppe genetics, or probably both).

It is much probable to speculate about Neolithic links between Western Iberia (from Asturias to southern Portugal) and coastal North Africa : Moorish invasions cannot explain Galician or Leonese genetic results. Iberia is a vast land, with many hydrographic basins and local pecularisms up to nowadays. No reason to believe it was different by then (more specifically before proto-IE people conquered and united those lands).

André de Vasconcelos said...

@alobrix

Atlantic Cyst could have been one source, but it didn't have to stop there, so assuming a gradient based on it might be misleading. Galicia does not have high levels of steppe ancestry within Iberia, in fact it's lower than the regions I mentioned up above, while also having higher EEF-related ancestry.
Ironically more Mediterranean (geography) Iberians are less Mediterranean (genetically), with the exception of Andalusians. The East-West cline points to a CE origin of steppe genes in here, thus from behond the Pyrenees - although, again, there could have been more sources coming from elsewhere.

I'd be interested in seeing differences between North and South Portugal, with the split being around the mouth of the Mondego river, and the Serra da Estrela mountain system. Visual geographic help: https://upload.wikimedia.org/wikipedia/commons/c/cd/Portugal_topographic_map-pt.png

Karl_K said...

@Alberto

"one of the points that Koch makes about it, is that it might have spread Celtic languages along the Atlantic, but this left the Basque Country untouched. So it doesn't fit well for R1b (not only about Basques, but in general"

I don't think there is a big problem here.

I know that many people are skeptical that Celtic languages in Ireland came directly from R1b carrying Bell Beaker immigrants 4,000 years ago. Yet we have genomes from those immigrants, and they are extremely similar to the modern inhabitants. There is now no reason to suppose any other option without evidence to the contrary.

So. Let's assume that is true, and the R1b carrying Bell Beakers that arrived by boat to Ireland 4,000 years ago brought Celtic languages.

What is their relationship to the Atlantic coast and Iberia?

We know that the closest ancient genome we have to modern R1b-heavy Basque is a Bronze Age person with clear Yamnaya-like ancestry, but lower than in the Irish Beakers.

So, probably some groups of Iberians had a limited inflow of immigrants, who were almost 100% successful males. The original language survived along with a culture that remained for thousands of years.

In other places, whole families of immigrants arrived and made a much larger change in the language and genetics of many areas.

In general, the more steppe-like ancestry a population has, the more likely it is to speak an Indo-European language. The R1b story in Iberia and Sardinia and other places is interesting, because it shows that R1b men in Western Europe had some kind of ability to have successful sons after joining into a new culture. Their advantage was not their language.

Alberto said...

Yes, the North African connections seem to go back in Iberia at least to the Early Neolithic. It's already possible to detect that signal in those samples, even if they're not from the south. The signal weakens in the MN, but there's probably another pulse in the Chalcolithic. And then from the Iron Age to the Middle Ages, there are continuous contacts. It's actually surprising that the overall genetic impact is so small, given the geographic proximity.

A paper about the Early Neolithic contacts:

https://www.researchgate.net/publication/235915451_The_Mesolithic-Neolithic_transition_in_southern_Iberia

Helgenes50 said...

@ alobrix

Here's what I get with the same populations
for Galicia and Portugal

"distance%=0.3144 / distance=0.003144"

Spanish_Cataluna

Iberia_CHMN 42.20
Steppe 31.45
Lengyel 23.30
Levant_Neolithic 3.05

"distance%=0.4123 / distance=0.004123"

Spanish_Galicia

Iberia_CHMN 53.15
Steppe 31.10
Levant_Neolithic 14.40
Gambian 1.35

"distance%=0.2557 / distance=0.002557"

Portuguese

Iberia_CHMN 53.4
Steppe 29.9
Levant_Neolithic 14.7
Gambian 2.0

Alberto said...

@Karl_K

The problem I have with this is why should we prefer a model where R1b entered Iberia after 2000 BC with a culture that only flourished in the lesser populated Atlantic area, when we have a preceding culture (Bell Beaker) that flourished throughout all of Iberia and seems to explain better the modern distribution of R1b (i.e, high in all the territory)?

Unless people here already know that Bell Beaker samples from 2200-2000 BC from Iberia are not R1b and have no steppe admixture, and then we're forced to look for an alternative. If this is the case, then ok, we have to look for more complicated scenarios.

Karl_K said...

@Alberto

OK. We will have to wait and see. You are suggesting that the earliest Beaker Folk in Iberia probably already had R1b and steppe ancestry. Is that correct?

Rob said...

@ Alberto & Karl_K

A possible commodification could be that there was indeed flow during the Cist period, but perhaps this was limited to a northern subgroup, like L21, and perhaps would not explain the overall pattern of L51.

Karl_K said...

@Rob

But that doesn't change much. A Yamna man from Samara 5000 years ago was R1b-P297, so both L21 and L51 are steppe brothers.

So perhaps L51 was an earlier Bell Beaker associated haplogroup, but later L21 influx pushed up the steppe ancestry in Iberia with the cist culture?

That leaves a lot of questions about where exactly BB culture originated, and by whom.

Heraus said...

Now it is quite clear that proto-IE people reached - rather massively - Iberia (well it has been known for centuries actually thanks to Celtiberian placenames), how come the Iberian language, with obvious pre-IE roots, managed to survive ?

Must we theoricize more than one wave of Kurgan influx into Iberia (one from the East Med through naval contacts, another one crossing the Pyrenees)? Was Eastern Iberia already too organized to lose its autochtonous language despite massive influx ? It looks like the Celts might have been latecomers preceded by other people with already Kurgan genes ...

The ancient Cantabrian language might be a hint : after all, it looks like most Cantabrian surnames are not Celtic but only generic IE. A remnant of a first wave of IE people into Iberia ?

Or are calculators all wrong when it comes to Kurganor Yamnaya genetics ?

Alberto said...

@Karl_K

You are suggesting that the earliest Beaker Folk in Iberia probably already had R1b and steppe ancestry. Is that correct?

No, I said that I wouldn't be surprised if early Bell Beakers from Iberia (say, 2600-2400 BC) looked like the preceding MN-ChL population. But I would be surprised if the later Bell Beakers (c. 2200-2000) would also look the same, and had no R1b and steppe ancestry.

@Rob

Yes, I think that the Atlantic connections can explain some, but not all of the R1b in Iberia (or at least that would be my assumption for now).

alobrix said...

@Helgenes50

Thanks. It suggests a different pattern of neolithisation in northwest and northeast Iberia. Interesting.



Karl_K said...

@Heraus

"It looks like the Celts might have been latecomers preceded by other people with already Kurgan genes ..."

Could you put dates with your timeline? What do you mean by "latecomers"?

We know that likely Celtic speakers from 2026-1885 BC in Ireland were buried in single burial stone cists and had lactase persistence, blue eye color, Y chromosome R1b-L21 haplogroups, and the hemochromatosis C282Y allele. These graves were from exactly the time when copper mines, bell beaker pottery, etc. started appearing.

You then are suggesting that there was an earlier spread of R1b people over Iberia (when? 3000 BC?) that was somehow distinct from these immigrants to Ireland?

Palacista said...

This is all about the Celtic from the West books that admittedly do look interesting. The "Kurgan waves" may have reached the the Atlantic, but this talk is about a wash back from the British Isles that spread the Celtic languages in France and Iberia after forming from general IE in Ireland/Wales/Scotland.

Rob said...

