Proto-Albanians and proto-Romanians (Y-DNA)

[Ownstyler]

An interesting discussion started in the thread dedicated to Minoan and Mycenean origins, so let's continue in a more focused manner here. What can Y-DNA data tell us about affinities between proto-Albanian and proto-Romanian and the regions where their speakers might have lived? There seems to a consensus that Moldavian paternal lines are sufficiently different from Albanians to exclude a significant Carpi contribution to the Albanian ethnogenesis, so I will talk just about Albanian-Romanian connections.

According to linguists proto-Albanians and proto-Romanians had intense contacts for several centuries in the early Middle Ages, maybe up to the 10th. So to guess where this area of contacts was, we should look at paleo-Balkan Y-DNA lines younger 1200-1800 years old, that both Albanians and Romanians have. We already know the most common and likely paleo-Balkan haplogroups in Albanians are E-V13, J2b-L283 and R1b. Data I gathered from Bosch 2006, Stanciu 2010, Netea 2012 & Varzari 2013 (total n=363) show Romanians have 16.3% E, 12.1% R1b (<1% BY611), 3.6% J2b2.

Regarding E-V13, as I mentioned, it is difficult to classify short haplotypes into early medieval subclades, but, looking at the main Albanian clusters: PH2180, Y93102, BY4461 and 2 clusters under FGC11450, it seems unlikely that Romanians will show significant connections within 1200-1400 years with them. Nonetheless, because Romanians are so poorly tested, I would say that a thorough analysis of E-V13 needs more data.

Even if we consider the common J2b-L283 lines of Romanians and Albanians as western rather than generally Balkanic lines, this doesn't erase the fact that E-V13 looks more central, southern and eastern, in a Balkanic context

Why do you think that? And how does that play into the early medieval contacts between proto-Albanians & proto-Romanians? Have you found any E-V13 Albanian-Romanian clades with a mrca of 1200-1800?


When it comes to J2b-L283 and R1b-BY611, even though Romanians are not well-tested, the data is already one-sided. The main J2b-L283 Albanian clusters Z38300>Y20899, Y82533, CTS11100 and even smaller ones, like that J-Z631 haplotype found in some Romanians, all have western distributions. J2b-L283 is even more frequent and diverse in Apulia than in Romania or SE Serbia. Btw some Vlachs from Albania have been found to belong to Y20899, a diverse Albanian cluster also found in Apulia.

You write, "There is no reason at all to exclude J2b-L283. J2b-L283 in Tosks actually has some good diversity, and it is still more frequent than in neighboring countries.". Maybe there is not a reason to exclude it, bearing in mind that it is the second most prevalent haplogroup among Ghegs (which are also the ones living the closest to the proto-Albanian homeland), but we also have to consider the frequency among Tosks.

J2b-L283 frequency among Tosks is lower than in Ghegs, even though more frequent than elsewhere. However, the main indicator or historical presence is diversity, and in Tosks it is diverse, albeit less than in Ghegs. But this might not even be a Tosk-Gheg, as much as a highland-lowland difference: if you look at the largest highland of the South, Laberia, J2b-L283 is close to 15%. Other paleo-Balkan lines are also more frequent in mountainous areas, which makes perfect sense because we know the Slavic incomers were mainly settling in agricultural, flat lands, and the native element of the Albanian language has a strong pastorial, highland character. So, I don't see how proto-Albanians might not have had a significant % of J2b-L283.

R1b-BY611 it is at minimal frequency in Romanians. More importantly, I can classify the few haplotypes I have seen until today into either R-Z2705>BY105603>BY61976 or R-Z2705>BY38894>Y32147. Vlachs and some areas in Bulgaria have higher frequency than Romanians, but similar diversity, lacking 393=12 & 392=13 haplotypes (no such haplotypes have been found in Bulgarians, SE Serbs or Vlachs either).

Indeed, most Albanian R1b-Z2103 is R1b-Z2705, and i see that its earliest lines seem to be concentrated in the Dinaric Alps. In any case, the group's predecessor lived approximately 1450 years ago, which was a period of great upheaval and subsequent migrations in the Balkans. Is there any possibility he could have come from the eastern and central Balkans, subsequently pushed to the west by the newcomers? Do we have any significant information on R1b-Y10789 and R1b-Y23373?

Yes, the age of the branch is not high (although it should increase a bit soon), but the likeliest location of tmrca is NW Albania or Montenegro. And the closest relatives whose ancestry we know, at 3300 ybp, are in the W Mediterranean. A few unconfirmed, potential R-BY611 haplotypes, are mostly concentrated in North Italy and NW Balkans (check the thread I linked). Of course, we cannot say with 100% confidence that Moesia can be excluded. But right now, I do not see a single point of evidence suggesting the eastern half of the Balkans.

You write, "So, as I said, there is little shared paleo-Balkan paternal ancestry in the best researched branches, and current data suggests that this small amount spread from the western half of the Balkans.". But shouldn't there have been some, bearing in mind that you are also a proponent of the "Immigrationist theory" and Schramm? As for the current genetic data suggesting that it spread from the west, allow me to view it as more obscure. In any case it is very probable, not just because of genetics, but folklore does give some hints likewise.

