Genetic study A genetic history of the Balkans from Roman frontier to Slavic migrations

Lazaridis commented about this map posted on twitter by a random account,it's not from the study itself.
I see. I misstook that tweet for this one:


From what I understood from the supplement, Table 8 in mmc2 file, is what Olade and co used to base their whole thesis/analysis on.
The point is still very much valid, Lazaridis name is on a paper whose methodology he disproves of.

So he goes from this:
No, as there are 7 different sources in Table 8 and it's usually not possible to estimate ancestry simultaneously from many sources so smaller subsets are considered. E.g., Table 8 shows that only Albanians have ancestry from Albania_BA_IA which is surely not the case.

To this:



I mean, the whole methodology of the paper hangs on table 8, what do you mean by:
Unfortunately, there is a tendency of people to read what they want to read into models, so what can one do.
When the ones reading it quite literally as they want are your co-authors.

Either way, still think Lazaridis is the greatest to do it. But this is very confusing.
 

Attachments

  • lazaridis.png
    lazaridis.png
    100.8 KB · Views: 233
The Slavic input isnt as crazy as the Anatolian input that is being claimed. With the numbers provided in this study it would imply some of these ydna came from Anatolia during Roman era - j2b l283, e-v13, r-z2103, r-pf7563, i1 etc.

J2b l283 definitely didn't, i1 definitely didn't, so now we need more iron age and early middle age Anatolian samples to determine z2103, v13, pf7563
 
@Ghurier Statistical evidence clearly refutes your non eviential claims.
That's quite the oposite in fact. I provided the related arguments based on public data (you failed to comment).
I'm waiting for more arguments from your side to consider your position, that seems ideoligically motivated.
Ignoring arguments didn't make them disappear.

You haven't answered my question with regards to which archeological complex should have yielded a major Slavic patrilineage in the Paleo-Balkans. It is a paradoxon of course.
There is no paleo-Y3120 at all ... whatever you mean by "paleo" ... Reasoning by lack of sample just didn't works with the current poor coverage we have for ancient DNA. Particularly when tracking isolated branch pre-diversification. Finding a ~Y3120 pre-diversification would be fairly lucky (keep in mind that by this time some haplogorups were crossing Europe in few centuries).

Let take another exemple :
The same way that Balkans never yields BCE Z631 samples (you still might expect some from ~400 BCE on in the south-western Balkans), however if I remember well, you are a strong believer that Z631 cradle is around Albania ? Another claim that is totally un-substancieated from a statistical point of view.

As you see, the elements on which you base your conclusions seems to vary from case to case. This is a strong indication that your positions are ideologically motivated ... mine are not, they always follow robust (calibrated on known cases) and objective statistical probes.

As for branch reading and its interpretation, you showcase a lack of understanding for both but this was already pointed out to you by various members before and on multiple occasions.

The point is that you are blinded by ideology. I'm not surprise to see that you are not commenting the statistical elements I'm bringing (Maybe you don't understand them).

I think, that as all the Albanian members you mention, you only look at ancient samples without look-back, making strong claims even when you are thousands of years after TMRCAs.
Small remainder : science is not a democratic process, but an argument based process ... you can be 100 ideologues against my statistical proxies, they will remain more interesting scientifically speaking than your ideology.
When you'll be able to challenge the elements I bring, I might consider your opinion. Yet, it is clear that you have a very limited understanding of basic statistical probes.

As we observe, I-Y3120 population that source Greece and I-Y3120 population that sourced ~Poland decoupled around 300 CE.
This is what we observe. It is not a question of interpretation this is an observation.
Whereas the I-Y3120 population around Poland and the main Balkanic one Decoupled around 600 CE.
Handle it !

By construction to have clade decoupling you need either :
--> the population to stop having overlapping terrtories == migration.
--> the population practicing some sexual segregation between sub-groups.

You also seem to have missed that the earliest sampled individual carrying I-Y3120>Y18331, which I remember was some sort of pet theory material used by you in the past for some ancient Slavic homeland in the Southern Balkans, has a clean Slavic auDNA profile. I mean one can, too, light out the modern basal Polish sample.