@ Karl_K

* "But that doesn't change much. A Yamna man from Samara 5000 years ago was R1b-P297, so both L21 and L51 are steppe brothers.
So perhaps L51 was an earlier Bell Beaker associated haplogroup, but later L21 influx pushed up the steppe ancestry in Iberia with the cist culture?"



-> I was specifically replying to Alberto's statement about the R1b link with the BA Atlantic system has some difficulties, not hypothesizing about whether or not R1b-L51 expanded from the steppe.

As for Beaker, who knows there are still several possibilities, including the long recognized fact that it wasn;t a homogeneous phenomenon, and (as I've oft argued) modern pattern continued to evolve until recently.

* "That leaves a lot of questions about where exactly BB culture originated, and by whom."

-> I get the feeling that BB isn't from the steppe at all, some steppe derived groups joined it and expanded within it c. 2500 BC. There's a trail from Yamnaya to Hungary and SW Poland of individuals of steppe origin arriving to central Europe, entering the BB network at expanding, sometimes with force, as shown in the site at Le Petit Chasseur.


* "We know that likely Celtic speakers from 2026-1885 BC in Ireland were buried in single burial stone cists and had lactase persistence, blue eye color,"

Karl, how do languages like Ligurian, Italic, Venedic, etc fit in your scheme ?
If BB was "obviously Celtic", how did (more 'archaic') Lusitanian get to western Iberia ?

In fact, much of BB might not even have been IE.

Karl_K said...

@Rob

I don't really have a scheme.

Genetically and culturally, Ireland looks like a clear case. Around 2000 BC, the European Neolithic population was largely replaced with a population that looks extremely similar to the people who still live there today.

That gives strong suggestions about the Celtic language group and R1b-L21.

But, as you suggest, the Italic language group looks like it may have spread with a distinct migration of a people that were quite similar genetically.

But was their last common homeland still in the steppe area, or somewhere else?

Rob said...

@ Karl_K

I don't disagree at a broad level, but some important points remain:

1) there were several other IE languages in western Europe aside from Celtic

2) Celtic cannot be solely be linked with L21 (at least not on basis of present evidence)

3) The Bell Beaker phenomenon as a whole probably didn't originate on the steppe.

So one possibility is that BB indeed began in Iberia, was therefore not IE, apart from in the east which was infiltrated by steppe groups. You're probably aware the 2000 BC figure you quote above is after the BB culture ends

Plains Wanderer said...

@Karl_K

We shouldn't make too many assumptions about language from only genetic evidence. Yes, the British Isles became mostly genetically modern around 2200-2000 BCE, but that doesn't necessarily mean that they were linguistically modern as well.

Does linguistics support the idea that Celtic languages could have spread this early? It has long been suggested that Celtic languages and culture spread during the Urnfield or Hallstatt period at the earliest and that they spread with relatively minor demic impact. The genetic evidence that we have so far still supports this possibility.

Karl_K said...

@Rob

I agree with you all around. I was simply trying to tie R1b-L21 to Celtic based on the actual ancient and modern DNA evidence we have from Ireland.

I was not suggesting that the language and Y-chromosome went hand in hand at all times in all other areas of Europe.

I think the term "Bell Beaker Culture" confuses a lot of this discussion. It is just not precise enough. I am aware when the "BB culture ends", yet there are strong indications of continuity.

So most likely the overall "BB culture" was actually an example of an easily passed cultural package. But there were groups who picked up key parts of it and ran with it all over Europe, and by 2600 BC the Bell Beakers in Germany had R1b-M269 and an autosomal composition quite similar to modern North-West Europeans (such as the Irish).




Karl_K said...

@Plains Wanderer

"It has long been suggested that Celtic languages and culture spread during the Urnfield or Hallstatt period at the earliest and that they spread with relatively minor demic impact. The genetic evidence that we have so far still supports this possibility."

It has long been suggested. But that doesn't make it more likely to actually be true. The linguistics can be complicated without all the information.

What ancient genomes have been showing us in the last few years is that major sweeping changes in language across large areas usually have at least some kind of genetic component, even if it is subtle.

I don't think that 'it has long been suggested' is going to cut it in this case.

Ric Hern said...

Polada Culture. Originating from the Unetice.

Al Bundy said...

@Rob I'll look at your earlier posts.I was chiming in too quickly my mother says I do that a lot.

Rob said...

@ Karl _K

" But there were groups who picked up key parts of it and ran with it all over Europe, and by 2600 BC the Bell Beakers in Germany had R1b-M269 and an autosomal composition quite similar to modern North-West Europeans (such as the Irish)."

Agreed, but what Plains W and I are getting at is that this language must have still been something significant more basal to Celtic proper.

@ Ric H

Yes something like that made a big impact in post-Remedello, post -BB Italy; which then spread south later as the disintegrating Apennine culture.

Cossue said...


Galician and Leonese medieval documentation shows the presence of Moor slaves in the NW of Iberia from the 9th till the 13th centuries, either because they were war prisoners brought from the South, or because they were simply bought to slave traders. And they usually admixed with the general population after baptism. Just a few examples (I know tens more):

* Donation of king Alfonso III to the Cathedral of Lugo in 897:
“mancipia ex Hysmaelitarum Terra captiva duximus L, quibus precipimus expleri obsequia Ipsius Sedis”
(‘50 serfs who we brought as prisoners from the land of the Ismaelites, who we order to be given for the service of this See’)

* Donation of king Ordoño II to Santiago de Compostela in 911:
“donamus etiam glorie uestre ex mancipiis quos sca. intercessione uestra de gente hismaelitarum cepimus; nominibus Froilanum, Leodericum cognomento Abdela, Froritum cognomento Abderahamam cum sua muliere Maria et sua filia Guntina, Zahit, Zahim, Scahit, Zahaton, Iausar, Lallus, Fetta, Melchi, Zahit, Aloitus, Fare, Adosinda cognomento Anna, Teodegundia cognomento Anza, Carrataim, Belita, Rahama, Kerita, Aissima cepta cum filia sua. item et alios Zahat, Eikar, Abdel, Gatel, Calaph. item Cahat, Alfarach, Abuzahat, Feta et Alazath”
(‘We give to you, of among the serfs who we captured of the Ismaelites thanks to your intervention, those called Froila, Leoderico alias Abdela, Frorito alias Anderahaman with his wife Maria and his dauhter Guntina, Zahit, Zahim, Scahit, Zahaton, Iausar….’)

* Some serfs are donated in a 1029 charter from the monastery of Cellanova:
“mancipios et mancipellas quos fuerunt ex gentes mahelitarum et agarini, id sunt: Petro, Martino, Domengu, Halephe. item post Alveidar, Maria, Gigenia, Marina, Semza”
(‘serfs and maids who were of the people of the Ismaelites and Agarenes...’)

* Some serfs are donated in a 12th century charter from Pontevedra:
“In Ribas Iº kasal et IIIes mauros et IIIes mauras, totas suas equas bravas et totas meas vaccas” Pontevedra, 12th century”
(‘in Ribas one hamlet with 3 Moor men and 3 Moor women, all the wild mares and all my cows’)

* Some serf are donated in a 1154 charted from Oseira:
“vacciis XXXª et ovibus Cm et ethiopibus XIIIIcim, inter sarracenos et sarracenas”
(‘30 cows, 100 sheep, 14 Ethiopians, both Saracen men and women’)

* A legal charter from 1230 (‘foro do Bo Burgo de Castro Caldelas’), in Galician language:
“Iten se alguun extranyo uender mouro ou moura, de in portagen xij dñ., e se in casa do ospede uender, de a el otros xij dñ.”
(‘Item, if some foreigner would sell a Moor man or a Moor woman, he shall pay 12 coins...’)