Yes, there definitely is some, but not too much. Looking at the current data, it seems likely that proto-Romanian developed south of the Danube, as we agreed, and most linguists and historians suggest. I think there the likeliest scenario that can explain why the shared Y-DNA lines of Albanians and Romanians are relatively rare and more diverse in the west, is that the proto-Romanians lived in an area stretching through Moesia and further west, only the western part being in contact with proto-Albanians, and absorbing a limited number of proto-Albanian Y-DNA lines. Of course, if new data is revlead and it contradicts what we have until now, I might change my mind.

[Johane Derite]

The shared "Albanian-Romanian" substrate doesn't exist apart from a tiny few words. The majority are straight up loan words from Proto-Albanian into Proto-Vlach/Aromanian/Romanian, and don't constitute a "substrate". It would be like saying English loan words in Albanian today like "computer" constitute a shared substrate between Albanian and English. There is a big difference between loan words and substrate.

A substrate by definition is scattered across many domains of life, whereas all the shared words between Albanian and Romanian are related to shepharding and animal husbandry, signifying that proto-romanians were either taught these things from a proto-albanian population or were buying dairy, cheese etc from a proto-Albanian population.

The most probable areas for this exchange is around modern day Kosovo/Nish or around Shtip.

But this only tells us about where proto-romanians and proto-Albanians had contact, it doesn't comprehensively show where proto-Albanians were, or that all proto-albanians were shepherds living only in those regions. These are nuanced points which make a big difference.

[Aspurg]

There exist limited links but some of these might have some importance. Including this connection in vocabulary between Romanian and Albanian. I have all haplotypes, including around 120 ethnic Romanians/Moldovans from FTDNA, overall over 700. I'll start with most significant ones.

1) J-L283>Z631>Z1043*
No less than nine individuals with Basarab surname belong to this cluster. 8 are from Sibiu, one from Gorj. Gorj Basarab shows some differences. Also there are Greeks with this haplotype. Albanian Korbi cluster shows some differences. at dys385ab, also dys635. So there might be 1000 ybp+ between them. Still as this is actually the most common haplotype in individuals with surname Basarab it might be of great importance. Additionally this cluster is found in other studies. ht96 from Petra Neamt/Buhusi, 169 from Ploiesti. Overall 11.

2) R-Z2705
Two Romanians from Dolj and one from Olt are R-BY105603, while there is one under Y32147 (YF08832)

These are only cases where Romanians and Albanians share more than two haplotypes.

3) E-FGC11450>Y146086
One Romanian at FTDNA from Baile Herculaine (one from study from NW Romania based on 385b might fit too). Related to him is likely Ukrainian (of Bulgarian origins). These maybe form a subcluster under Y146086 with Albanians (with Salihu who has 385b=17, but Albanians seem to have another subcluster defined by 447=25, common also in Greeks of obvious Arvanite origin).

4) E-BY4461
One Romanian RO106 from Vrancea.

5) E-BY4684>FGC71980
Unlike previous, there is only one Albanian in this cluster who has a surname with suffix -ul (definite suffix in Romanian/Aromanian), Albanian is in cluster with several Greeks, he is closer to one Greek than is this Greek to the other Greek. Members of this cluster are also 4 Basarab individuals from Sibiu. There is this widespread cluster and there is lone Bulgarian 2000 years away from them. So seems like a Vlach cluster of Eastern Balkan origin.

6) E-Y81468 (likely). Three Moldovans from Karahasani from a study have 389=14-31 (modal for PH1246) and high dys385b=21 (one of defining mutations for Y81468, one has mutated to 22). They are likely Y81468, have unusual 393=14, but hard to say how distant they are to Albanian Dushmani family. One English almost certain Y81468 also has 393=14 but same family member has 13 so not sure which one is older. These are distant to Dushmani.

7) R-L584>A12332*
YF08509 at YFull who is a Vlach from Albania I see has 390=25, 389b=31, dys19=13. Same haplotype is shared by one Gheg from Arberesh study, but also two Greeks, one from Macedonia and one from Athens (study), N113044 Romanian from Putureni, but also Romanians RU380 from Dolj and RU269 from Cluj. RU269 is the only one with dys390=24 and also has off-modal dys635=24, so this clade's home might be actually Romania.

8) J-Z38300
H44 from Oradea belongs to this clade very common in Albanians, there are also Szekely and Hungarians of this clade, one has 23 STR's, looks likely J-Y20899. There is one Aromanian from Romania also, so Vlachs could have brought this clade North of Danube.

9) J-Y22059
You have two Albanians, and there are also two Romanians of this cluster, one from Oradea other from Sibişeni (FTDNA). This cluster is dominated by Serbs who descend of Krici and Usorci tribes, some Bulgarians too.

What is interesting to me is presence of these Vlach J-Z1043* and E-FGC71980 both with surnames Basarab and both from Sibiu. As I've said Albanian Thaci-Korbi differ on at least 3 STR's from Basarab Romanians so they might have branched off in Early Medieval times. Basarab surname might have implications as these might have been part of Wallachian nobility.