Apparently you failed to read me correctly. Repeated strawman argument is not really helping you to get credence.
If you attack a position that no one is defending, you are just fighting wind.
As of now, my claims regarding Y3120 :
--> pre-200 BCE ? I don't know the location, and I challenge anyone to determine the correct location. For timing consideration, a location in north western Balkans, central Europe or central eastern Europe are not out of the picture.
--> 200 BCE ? Likely located around Southern Poland.
--> Post 200 BCE, individual clades with different histories involved in different movements, the statistically most obvious being around ~300-400 CE and 600 CE.

An origin in Southern Balkans ? I never claimed that. I mentioned a potential origin in North-western Balkans, among others due to interesting timing of the diversification (with the Romans attacks being the potential triger of population movements). And even in that setting, this clade would have been pushed toward Poland before expanding around Poland.
Other setting do work, for instance an original Central European setting.
Again, you appear to be blinded by ideology, and you fight something that no-one ever claimed.

This is called bad ethic and a strong bias as to what you wish the results to be. Every STEM major has to become familiar with the scientific method at the beginning of their undergraduate degree.

Interesting projection of yourself.
I think I have probably more peer reviewed paper than you, have you even a single one ?
Thus, your opinion about scientific methodology seems like a weak authority argument attempt.

Again, my position is based always on the same statistical probe, independantly of my opinions.
Your methodology seems to fluctuate strongly depending on the concerned haplogroup. Particularly when we get close from your country ... this is a strong indication that you are indeed blinded by ideology.

Your total silence about the statistical probe I kindly provided also speak volume about you real understanding of such statistical analysis.




IBD sharing is what should be used and not calculators made by armchair scientists.
IBD sharing is close from what is done when doing a segment by segment analysis, However IBD sharing might be too restrictive and require huge database (even for modern sample, using IBD you can't map your whole DNA).
I repeat for the third time : PCA based diagnostic are heavily biased and strongly affected by over fitting.
Segemnt based analyses barely have more than ~400 years depth ... after you enter into noise dominated stuff.

You have these people that create models and give people 30% Roman Imperial but where is all the mtdna and ydna to back up such admixture or even IBD sharing for that matter ?
The point is that Y-DNA and mt-DNA can be transmitted without significant autosomal signal, and some autosomal signal can be transmited without Y-DNA and mt-DNA injection.
A good exemple are Neanderthals having their Y-DNA and mtDNA being replaced by Sapiens like markers, while their global admixture didn't significantly changed.
Admixture is what tells you the origin of an individual, not Y-DNA or mtDNA that only concern a very limited subset of ancestors (for static population they will tell you the same picture).

Again, I'll take J-Z585 lineages found in Nuragic Sardinia.
These individuals have all or nearly all their ancestor coming from Neolithic Europe.
However, he have Y-DNA J-Z585, which is most likely a Chalcolithic arrival or as many Albanian nationalists would like some steppic stuff (possible, but far from a given).
You have here the perfect exemple of how Y-DNA and origin can be orthonal stuff. Because a paternal lineage few generations post migration is a very limited proxy of your origins.

Nowhere did I say we didn't mix with Slavs, what I am saying is that I don't have such lineages within my families, at least not paternally which is actually the most important thing and consistent with the language and culture of a people since all my lineages are literally native Balkan and Albanian is believed to be native Balkan which makes your whole argument nonsensical regarding some J-L283 back in 2500 BCE.

I quote you :
We didn't mix or assimilate Slavs or Gypsies or any of that other trash that came along there .

Also, you claim that "paternally" is the most important ? Why so ? What is making it so "important" to you ?
I suspect it is emotional reasons, because DNA-wise it is nothing special at all.

Apparently, as many you seem believe that Albanian culture and language derive from J-L283 culture and language.
Let look at the numbers.
If we use YFULL v11.05 and if we look a ~0 CE men we surviving descendent among modern Albanian:
-For Z638, we find 14 men with Albanian descendent out of 137 men (roughly 10%) !!! This is not a lot (and some of them are later re-injection ... but let not enter the details).
-For I-Y3120, we find 5 carrier with descendents in Albania out of 15 men in 0 CE (33% of them).
-For E-V13, it is 34 men from 0 CE that have modern dscendent in Albania (out of 426 men ~8%).
-For R-Z2103, it 8 men out of 405 (~5%)
-For R-PF7562, it is 9 men out of 40 (~24%)

Thus, when you look at it, E-V13 men from 0 CE that are at the root of modern Albanian population, are nearly as numerous as J-L283, I-Y3120, R-Z2103, and R-PF7562 men from 0 CE combined.
Also, most of Albanian founder-effects are posterior to slavic invasion and even Muslim invasion.