* A 13th century ‘cantiga’ (song), by the Galician Pero da Ponte:
“se xe foss'en corredura e podesse prender mouro, tenho que x'o venderia”
(‘If I would go for a ride, and I could take a Moor, I think that I’ll sell him’)

These people must have contributed to the current gene pool, and written history should not be neglected.

Karl_K said...

@Rob

It may not be a popular opinion, but several papers have suggested older dates.

Gray and Atkinson estimated a date of 4100 BC for Proto-Celtic, while Forster and Toth suggested a date of 3200 BC for the arrival of Celtic in Britain.

Rob said...

@ Karl K

I guess it's possible, but that would require Celtic being learned from Neolithic or even Mesolithic women from western Europe ?

truth said...

@ samuel Andrews

Ancient samples from Iberia already score north-african in calculators. I would say that 90% of this presence has nothing to do with the muslim period. In the muslim period, the "moors" were a very small minority, mostly in the elites and the armies, being the vast majority of muslims natives converted, and we can observe the north-african pattern distribution doesn't fit with history, for example Galicia (being only about two decades under Islam) has more NA than Granada (8 centuries under Islam).

Karl_K said...

Good one Rob. The point is that we likely do not have enough evidence of dead languages, or ancient language interactions, to pin this one down very accurately by linguistics alone. There is just not that much evidence to build an accurate tree, or determine where the root would be.

Karl_K said...

It is just as likely that the Continental branch of Celtic is an offshoot of the Insular branch. Or, more likely still, that they just diverged from each other at a slow pace.

truth said...

@ Cossue

having muslim names doesn't mean being ethnically north-african. The native population converted to Islam and changed their names for an arabic one.

truth said...

@ Cossue
Also ther term "moor" used in Spain was to refere someone "muslim" , regardless of race. The Indonesians from Molucas (muslims) were also called "moors" by the Spanish conquistadors.

Cossue said...

For the records:

According to http://www.nature.com/ejhg/journal/v24/n3/full/ejhg2015114a.html, in the Iberian Peninsula:

R1b-u152 peaks in Southern Portugal (13%), Galicia-Asturias (9-8%), and in the East (6%), but it falls to a 2% among the Basques (in the rest it is at some 4%)

R1b-m529 peaks among Galicians (7%), Asturians, and Cantabrians (6%); it falls to a 3% in Portugal, 2% among the Basques and 0-1% in the rest.

R1b-u106 is >5% in N. Portugal, >4% in Galicia, and also in Alicante. It falls to around a 2% in the rest.

R1b-DF27 is “just” 40% in Galicia and Portugal; some 48% in Madrid and Barcelona; and it raises to 70% among native Basques.

Karl_K said...

Thanks Cossue.

Any thoughts on that distribution?

Ric Hern said...

There is a difference between Q-Celtic and P-Celtic. Why did Ireland not become a P-Celtic speaking country ? There were certainly very minimal linguistic influence from the previous Neolithic inhabitants of Ireland as can be seen within the DNA and even the Hallstatt and La Tene did not show massive influence. So why will people totally change their language if there were no real military or economic pressure for them to do so ? We see how the English accomplished it. Massive pressure and absolute dominance accompanied by Population replacement in a substantial area. Why is this same scenario not visible throughout the Iron Age when Celts were said to have spread into Ireland ?

Karl_K said...

@Ric

Exactly. Classification of Celtic languages is a mess historically. Ancient DNA should now guide this field.

Cossue said...

@truth:

Yes, having an Arab or a Berber name and being Muslim didn't imply that these people were purely North African/Arabian, but it certainly increases notably the probability of them having North African/Arabian ancestors. But also the use of terms such as "Ethiopians" is very clear and concise.

Now, if you look at the miniatures of the Cantigas de Santa María, you can see clearly that whilst there were blonde Moors, also there were much darker people, notably among their serfs: https://goo.gl/eLQ2jw. Keep in mind that Arab slave trade moved hundreds of thousand of people during the Middle Ages.

Anyway, I'm not implying that all N African admixture in Iberia is recent, but that we have plenty of written sources that say that there were Muslim slaves in the NW of the Iberian peninsula during the Middle Ages, and that they admixed with the local population and converted to Christianity. So, simply saying "Middle Ages can explain N Afrian admixture in Iberia because it is also present in Galicia which was not really under Muslim control" is a fallacy, because there were known presence of Muslims in Galicia, not as rulers, but as serfs and POWs.

Ryan said...

I'd think this migration would be a much, much better candidate for Lusitanian rather than Celtic languages, no?

Cossue said...

@Karl_K

Well,
U106 = Sueves and Visigoths (that's probably the easy one :-)
R1b-m529 = (Late?) Atlantic Bronze Age, I guess.

But that's it.

Cossue said...

Ok, where I wrote

"So, simply saying "Middle Ages can explain N Afrian admixture in Iberia because it is also present in Galicia which was not really under Muslim control" is a fallacy, because there were known presence of Muslims in Galicia, not as rulers, but as serfs and POW"

I meant
"So, simply saying "Middle Ages can't explain..."

Plains Wanderer said...

@Karl_K

There's no point in throwing out long held ideas if available genetic evidence doesn't contradict them.

U152 is a strong candidate for the spread of Celtic languages in Iberia (and elsewhere).

DF27 is probably older and spread across Iberia during the late Bell Beaker period, hence its presence in pre-Celtic IE and non-IE groups.

Ric Hern said...

Absolutely Karl. I think the reconstruction of Proto-Indo-European is also problematic. Why I say so is because languages that split earlier from the common source usually retain many of the original and remain Conservative for longer than languages staying close to the original area of development. How can a Later Indo-European Language really contribute to the reconstruction of the original Proto-Indo-European ?

Ric Hern said...

What if the Basque peoples R1b were introduced by a small group of traders who did not have the massive influence of a full scale invasion ? Why would those people deem it necessary to change their language within such a scenario ?

postneo said...

@Ric Hern

"because languages that split earlier from the common source usually retain many of the original and remain Conservative for longer than languages staying close to the original area of development. "

You cannot say that earlier/later/peripheral/central branches etc..are "usually" more conservative than the other. This is a subjective statement with no statistical or empirical weight.

Two branches can preserve completely different archaic traits not preserved in the other.
Same as DNA. The only time the mutation clock slows would be when the fielding population stays small for an extended period of time.

AWood said...

If Poland had YDNA turnover during the Neolithic period transition to Corded Ware, why are we assuming this not to be the case for Iberia? You folks realize DF27+ is younger than the European Neolithic right?

andrew said...

@Karl_K

The diversity of the Celtic language family is simply far too low for it to be that old.

Gray and Atkinson's work is quite frankly total crap that has been widely canned in the linguistic community for its failure to accurately model the fact that language change goes faster during periods of language contact and language formation, and slower during periods of isolation. I've read the papers and they consistently miss the mark clear calibration points.

We absolutely know that we are getting major gene pool change in the Bell Beaker period as the Irish example illustrates. But, we do not know what language those people spoke and the linguistic evidence does not support proto-Celtic in the Bell Beaker area. It also ignores Northern Europe (e.g. Balltic Sea coast) where there were Bell Beakers and there are people with classic Bell Beaker genetics, but there was never any Celtic or even Celtic-substrate indications.

Ryan said...

@Ric - "There were certainly very minimal linguistic influence from the previous Neolithic inhabitants of Ireland as can be seen within the DNA and even the Hallstatt and La Tene did not show massive influence."

Irish and Scots Gaelic are vigesimal and that seems like an obvious and significant relic from a previous Neolithic (or perhaps later) group that they (and other insular Celts) encountered. Celtic languages were originally decimal (as PIE was).

I think the fact that Celtic languages don't divide well into a tree structure is good evidence of the influence substrates too, though those substrates need not all be non-IE or even non Celtic. Gaelic replaced Pictish in some parts of Scotland after all, and I doubt that was unique.

Ariel said...

@Adrew

"The diversity of the Celtic language family is simply far too low for it to be that old."

Devil advocate here. There were a lot of languages that have been proposed as pseudo-Celtic/para-Celtic languages like ligurian and Lusitanian, they could be related to an older founder language from the hallstatt culture territory. The main branch of celtic could simply be a dominant version that had a big expansion on a later date.

Ric Hern said...

The Neolithic Admixture seems to have been Continental so where did Celtic get the substrate ?

Davidski said...

The Neolithic Admixture seems to have been Continental...

Not all of it, as you shall see, hopefully soon.

olga said...

@ André de Vasconcelos sais " Why do Galicians, Leonese and Asturians score more NA ancestry than Valencians, when the former were only occupied for decades - or less than that - and the latter up to 1238 (ignoring the few years in which it was controlled by El Cid)? "

History is far more complicated that it looks in the average books, mainly for political reasons.
In Spain there were many migrations from North to South and from South to North after the arrival of the north african invations in 711 AD.There is research about these facts in: The Mozarabs in the kingdom of León: 8th-11th Centuries by Javier Fernández Conde, LA EMIGRACIÓN MOZÁRABE AL REINO DE LEÓN,
SIGLOS IX Y X, Gonzalo Martínez Díez,LOS MOZARABES EN EL NORTE DE ESPAÑA, Joaquín Perez Fernández. The mozárabes were cristians whith arabic culture, that spoke arab and romance, and of course intermingled with the north african invaders.

Samuel Andrews said...

@Andrew,

I trust linguistic experts. But I do want to say just because Bell Beaker existed in the Baltic doesn't mean they left a genetic or linguistic impact on modern people there. The lack of Celtic languages in the Baltic doesn't mean anything.

Also there is a clear correlation in R1b P312, Bell Beaker, and Celtic languages. But there's also a correlation in R1a Z280, Corded ware, and Balto Slavic languages. Maybe Celtic is defended from Bell Beaker but didn't expand till later and maybe the languages they replaced were relatives like Italic, Lusatian, and more

EastPole said...

There is a new paper on Cucuteni-Trypillia mtDNA by Nikitin, Reich et al.:

http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0172952#sec009

Cucuteni-Trypillia mtDNA seems similar to Funnel Beaker, Salzmünde and other European and West Asian/Anatolian Neo-Eneolithic farming populations.

Samuel Andrews said...

It's about time we got Ukrainian farmer DNA. ROmanian Neolithic mtDNA frequencies are quite differnt from Neolithic Spain, Germany, and Hungary. I'll se see how Ukraine Neolithic compares.

Samuel Andrews said...

Can already see one had H1b, a typical version of H1 in modern Danish and White Americans(There isn't a lot of high coverage data aviable). H1b hasn't been found in Neolithic Europe yet.

Samuel Andrews said...

@EastPole,

There aren't enough samples to say Ukraineian farmers had especially similar mtDNA to anyother other Neolithic Europeans. The authors made a premature conclusion. We can confidently say their H1b, H5, U8b1, HV6-24, T2b make them EEF-like and set them apart from Yamnaya and Catacomb. But that's all we can say.

Amanda S said...

Out of interest, Ryan, how do you know that Celtic languages were originally decimal. All the living ones are vigesimal (you can add Welsh to your list). Also French has some vigesimal elements (quarter Vigny for eighty) which I thought it got from Gaulish.

EastPole said...

@Samuel Andrews, they were probably using SNPs to determine similarity, for example they have noted that Maikop has similar lineages but displaying different polymorphism patterns:

“An mtDNA analysis of the Novosvobodnaya and Maikop cultures (3,700–3,000 BCE) from the northern
foothills of the Caucasus mountain range produced mtDNA lineages of T2b and U8b1a2 [30], although displaying different polymorphism patterns compared to the T2b and U8b1a2 lineages reported for TC (this report) and ancient farming groups from central Europe.”

Rob said...

@ Sam

"Also there is a clear correlation in R1b P312, Bell Beaker, and Celtic languages. But there's also a correlation in R1a Z280, Corded ware, and Balto Slavic languages. Maybe Celtic is defended from Bell Beaker but didn't expand till later and maybe the languages they replaced were relatives like Italic, Lusatian, and more"

Exactly. The first P312 groups were probably a sort of generic west European I.E., from which later groups like Celtic, Lusitanian, Italic & Venedic evolved.
Later, Celtic expanded over many of these, but not completely, as Venedic & Lusitanian survived into Roman times.
Moreover, contrary to Karl's suggestions (and I do realise he was trying to be funny), most linguists date Celtic to after 2000 BC for a reason - the Celtic languages are so close and shared vocabulary suggesting of mid-late Bronze Age cultural stage, which means that the earliest P312 arriving in the Atlantic weren't (yet) Celtic, but likely later evolved into Celtic due to contacts with central Europe (e.g. with arriving elites).

Rob said...

@ EastPole

Interesting paper.

" These findings are consistent with the notion that maternal genetic contribution from Mesolithic Europeans was minimal in the TC. Thus, any robust features of craniology, if they are genetic in origin, are more likely to be explained by paternal contribution, which should be visible at the whole genome level of analysis.'

I wonder if they're suggesting that R1 & 'robusticity' began to permeate into CT by Mesolithic Ukrainians males marrying in ?

Rob said...

And

"The frequency of haplogroup H and the presence of H1 in mtDNA lineages found in TC population at Verteba further strengthen the genetic connection between TC and populations of the Beaker cultural complex."

André de Vasconcelos said...

@olga

Most recent migrations were North->South due to people resettling depopulated areas (read: places that belonged to expelled muslim inhabitants).

Yes I am well aware of the Moçarabes, however they were at the very bottom of society in Al-Andaluz, so to say they "of course intermingled with the north african invaders" doesn't sound plausible, or significant at all. If they were female, they would probably convert to Islam, unless they were concubines of a powerful man, regardless their offspring would be raised as muslim. If they were male, what would be the odds they'd get a muslim wife, thus from a higher social standing (which would overwhelmingly be a native Iberian, rather than Berber, let alone Arab)? Very slim. Moçarabes and even Muladi did not live in a dream land, revolts did happen, the most notorious perhaps the one led by Marwan Ibn al-Djilliqui, a Muladi from modern northern Portugal who lived in the south in the IX century. His revolt, supported by the king of Leon, led to the creation of an autonomous - de facto independent - territory until the mid X century. It would form the basis for the Taifa of Badajoz.
Besides, this argument could work on the whole peninsula, not just the west, as Moçarabes existed in all Christian kingdoms, not just Asturias/Leon.


@Cossue

Yes, we've had this talk months ago, however, even if those people leave a genetic legacy (one we could actually see), what about the places these NA men and women were free rather than slaves? Wouldn't those places have much better conditions for their offspring to thrive and expand?

Ryan said...