Thus, who brought the culture ? Who brought the language ? Good question, the two answear are not nessessarly the same.
This is from a guy with mainly a Bell-beaker like DNA profile, living a latin-language speaking country, that derives its name from a Germanic tribe.
Conflating DNA/Culture/Language is just very naïve. We could list the exotic cultural elements that people living in Albania have adopted, there is quite a lot.

All your lineage are native from the Balkans ? What does that even mean ? When did you set the cursor ? All Human lineage are of African origin to my knowledge.
When and why did you decide to put a cursor to determinate a geographical origin ? For exemple, why not starting at J-M241 level ?
Sounds like at ethno-nationalist obsession to me, but maybe I'm wrong.

Mtdna can tell us a lot too by testing for various sides and I am sure I am Albanian on those sides too. Why shoud I as an Albanian not identity with my lineages ? I don't care about Slavs or Anatolians that were assimilated or that we absorbed, but at the end of the day , they are Albanians too today.

Should you identify yourself to your lineages (Y and mt) ? This is an interesting question.
But if you consider all the line that mixed to "create" yourself ... this incorporates some Slavic lines.
mtDNA and Y are just odities of DNA transmission that make some mutations having a coherent transmission patern accross generations. This is what makes it relevant for population genetic analyses.
But regarding yourself it is a negligible portion of your origin once you got back few centuries in the past.

For instance, why identifying yourself more with your Y-DNA than the mutations that define your skin color ? your blood-type ? your hair/eyes color ? ...
Why identifying yourself to some mutations and not to all the mutations that are making you "you" ?
This is what I never got in this "haplogroup" obsession (unless going for very sexists approach of the genetic origins).

So your whole argument here is that the Albanian language is not related to people who were J2b2 and E-V13 or at least R1b despite overwhelming majority of Albos belonging to these lineages ?

I don't know personnally, and I suspect no-one knows.
J-L283 is attested in Cetina (unknown language), in Maros (likely IE, BB-related language), in Myceanean (IE-language), in Sardinia (non-IE-language), in Caucasus (unknown language).
What is the language that the J-L283 MRCA was speaking ? I don't know, but at least some of the descendents changed of language and culture.
Diversity cluster in Sardinia is around 3100 BCE or 3500 BCE, diversity for Cetina is around 2500 BCE, Caucasus lineage seems quite lonely.
I don't see any specific reason to think that J-Z597 (~2500 BCE) emerged from a proto-Albanian speaking population (it could be, or not).

Who were the original speaker of proto-Albanian ? That's really an open question. Were these speaker mono-haplogroup ? Likely not ... even in upper paleolithic mono-haplogroup models are just to simplistic, thus by Iron age/Middle age it is quite riddiculous to summarize a culture with one haplogroup.
Yet, in BA we don't find J-L283 significantly next to E-V13 or R1b, therefore it seems unlikely for these population to have spoken the same language, which one brought the language ? I don't know.

How can these calculators be taken serious ? And how on earth do they supposedly know what admixture occurred or what populations to use ? Doesn't make any sense. It's just a bunch of guess work.
Depends, when your populations are strongly different, a PCA approach can give reasonable results (but you need to make sanity check to prevent overfitting). But also, if by lack of sampling you miss some "source populations" in your model, it will bias your results.
It is when you try to apply it to very fine populations that it starts to fail and significantly over-fit.
DNA transmission is nearly a linear process (in average it is).

Then, the correct admixture should "work" in a PCA space, but inverting PCA can yield to multiple possibilities.
Identifying the good one might be tricky.

That's why I always advise people to check their model on a PCA, but not try to invert admixture from a PCA (unless you have already strong constraints and you are just looking for ~1/2 very exotic components).
 
@Ghurier Ma cousine, you're saying a lot without actually saying anything. I won't dwell into your enumeration of pseudoarguments as there is no purpose in doing that. Nor the ad hominems. They are without substance. Besides, the thread shouldn't be derailed another time.

FYI: I have never claimed J2b-L283>Z631 to have originated in Albania, I rather think it emerged around lesser Pannonia, mod. day Bosnia, Western Moesia and North of the Labeates.
 