Amanda - It's actually only some dialects of French that have vigesimal relics actually. In Switzerland or Belgium they would say octante instead of quatre-vingts. Franco-Provençal and Occitan are both purely decimal too (aside from the same switch in number form that English has at twenty - ie nineteen vs twenty-nine). The only inscriptions we have for Gaulish numbers has thirty as triconti - a decimal form - though that is from southern France in an area that was decimal until recently. Gaulish could be the outlier there (and we only have one number in one inscription to go buy), but proto-Indo-European was decimal, so I think it's more likely that the insular Celtic languages acquired this feature well after proto-Celtic split up. IIRC proto-Indo-European has common roots up to 10, and mostly similar ones up to 99. There's a common root for hundred as well, but it meant different amounts in different languages (for example hundred was originally 120 in Germanic) and probably meant something like "many" originally. There's no common root for 1,000.

Ric Hern said...

Some languages change faster than others. How does shared vocabulary point to a later formation date ?

Ric Hern said...

Take Q-Celtic vs. P-Celtic. P-Celtic is postulated as being the oldest of the two. Now we look at the word for horse in PIE and we see it is nearer to Ekwos than to Epos. KW sounds more like Q than P. So now I ask you with bloodshot teary eyes, How can P-Celtic the older form of Celtic ?

Ric Hern said...

Call it Proto-Celtic or Proto-Italo-Celtic if you want but those people in Ireland surely spoke some basal form of Celtic.

Aram said...

U8b1a2 in CT is the same as in Maykop.!
Today it is found as far ss in Algeria.
U8b1a1 is a Iranian thing.

Krefter said...

Post at my blog about the new Cucuteni-Trypillia mtDNA.

http://mtdnaatlas.blogspot.com/2017/02/european-farmer-mtdna-in-cucuteni.html

The H1b1* result is especially interesting. H1b is pretty much an exclusively European lineage today. It peaks in Eastern Europe at as high as 4%. It has been found in Late Neolithic Poland and Mongolia(alongside R1a Z93). It hasn't been found in any remains from Neolithic Hungary, Germany, and Spain.

Hodo Scariti said...

I'm sorry, but doesn't 4000 years before present coincide with the full blown beaker package in Iberia (2200-2000 b.C. cca)?

Ric Hern said...

In Ireland ?

Karl_K said...

@Blasonario

Why are you sorry? You are correct. The only problem is that some people see that stage as the end, because parts of it had been around much earlier.

Ric Hern said...

Are the majority Neolithic Admixture in the Irish not Continental ?

Rob said...
This comment has been removed by the author.
Cossue said...

@André

Yes, of course: they would have contributed to a much larger extent. The problem is that many didn't had a chance, since most of their descendants, hundred of thousands of Muslims, were expelled from Andalusia and from the East of the peninsula being replaced by colonists from the North -also from Galicia-. This replacement already took place in Coimbra, when it was conquered in 878: according to the Asturian chronicles the population of the city was expelled and substituted by Galicians ("Conimbriam ab inimicis possessam eremauit et Gallecis postea populauit"). So the most promising point is South Portugal, were as long as I know Muslims were not expelled. But since S Portugal is simply closer to Africa it is only natural that it shows higher N Africa admixture/affinity.

Again, I'm not denying old N African admixture: I'm simply pointing out that for interpreting modern distribution you must take into account historic evidence. Or at least this was customary in the pre-post-truth world.

Alberto said...

Yes, André, what happened with the Muslim population in Portugal? After religious freedom was banned in Spain and a large part of the Muslims abandoned the country (most to Africa, but some to Portugal, if I'm not mistaken), and the ones who stayed and converted were prosecuted quite aggressively, it also becomes plausible that if Spain lost most of this population, but Portugal didn't (and even received some more), that would contribute to a difference between Spain and Portugal.

But I'm not sure what happened eventually with that population in Portugal. Did it get absorbed into the Christian population? Did they leave at some point?

The fact that Spanish regions adjacent to Portugal also have higher NA admixture could easily be explained by the last 500 years of history, where contacts across the border have been continuous. But this wouldn't affect the populations of Valencia, Aragon, Basque country, etc...

Simon_W said...

@ Ric

"Take Q-Celtic vs. P-Celtic. P-Celtic is postulated as being the oldest of the two. Now we look at the word for horse in PIE and we see it is nearer to Ekwos than to Epos. KW sounds more like Q than P. So now I ask you with bloodshot teary eyes, How can P-Celtic the older form of Celtic ?"

I don't know where you got that from. Q-Celtic is for sure the more archaic version, there is no doubt about this. Q-Celtic, that's basically Gaelic (Irish and Scottish) and Celtiberian. Which indicates that they reached their respective places in Iberia and the Isles before the continental heartland switched to P-Celtic.

Confusing might just seem that the oldest evidence of Celtic, namely from Lepontic, is P-Celtic. But these texts are dated to the 5th century BC, which isn't that early. On the other hand it seems very sensible to derive the Lepontii from the Canegrate culture, which is in the LBA. So at least it looks very much like the north Alpine / East French area in the LBA was already P-Celtic.

Simon_W said...

Slight correction: The older Lepontic inscriptions date back to before the 5th century BC, the item from Castelletto Ticino being dated at the 6th century BC and that from Sesto Calende possibly being from the 7th century BC (Prosdocimi, 1991). (according to wikipedia)

Gioiello said...

Soon you all will realize that Celt came from Italy to Iberia and from there Northward. Lusitanian is clearly linked to the Italic languages and the oldest form of Celt is the "Stele di Novilara", not found amongst the Scots...

Simon_W said...

And next you're going to say Celtic is derived from the Cardium pottery? :D

Gioiello said...

I already said that from so long... and tyhe worst thing for you is that all what I said is becoming true...

Karl_K said...

Wow.

Gioiello said...

Would you be ready to believe that these two haplotypes, of the mainly Sardinian cluster of R-U152-Z192-PF6652, the first from France and the second from Switzerland, with 71 mutations out of 111 markers, are separated only from 4400 years, when also the usual calculation (454 x 71) : 222 x 30 years would give that age, when we know that all that should be multiplied at least for a 2.5 factor?

285324 Boucher BOUCHER, Cathelin (b.~1620, Chaix)>Jean (b.~1650) France R-BY3631
11 24 14 11 11-15 12 12 11 13 12 30 17 9-10 11 11 26 14 19 29 16-16-16-17 10 11 18-21 16 15 20 17 34-43 11 12 11 9 15-16 8 10 11 8 10 10 12 21-23 16 10 12 12 15 8 12 22 20 13 12 11 15 11 11 12 11 34 15 9 16 12 26 26 19 12 11 14 12 10 9 13 12 10 11 12 30 12 14 24 13 10 10 20 15 18 14 24 18 12 15 24 12 23 18 11 14 17 9 12 11
194748 Probst Rudolf Probst 1597 Luzern,Sw>Bartel d.Kandel,Pfalz Switzerland R-PF6652
13 24 14 11 11-15 12 12 11 13 13 29 17 8-10 11 11 21 15 20 29 15-15-15-16 12 10 19-23 15 15 18 18 26-38 12 12 11 9 15-16 8 10 10 8 13 10 12 22-23 16 10 12 11 16 8 12 22 20 14 12 11 13 11 11 12 12 36 15 9 16 12 25 26 19 10 11 12 12 11 9 12 12 10 11 11 30 12 13 24 13 9 10 23 15 18 13 24 16 12 16 25 12 23 18 10 14 17 9 12 11

Karl_K said...

I would be ready to believe that R1b boat men hit Sardinia pretty frequently in the early part of the migration that hit Iberia.

truth said...

Obviously the reason he didn't mention "north-african" it's because :

1\ It's a very small part of spanish ancestry (3-4%) , and
2\ The interview was about the Large and Important ancestriees that makeup iberian (or Western European) genetics (Neolithic, Hunter-gatherers, and Kurgan),

Gioiello said...

Of course the FTDNA Project is wrong and 285324 Boucher is put from YFull tree in an independent subclade of R-U152, R-BY3851, but a little changes about what I said, because the separation between these two subclades is always 4400 years as to YFull.

André de Vasconcelos said...

@Cossue and Alberto

We are in agreement, then. Not sure about the remaining muslim population in the south, I know they were frequently expelled in the early periods in the north (when the county was a vassal of León, or recently independent), but afterwads I really can't say for sure. I know that massacres were not unheard of, for instance in Paderne (Algarve) this did happen, and also that not many buildings survived to today, you'll have a hard time finding old muslim buildings here, the largest might be a church in Mértola, but even that is fairly small when compared to the stuff you have in Spain. Then again, our Reconquista ended in 1249 and the West has been historically poorer than the East, so that might be part of the reason.

Don't think freedom of religion - or lack thereof - was responsible for any potential small differences, when the inquisition was established the prosecution was aimed at Jews rather than Moors, maybe because the latter were (virtually?) non-existant. Whether this was because they fled, died of converted beforehand I do not know, but considering New Christians were persecuted here aswell I'm inclined to discard the latter. I have a book, almost a tome, about the history of Portugal during this period, when I have time I'll try to read it a bit, and I'll have more informed things to say regarding muslims in Portugal after our Reconquista was finished.

Cossue said...

Yep. I was in the Algarve last summer and my impression was that popular architecture -and in general, language and people- were much closer to central or northern Portugal (in a lot of senses, even to my Galicia) than, say, western Andalusia and Old Castille/León. I think that the Castle of Silves is the largest Muslim building in Portugal?

A little off topic, but it is well known that a number of crypto-Jew families established themselves in the raia, the frontier in between Tras-os-Montes, Galicia and Zamora during the 17-18th centuries. You know, because just in case thing went really wrong with Inquisition in one country they could pass the frontier into the other...

Alberto said...

It turns out that actually Portugal was the first one to outlaw Islam, in 1497. I didn't know that. Castile and Aragon followed suit in 1502 and 1526 respectively. Portugal also established its own inquisition in 1536. So things were pretty much the same in both countries.

The inquisition did prosecute mostly crypto-Jews, but that's because they were influential people, often rich and the main loaners to the crown. The crypto-Muslims (Moriscos) on the other hand were mostly a working force who served their Christian lords, so the situation was quite different. But their numbers were pretty big. It's estimated that in Valencia they could reach 33% of the population, and since their expulsion was quite successful when it happened (early 17th century), it lead to an economic collapse.

Anyway, the picture I get is that contrary to what I thought some time ago (I remember I discussed his with Maju back then), the historical (Medieval) events don't seem to correlate well with the genetics, so Maju and André are probably right that the pattern we see is more determined by prehistoric events. But we'll need to get some Portuguese Mesolithic to Chalcolithic DNA to really understand this in better detail.

Simon_W said...

@Gioiello

Which FTDNA project are you talking about? At least the FTDNA R-U152 project has it right.
Indeed, according to http://www.ytree.net/SNPinfoForPerson.php?personID=1561 that Boucher is >Z193>Z40483>etc while the Probst is >Z193>Z192>etc , and that's also indicated in the R-U152 project. But as you say, it doesn't change a lot, because according to yfull the whole R-U152 clade has a TMRCA of 4400 ybp. Now, 71 differences in 111 markers is a lot for sure, but STR markers are not very reliable for TMRCA calculation. SNPs are more reliable. STR markers sometimes vastly underestimate the TMRCA and sometimes overestimate it. By the way I can't see how PF6652 could be a mainly Sardinian cluster as you called it. I see specimens from Germany, Switzerland, Hungary, Ireland and Scotland in the R-U152 project, and a German and two Tuscans on yfull.

truth said...

@ Giollelo

Lepontic must also have split off from the other Celtic languages relatively early, but it is already a P-Celtic language.

Simon_W said...

And as for the idea that Celtic comes from the Cardium Pottery expansion... Most people regard this as outlandish, for some good reasons which have been known for years now:

1. Ancient yDNA. From the Epicardial of Catalonia we have three G2a and one E1b1b. From the equally old El Trocs site which is a bit more inland we have one I2a1 and one R1b-V88, the predominantly African clade of R1b which has nothing to do with the predominantly west European M269. And don't forget Treilles from the Mediterranean coast of France, 3000 BC which yielded no less than twenty G2a and two I2a1. Note how much these haplogroup compositions differ from the composition of typical Celtic people. They are like different worlds. And none of these individuals had a single R1b-M269.

2. Ancient autosomal DNA. The Cardium Pottery people were typical Early Neolithic Europeans. They were rather dissimilar to Insular Celts or the Romanized Gauls, the French, and quite dissimilar to mainland Italians, too - they are most similar to Sardinians. The legacy of EEF in the British Isles is about 40% and about 50% in France. So while there was a time where their autosomal signature expanded - the early Neolithic - they have since then been a lot on the downgrade at the expense of steppe admixed people.

3. The reconstructed Proto-Indo-European vocabulary suggests a Chalcolithic cultural level, nothing early Neolithic or even Mesolithic. And if you look how much the Romance languages have changed since the time of Latin it's impossible that Celtic has started to dissolve 8000 years ago.

Ceterum censeo: North Picene doesn't look Celtic to me.

Simon_W said...

@truth

Yes, like I said, it's highly likely that Lepontic goes back to the Canegrate culture of the LBA, which shows strong evidence of North Alpine / East French infiltration.

Gioiello said...

@ Simon_W

I found those data firstly in the Probst project just searching for 194748, but after, verifying the SNPs, I saw that they were wrong. The huge presence in Sardinia of the Z192 subclades are in the Rocca's tree. Now I am going to my wife for "cena". After I'll study and, probably, answer your posts about linguistics. I read these last days a paper on Lusitanian which seemed to me clearly derived from the Italic languages (what I spoke about also on "Dienekes' Anthropology blog", where perhaps you read me too).

André de Vasconcelos said...

@Cossue

Yes I was excluding military fortifications, all of them have been rebuilt or improved after being taken. The castle in Lisbon, São Jorge, was also originally moorish. But it used to be in ruins up to 1940, so probably it looks nothing like the original.
In the south the biggest differences are the high amounts of whitewash (to reflect sunlight and avoid the torrid heat inside houses) and the building materials since there is no granite down there. And the Algarve chimneys, don't know if you noticed them, but they might be an architectural remnant from the muslim period.
https://www.google.pt/search?q=chamin%C3%A9+algarvia&source=lnms&tbm=isch&sa=X&ved=0ahUKEwiZ0JTu4KvSAhXIwBQKHWuOCc4Q_AUICCgB&biw=1280&bih=675
I avoid it like the plague, I can't stand the heat and I'm not really a beachgoer.

About the Jews, what you say is true but not just in NE Portugal, for instance my girlfriend is from Idanha (old Egitania), near the border with Extremadura on the northern bank of the Tagus river, and she did have a jewish great-grandfather. Most crypo-Jews, however, seem to be from Belmonte, but there's really not many of them.



@Alberto

Yes, but eventhough there is no solid data, it does seem the number of Moors in Portugal was very small. It's a bit hard to track who was of foreign origin or not, because authorities mostly cared about the religion people followed, rather than their ethnic origin.