@Ghurier Ma cousine, you're saying a lot without actually saying anything. I won't dwell into your enumeration of pseudoarguments as there is no purpose in doing that. Nor the ad hominems. They are without substance. Besides, the thread shouldn't be derailed another time.

Cousine ? Nice attempt to speak French, but it looks like a fail.

You admit that you are unable to challenge my argument about subclade decoupling ?
Thank you to acknoledge this incapacity.

I have never claimed J2b-L283>Z631 to have originated in Albania, I rather think it emerged around lesser Pannonia, mod. day Bosnia, Western Moesia and North of the Labeates.

Just for me to have fun, how do you generate RMPR116 with his autosomal profile under this setting ?
Reply me by PM to not deviate this topic.
 
Cousine ? Nice attempt to speak French, but it looks like a fail.

You admit that you are unable to challenge my argument about subclade decoupling ?
Thank you to acknoledge this incapacity.

Just for me to have fun, how do you generate RMPR116 with his autosomal profile under this setting ?
Reply me by PM to not deviate this topic
I could comment on your non existent English skills too 😆

That is a good idea. I'll think about sending you a DM later but will probably mysteriously forget about it 😁
 
Does anyone have the list of Slavic samples in the Balkans from Southern Arc and this paper? I know only of a few, including some of the Kuline, some of the Doclea etc. The issue is that they are labeled together in the Reich dataset which is not ideal for qpADM runs. This means that the label Doclea includes pre slavic samples, and if you run Doclea like that without changing the labels its of no use. Same for Kuline.

If someone can create a G25 sheet, I can appropriately label the Reich dataset and run qpADM on them.
 
So please tell me Einstein, how on earth is is possible that all my lineages from all sides are native Balkan ? Even from my fathers mothers side they are J-L283 . We didn't mix or assimilate Slavs or Gypsies or any of that other trash that came along there . We might of absorbed such autosomal DNA from other Albanians or taken some brides from some sides but nah we didn't mix with that shit. And how on earth is such a thing not more important in tracking your ancestors than some calculators made by armchair scientists ? And how is it not consistent with the language and culture that people have and speak ?
Have you done an autosomal DNA Test?
 
Are the G25 coordinates available yet?

Specifically trying to see if these are early Slavic samples in the Balkans.
35010​
Croatia_Gornji​
1095368​
F​
1000-1200 CE​
I35095​
Croatia_Gornji​
1092373​
F​
1000-1200 CE​
I28390​
Croatia_Nustar​
832362​
F​
771-947 cal CE (1175±15 BP, PSUAMS-12038)​
I15742​
Croatia_TrogirMatijeGupca​
792830​
F​
661-775 cal CE (1311±16 BP, MAMS-38382)​
I2530​
Macedonia_Bitola​
676563​
F​
992-1121 cal CE (1000±15 BP, UCIAMS-186345)​
I15540​
Serbia_TimacumKuline​
607920​
F​
775-978 cal CE (1140±15 BP, PSUAMS-9670)​
I15543​
Serbia_TimacumKuline​
589616​
M​
800-1000 CE​
I15542​
Serbia_TimacumKuline​
539674​
M​
897-1021 cal CE (1075±15 BP, PSUAMS-8555)​
I15537​
Serbia_TimacumKuline​
422691​
M​
800-1000 CE​
I15541​
Serbia_TimacumKuline​
409930​
M​
800-1000 CE​
I15743​
Croatia_TrogirMatijeGupca​
818204​
F​
774-979 cal CE (1141±17 BP, MAMS-38381)​
I5404​
Serbia_Lepenski​
374622​
M​
1432-1611 cal CE (421±23 BP, OxA-25212)​
I17980​
Bulgaria_Ryahovets​
739203​
F​
1051-1217 cal CE (895±15 BP, PSUAMS-8331)​
I26766​
Croatia_Sipar​
899431​
M​
650-800 CE​
I26767​
Croatia_Sipar​
805277​
M​
500-700 CE​
I28394​
Croatia_Nustar​
827610​
F​
700-900 CE​
I28393​
Croatia_Nustar​
823071​
F​
700-900 CE​
I28400​
Croatia_Nustar​
844068​
M​
700-900 CE​
I28398​
Croatia_Nustar​
841085​
F​
700-900 CE​
I28388​
Croatia_Nustar​
830480​
M​
700-900 CE​
I28397​
Croatia_Nustar​
824487​
M​
700-900 CE​
I35008​
Croatia_Gornji​
1095761​
M​
1650-1750 CE​
I32299​
Serbia_ViminaciumSvetinja​
481766​
M​
1100-1300 CE​
I32301​
Serbia_ViminaciumRudine​
87750​
F​
1100-1300 CE​
 
Have you done an autosomal DNA Test?