When the southern reaches of Alentejo and the Algarve were taken, after Navas de Tolosa, most of them were apparently not done by the king, but by military orders and nobles, the most notable of which Dom Paio Peres Correia, Grand-Master of the Order of Santiago, also famous for the conquest of Murcia and Sevilla. Some of the cities, however, were abandoned before the Christian advance, like Elvas. I have no idea how the Order of Santiago treated the muslim population that remained, but somehow I feel most wouldn't be inclined to stay for long. By the way, cities in Portugal were a lot smaller than most people think, for instance by 1477 Braga had a population of under 1800. Here's a graphic representation of the population by 1422AD (based on the number of Besteiros do Conto) http://i1255.photobucket.com/albums/hh621/Sawelios/20170225_174322_zpswxjn6apk.jpg~original

Prof. José Mattoso - possibly the greatest medievalist in Portugal today - says the following (rough translation):
"The Moors, after mid XIV century, declined rapidly. We don't have any solid figures, but documental references regarding their presence progressively disappear. In 1472, when the Cortes talk of Moors, they refer only to the slaves brought from North Africa".
"Alas, by the last decade of the XV century, the religious presecutions will not leave, in respect to the Moors, any significant record. And the opposite would have happened if this minority had preserved the slightest democraphic or socio-religious weight".

So, in conclusion, I don't think the post-Reconquista muslim population in Portugal had any sort of relevance for the modern genetic makeup of Western Iberia. I still think the most significant impact happened before, maybe in the Neolithic/Chalcolithic (there are archaeological findings of NA origin in the Douro basin dated from circa 2000BCE), maybe in the Roman period. But I'm open to any explanation.

Gioiello said...

@ Simon_W
"one R1b-V88, the predominantly African clade of R1b"

Perhaps you are the only one on all the fora not having realized that I definitely demonstrated that R-V88 doesn't come from Middle East or Africa but from the Italian Refugium. My letters (and their answers) to Cruciani and Scozzari date only a little after their paper of 2010. Thus put R-V88, and very likely hg. G, in spite of the hg. G found in Middle East, and E-V13 (not found in aDNA in the Balkans when I am saying from so long that the oldest haplotypes are in Italy), as migrated from Italy to Iberia. They may have carried (with I-M26) the Basque language (Sardinian was very liely a similar language) and Italy hasn't only that but very likely IE and an intermediate language as Etruscan/Rhaetic (this was the theory of Alfredo Trombetti).
But are you a cathedratic likel Gordon? Certainly after that you lost your Jewish scholars (who had at least 51% of my Tuscan ancestry) also Goettingen and the other German universities aren't more those of Schleicher, Brugmann and all the others.
But you should know the great Pole linguist Witzak (congratulation to Davidski) and the paper of Blazek on Lusitanian.

Grey said...

On the languages thing - if a small migrating population merged with a large population of a pre-existing culture in one region then they might adopt the majority language however if the same thing happened in a nearby region that was also part of the pre-existing culture but much less densely populated because of the climate such that the incoming population became the majority then the incomers might keep a modified version of their original language.

(If so then maybe Basque is descended from the Atlantic Megalith culture language?)

Ric Hern said...

You have +-60% R1b is the Spanish Male population which is +-21 Million combined. So about 12 Million Males are R1b of which the Basque only make up about 1.5 Mil. So you are left with 10.5 Mil that speak Indo-Europeañ all around the Basque. The Spanish are also heavy with R1b DF27. So what does this facts basically tell us ?

Karl_K said...

The Basque show us that Indo-European speakers have substantially more Yamnaya-like autosomal ancestry than non-Indo-European speakers. Y haplogroup frequency is not always a good marker for language. The real question is did the Basque ancestors acquire their R1b at the same time as the surrounding populations, or in a seperate independent event.

Davidski said...

There were four major population shifts in Iberia.

1) from Mesolithic foragers to farmers

2) from farmers to early Indo-Europeans

3) from early Indo-Europeans to late Indo-Europeans (Romans)

4) from late Indo-Europeans to modern Iberians with some North African admix

Basques differ from other Iberians in that their ancestors experienced 1 and 2, but managed to keep their language, and then they were largely spared 3 and 4.

So they have about as much Yamnaya ancestry as other Iberians, or even more than most, because they were largely spared the more southerly Roman and North African influences.

Grey said...

Ric Hern

"So what does this facts basically tell us ?"

DF27 expanded first and then got rolled over by a later second wave?

(just suggesting that as a possibility)

the key thing i think is Iberia was full of shiny metal and people will travel a long way and fight a lot over shiny metal

truth said...

The Roman influences are small actually. Most R1b in Italy is U152, which is low in Iberia, unless roman influence was female-mediated, wich doesn't seem plausible.

truth said...

If we take for example Andalucia (one of the most Romanized regions of Spain) , then we can see very low levels of R-U152 :

Myres et al. 2010 : 0.8% (1/127)
Valverde et al. 2015 : 4.0% (4/100)
That gives a total of about 2.20% R-U152 out of 60% R1b for Andalucia.






Josep Coderch said...

Hi, I'd like to know if it's possible that there was a migration wave into the Iberian peninsula of R1b males who didn't speak an indo-european language.
I'm catalan myself and from what I've seen we have very high R1b ydna (around 80% I recall) but we didn't start speaking an indo-euro language until the romans came. The iberian language seems to be unclassified, but my doubt is whether its origins are from the local neolithic people or the later R1b settlers brought it.

Simon_W said...


@Gioiello

About R1b-V88, relax! By calling it „predominantly African“ I wasn't saying anything about its origin, I was just referring to the place where it's most common now. Just like I did with calling R1b-M269 „predominantly West European“. Of course I don't believe that it originated there.

And no, I'm not dogmatic, I just try to follow the evidence where it leads me. Whereas you have made up your mind ten years ago and now you're never going to change your opinion again, no matter what new evidence comes to light. That's an awkward attitute which seriously hampers the ability of learning.

And no, I'm not a German, so I didn't „lose“ my Jewish scholars, and anyway I don't share your obsession about Jewish scholars. It's completely off topic. I don't agree with you that Ashkenazi Jews are 51% Tuscan or Italian, but that's also off topic.

About Lusitanian, afaik it seems to belong to a pre-Celtic IE layer in western Europe which is closely related to Celtic and Italic. Hence I don't deny any similarities with Italic, but it would be very premature to explain this with an origin in Italy.

I'd like to come back to your point about the language of the Novilara stele. In fact there are on the whole four pieces of epigraphic evidence of the so called North Picene language. The long text of the Novilara stele and three short fragments. To me it looks like just one short fragment is real, and the rest, including the famous, beautiful Novialara stele are probably forgeries, hence it's not in any real language, it's a nonsensical jabbering with some Greek, Etruscan and Italic influences. The most convincing argument for this opinion: The script has a sign in shape of an R, and since the sign that looks like a D probably represents the phoneme r, it's assumed that the R represents a d – just like centuries later in the Oscan script. But that would be a strange coincidence. In general it can be said that the R sign was developped rather late, in the Oscan script as a sign for d and in the Latin script for the r phoneme like we still use it. What did occur early on, already in some Greek scripts, was a P with a short downward stroke, as a variant of rho, but a complete R that's an anachronism. And if it's not meant to signify rho it's a double anachronism. There are other aspects of the Novilara stele which are anachronistic – the interpunctuation for example, but the thing about the R convinced me most. It has also been noted that on the backside of the stele the helmet looks rather unusual for this time and place, it looks like Hittite-derived. And in fact, the stele and two of the fragments were not found in scientific excavations, but under unsupervised circumstances by private collectors. And the only fragment that was found in an archaeological excavation has a linguistically different character, with similarities to Etruscan.