The Slavic distribution between Albanians is not uniform some Have more some have none. Normally the low lands have more.
 

Attachments

  • 9699C65A-D910-4CEF-BCDD-5435218D95D6.jpeg
    9699C65A-D910-4CEF-BCDD-5435218D95D6.jpeg
    316.8 KB · Views: 210
What about the 43% Anatolian, where is it more and less?

That one is not very clear to me because it is not just Anatolian for Albania

The label is CroatiaSerbia_RomanAnatolian, not just Roman Anatolian.

I am not sure how this is calculated, also Lazaridis one of the Authors does not know. Hahahah. But if we assume that what is said in the other study is correct than this does not seem very dramatic.
 

Attachments

  • 98967098-C2E0-437B-AC0C-9EB0D8159750.png
    98967098-C2E0-437B-AC0C-9EB0D8159750.png
    1.2 MB · Views: 205
That one is not very clear to me because it is not just Anatolian for Albania

The label is CroatiaSerbia_RomanAnatolian, not just Roman Anatolian.

I am not sure how this is calculated, also Lazaridis one of the Authors does not know. Hahahah. But if we assume that what is said in the other study is correct than this does not seem very dramatic.

CroatiaSerbia is just where the sample was found, wasn't the sample fully immigrant? How much was it Anatolian and Illyrian?
 
CroatiaSerbia is just where the sample was found, wasn't the sample fully immigrant? How much was it Anatolian and Illyrian?


Male lines do not show Anatolian lines in these -% so if true, it seems that the proto-Albanians took Anatolian women, can’t imagine why this might have happened, maybe slaves not sure?

Or For some people fiku zi eshte me I mire se i Bardhi :)
 
Male lines do not show Anatolian lines in these -% so if true, it seems that the proto-Albanians took Anatolian women, can’t imagine why this might have happened, maybe slaves not sure?

Or For some people fiku zi eshte me I mire se i Bardhi :)

I mean I'm asking if it was because I don't known how much anatolian or how much illyrian it was. I don't think 43% Anatolian input in Albanians is possible, this study doesn't make much sense
 
Last edited:
From this study: https://www.biorxiv.org/content/10.1101/2023.06.05.543790v1.full

"We have shown that modern Albanians from Tirana derive 25-48% of their ancestry from a South-Slavic-related source"

"the ancestry of South Slavs as 55-70% Balto-Slavic-related, with the remainder originating from the pre-Slavic inhabitants of the Balkans

"Although models using the unadmixed Balto-Slavic-related proxy (Russia IA Ingria) receive higher statistical support, they are less likely historically. . . It is therefore likely that Slavic admixture entered the Albanian population via an ancestry profile related to modern South Slavs, rather than unadmixed Balto-Slavic populations."

____________________________________________________________________________________________

If we take the average of 55-70, 62.5%.
62.5% x 25-48, gives a rough average of 15.6% to 30% purely Slavic ancestry in these Tirana Albanian samples, and as low as 13.75% to as high as 33.6% depending on how high the Balto-Slavic and/or how South Slavic shifted the Albanian samples are.

Now... this study shows 31% Slavic 'average' in Albanians. Why are there such large discrepancies between all these models...? How is it possible for Albanians and Greeks to be THAT Slavic shifted?

My question (as a complete amatuer):

1) Is it possible these genetic calculators, including Vahaduo/G25 ones, all prioritize a "better" statistical fit by using overly Southern and Northern sources (rather than genetically intermediate), thereby conflating Slavic ancestry higher than it actually is?
Or is the science solid and amateurs just don't get it?

2) If not, then it seems most of the Slavic ancestry in Albanians came from Slavic women, since Slavic paternal lineages are so low... How did that happen? A war that Slavs lost, and Albanians married the remaining Slavic women? Or what other theory?

Sorry if this is amatuer-tier. Just trying to learn.
 