Gioiello said...

@ Simon_W

I was expecting some answer of yours from the time of "Dienekes' Anthropology blog" about some inscription on Crete of probable Italic origin... but much water passed.
1) R-V88, R1b1-L389+, R-M335 surely come from the Italian Refugium, for that I didn't change my position, because it has been demonstrated true.
2) That R-M269 didn't come from Western Europe... be prudent. About a wrong presupposition that R1b1 in an aDNA in Iberia of more than 5000 years ago that Richard Rocca had against me and Genetiker he arrived to announce his dimissions from the "Italy FTDNA project", after having been demonstrated an ass in genetics, and read what I am writing from ten years about that.
3) That Ashkenazic Jews are 51% Tuscan was said just from Davidski, and I said that his about 30% Samaritan is overestimated for a 10/20%.
4) During my life I did other, but I'd be able to put hands also in facts of linguistics. I don't know if you know a paper on "La parola del passato" of a Dutch linguist about the Stele of Novilara. It seemed to me not less reliable than the analyses of Witczak and Blazek about Lusitanian.
5) But you are right. Not "ciacole", but papers.

Ric Hern said...

What I'm trying to say is that percentages can be misleading. Although the Basque can be 98% R1b DF27,they are only +-1,5 Mil Males while in the rest of Spain the Percentage could be 30% R1b DF27 but they would then be 7 Million Males. This could make the amount of R1b DF27 Men outside Basque Territory almost 5 times more. Does this then really point toward a replacement of R1b DF27 within Iberia ?

Ric Hern said...

Maybe their percentage from the beginning never exceeded 30% on average throughout the combined territories.

Ric Hern said...

This also counts for the Language spoken by R1b DF27. If 7 Million R1b DF27 Men speaks Indo European and only 2 Million R1b DF27 Men speak a Non-Indo-European Language,what is the chances of Indo-European not being the original language of R1b DF27 people ?

Gioiello said...

@ Simon_W
This is my last analysis of the situation of the hg. R-M269-PF7562, not different from what I said ten years ago, that I published in some blogs and fora. If you have something better, you may post that here. About the Novilara stele... I don't agree with you. One thing is to forge "Manios med phephaked Numasioi" and another is the test of the stele. The fact is that tens of languages disappeared amongst the European languages before they could have been witnessed. If I were you, I'd begin to think that also Villabruna is a forgery... otherwise you are inevitably defeated by me.

"Of course that the "basal branches" of R-M269 are in the Balkans isn't true.
1) It is true that so far there is a sample of R-M269-PF7562+ and PF7563- found in Anatolia, but this is just a tiny subclade which isn't the ancestor of all the others, as nobody would think that R1b1-L389- found in India or Asia is the ancestor of all us who are R-L389+ (now the thing is more and more clear with the tests on R-V88 and R-M335 all rooted in Italy).
2) The age of this R-PF7562+/PF7563-, which was 5600 ya against the subclades given at 4500 ya, is now reduced (after that I invited my Italian American friend Joe Merante to send his BAM file toYFull) at 300 years, being the subclades at 5300 years old.
3) Anyway these subclades are younger than others, not only the oldest R-V88, R-M335 etc all rooted in Italy, but also R-M73 which is older in Western Europe than in Eastern Europe/Asia, where there is the subclade R-M73-M478, and also my R-L23-Z2110 seems older than the others at the R-L23 level.
4) Beyond that, R-M269 has the hugest subclade (Z29758) deeply rooted in Italy and Western Europe and it is clear now that the other samples found in Lebanon and around the Black Sea derive from Italy. The Jewish subclade (A11710) seems young, with 19 SNPs after a bottleneck, and may be derived from Italy or Western Europe as the most part of the Ashkenazic and Sephardic subclades. The other single clade are pretty all from Italy and Western Europe.
5) If other data will change this situation, I'll be glad to take it into account".

Grey said...

Ric Hern

"This also counts for the Language spoken by R1b DF27. If 7 Million R1b DF27 Men speaks Indo European and only 2 Million R1b DF27 Men speak a Non-Indo-European Language,what is the chances of Indo-European not being the original language of R1b DF27 people ?"

good point

#

Josep Coderch said...
"Hi, I'd like to know if it's possible that there was a migration wave into the Iberian peninsula of R1b males who didn't speak an indo-european language."

i guess that's another possibility that shouldn't be overlooked

Ric Hern said...

Interesting. What this for me hints at is that some forms of Proto-Italic or Italo-Celtic was spoken in Iberia and it did not take much to Latinize the inhabitants.Isn't it interesting that the Romans focused their military efforts mostly within areas where Celtic were spoken. I think the close relatedness of the languages certainly helped withe the subjugation process. Spies could learn the languages much easier etc...

Gioiello said...

@ Ric Hern

If Latins would have been able to link Lusitanian "MVITIEAS" with "mutuus" not any more pronounced *moiteuus, they would have been the first Indoeuropeanists, but certainly other words were more similar, above all for the Osco-Umbrian people.

Ric Hern said...

Maybe the actual speaking of the Languages fell better on the ear than the spelling fell on the eye. Heheheeeh.

Ric Hern said...

Yes Q-Celtic and Q-Italic vs. P-Celtic and P-Italic.....

Ryan said...

@Ric - "This also counts for the Language spoken by R1b DF27. If 7 Million R1b DF27 Men speaks Indo European and only 2 Million R1b DF27 Men speak a Non-Indo-European Language,what is the chances of Indo-European not being the original language of R1b DF27 people ?"

That doesn't have much to do with anything. I speak English and French, yet go back 10 generations and none of them spoke either. Language shifts happen.

Ric Hern said...

True. Okay.Marija Gimbutas mentioned the Bell Beaker looks like an Amalgam of Yamna and Vucedol Cultures. Could DF27 have adopted the Non-Indo-European Language of the Vucedol and taken a Maritime Route via the Adriatic to Iberia while R1b L21 retained the more Yamna like character and migrated over land via Germany towards the West ?

Ric Hern said...

Basically two Bell Beaker waves with one reaching Iberia quicker and with a Non-Indo-European language ?

Ric Hern said...

A Bilingual Culture ?

truth said...

@ Ric Hern

How do you explain then the 20% R1b-L21 is some basque areas ?

Ric Hern said...

How do you explain the Gallaecian and Lusatanian Languages ?

Ric Hern said...

Traders and Miners from the North. What is 20% of 1.5 Million within the Basque area ? It is only 300 000 Men.

alobrix said...

If the population that brought 30% of steppe genes was 50-60% Yamnaya, then there was a significant turn-over in all the Bronze Age iberian populations. A predominant new population with a new language and culture.

truth said...

@Ric

The best explanation for Gallaecia is that the area used to be Lusitanian-speaking and was subsequently Celticized, resulting in the strange mixed Celtic-Lusitanian character of Gallaecia. One very interesting aspect here is also that the Gallaecians continued to worship Lusitanian gods.

As for Lusitanian, my interpretation is they were the first wave of Indo-European people to arrive on the Iberian penninsula

Ric Hern said...

Are the Celtic admixture as you mentioned closer to Q-Celtic or P-Celtic ?

Olympus Mons said...

@truth...
"Everybody" knows that celtics were the only ones allowed to enter the TURDULI tribes (lusitanians) and inhabit their territory.
Its known how the bardilli (Turduli of the sado river) roamed freely with celtici into northern portugal until Galiza.

Why were they allowed? Because were seen as the same stock as the locals.