From this study: https://www.biorxiv.org/content/10.1101/2023.06.05.543790v1.full

"We have shown that modern Albanians from Tirana derive 25-48% of their ancestry from a South-Slavic-related source"

"the ancestry of South Slavs as 55-70% Balto-Slavic-related, with the remainder originating from the pre-Slavic inhabitants of the Balkans

"Although models using the unadmixed Balto-Slavic-related proxy (Russia IA Ingria) receive higher statistical support, they are less likely historically. . . It is therefore likely that Slavic admixture entered the Albanian population via an ancestry profile related to modern South Slavs, rather than unadmixed Balto-Slavic populations."

____________________________________________________________________________________________

If we take the average of 55-70, 62.5%.
62.5% x 25-48, gives a rough average of 15.6% to 30% purely Slavic ancestry in these Tirana Albanian samples, and as low as 13.75% to as high as 33.6% depending on how high the Balto-Slavic and/or how South Slavic shifted the Albanian samples are.

Now... this study shows 31% Slavic 'average' in Albanians. Why are there such large discrepancies between all these models...? How is it possible for Albanians and Greeks to be THAT Slavic shifted?

My question (as a complete amatuer):

1) Is it possible these genetic calculators, including Vahaduo/G25 ones, all prioritize a "better" statistical fit by using overly Southern and Northern sources (rather than genetically intermediate), thereby conflating Slavic ancestry higher than it actually is?
Or is the science solid and amateurs just don't get it?

2) If not, then it seems most of the Slavic ancestry in Albanians came from Slavic women, since Slavic paternal lineages are so low... How did that happen? A war that Slavs lost, and Albanians married the remaining Slavic women? Or what other theory?

Sorry if this is amatuer-tier. Just trying to learn.

The Slavic autosomal input and ydna are fairly similar, it looks like Slavs settled initially in their own groups and then were assimilated over time. Also I doubt it was admixed South Slavs that moved to Albania but pure Slavs, if it was South Slavs then it means some other lines like e-v13, i1, j2b l283, r-z2103 also entered Albania with these South Slavs
 
The Slavic distribution between Albanians is not uniform some Have more some have none. Normally the low lands have more.

From this study: https://www.biorxiv.org/content/10.1101/2023.06.05.543790v1.full

"We have shown that modern Albanians from Tirana derive 25-48% of their ancestry from a South-Slavic-related source"

"the ancestry of South Slavs as 55-70% Balto-Slavic-related, with the remainder originating from the pre-Slavic inhabitants of the Balkans

"Although models using the unadmixed Balto-Slavic-related proxy (Russia IA Ingria) receive higher statistical support, they are less likely historically. . . It is therefore likely that Slavic admixture entered the Albanian population via an ancestry profile related to modern South Slavs, rather than unadmixed Balto-Slavic populations."

____________________________________________________________________________________________

If we take the average of 55-70, 62.5%.
62.5% x 25-48, gives a rough average of 15.6% to 30% purely Slavic ancestry in these Tirana Albanian samples, and as low as 13.75% to as high as 33.6% depending on how high the Balto-Slavic and/or how South Slavic shifted the Albanian samples are.

Now... this study shows 31% Slavic 'average' in Albanians. Why are there such large discrepancies between all these models...? How is it possible for Albanians and Greeks to be THAT Slavic shifted?

My question (as a complete amatuer):

1) Is it possible these genetic calculators, including Vahaduo/G25 ones, all prioritize a "better" statistical fit by using overly Southern and Northern sources (rather than genetically intermediate), thereby conflating Slavic ancestry higher than it actually is?
Or is the science solid and amateurs just don't get it?

2) If not, then it seems most of the Slavic ancestry in Albanians came from Slavic women, since Slavic paternal lineages are so low... How did that happen? A war that Slavs lost, and Albanians married the remaining Slavic women? Or what other theory?

Sorry if this is amatuer-tier. Just trying to learn.
I agree that the study seems to be partially wrong.
The Slavic autosomal input and ydna are fairly similar, it looks like Slavs settled initially in their own groups and then were assimilated over time. Also I doubt it was admixed South Slavs that moved to Albania but pure Slavs, if it was South Slavs then it means some other lines like e-v13, i1, j2b l283, r-z2103 also entered Albania with these South Slavs
I think that Albanians mixed with both original slavs,and then with South-Slavic populations.
 
Back
Top