July 31, 2008

Expansion of E-V13 explained

E-V13 is the main European clade of haplogroup E. It has been variously interpreted as a signature of early Balkan Bronze Age, or Mesolithic, the Greek colonization of Southern Italy, Greek ancestry in some Pakistanis, or Roman soldiers of Balkan origin in Britain. A proper understanding of its age would help resolve the problem of its origins.

Age, of course, depends on a proper choice of mutation rate, and as I have argued (part I and part II), the proper effective mutation rate is near the germline rate and not 3.6x slower as argued by Zhivotovsky, Underhill, and Feldman (2006). This is especially true for a relatively young haplogroup (very low STR variance compared to other lineages), which is also quite frequent in its area of origin, while much reduced away from it, giving a definite impression of a sudden and relatively recent expansion.

In my previous post, I estimated a Late Bronze Age for E-V13 in Greece and areas affected by historical Greek colonization. I now used Ken Nordtvedt's Generations2 program to obtain estimates of the age of E-V13 in three different datasets: the King set, 12-marker data from the E-M35 Phylogeny Project (Haplozone), as well as E-M78 data -most of which should be E-V13- from Bosch et al. (2006). In the latter set, I used two marker sets: all 12 markers common between Generations2 and Bosch, as well as 8 markers common between them, but excluding markers after DYS392 (in the Generations2/FTDNA order).

N


Age (25y/gen) Age (30y/gen)
Nea Nikomedeia 8
149
1725 BC 2470 BC
Sesklo/Dimini 20
71
225 AD 130 BC
Lerna Franchthi 20
120
1000 BC 1600 BC
Crete 13
68
300 AD 40 BC
Haplozone 103
134
1350 BC 2020 BC
Aromuns (12) 32
71
225 AD 130 BC
Aromuns (8) 32
73
175 AD 190 BC
Slavomacedonians (12) 13
51
725 AD 470 AD
Slavomacedonians (8) 13
59
525 AD 230 AD
Albanians (12) 9
70
250 AD 100 BC
Albanians (8) 9
59
525 AD 230 AD

Both the King et al. E-V13 data, as well as the diverse, mostly European Haplozone E-V13 agree in placing the expansion of this haplogroup squarely in the Aegean Bronze Age.

Aromuns (Vlachs) coalesce to the Roman era, consistent with the idea that they are Balkan natives who became Latinized linguistically at around that era.

Albanians also coalesce to Roman/Late Antique times, consistent with the idea that their high frequency of haplogroup E-V13 (which reaches very high numbers in e.g. Kosovars) is not associated with high diversity. Founder effects in that time frame are the reason for the high frequency of E-V13 in them.

Finally, Slavomacedonians from the former Yugoslav Republic of Macedonia coalesce well into AD times, at around the time of the first Slavic arrivals in the Balkans. This suggests that E-V13 in them is the result of local founders at around that time who adopted the Slavic language. However, Pericic et al. (2005) (see below) report high (but unspecified) diversity of E-M78α in "Macedonia", so it is possible that a larger number of earlier inhabitants were absorbed.

Pericic et al. (2005) give a 7.3kya estimate for the expansion of E-M78α (almost perfectly equivalent to E-V13) for Southeastern European populations north of Greece. Due to their use of the 3.6x slower mutation rate, this figure needs to be converted to equivalent years. The Nea Nikomedeia time depth was estimated as 9.2kya by King et al. Therefore, the equivalent age for the Pericic et al. (2005) expansion is (7.3/9.2) * 149 generations or 118 generations (1,540-950BC). They note that STR variance is higher in Greece, Macedonia, and Apulia, all areas with well-known historical Greek connections.

Cruciani et al. (2007) propose that E-V13 arrived in Europe from West Asia and underwent an expansion in Europe at 4-4.7 kya. This age is calculated using effective mutation rates that are 2.4 or 2.8 slower than the germline rate, which seems to suggest a Late Bronze Age or even later expansion with a rate closer to the germline one.

In the Balkans, it is fairly clear that E-V13 is mostly concentrated south of the Jirecek Line which separated native Greek from Latin speakers. In Italy, the highest frequencies are found in the south, the areas of historical Greek colonization. High frequencies are also attained in Cyprus. Cyprus also high STR diversity, consistent with an early arrival, suggestive of both early Mycenaean and later colonizations from the Aegean.

Conclusion

The age and distribution of E-V13 chromosomes suggest that expansions of the Greek world in the Bronze and later ages were the major causes of its diffusion.

Who was the E-V13 patriarch in Greece? He was perhaps one of the legendary figures of Greek mythology some of whom are said to have come from abroad. For whatever reason, his progeny grew, and were around to participate in the expansion of the Mycenaean world and the subsequent Greek colonization.

UPDATE (Aug. 1):

An additional piece of evidence is Y-chromosome distribution in Calabria, a Southern Italian region with well-known Greek connections. According to Semino et al. (2004) [Am. J. Hum. Genet. 74:1023–1034, 2004], the Calabrian sample has an E-M78 frequency of 16.3%, whereas "Calabria 2" representing the "Albanian community of the Cosenza province" has only 5.9%. This is consistent with the idea that E-V13 in modern Albanians is to a great degree due to Greek founders (Epirotes or ancient colonists).

74 comments:

Maju said...

Pericic: ... point to the Vardar-Morava-Danube river system as one of major routes for E3b1, in fact E3b1 α, expansion from south and southeastern to continental Europe.

How can this fit with Mycenaean or classical Greek expansion is a an absolute mystery to me, honestly.

Cruciani: Haplogroup E-V13 (...) [has] a clinal pattern of frequency distribution from the southern Balkan peninsula (19.6%) to western Europe (2.5%). The same haplogroup is also present at lower frequencies in Anatolia (3.8%), the Near East (2.0%) and the Caucasus (1.8%).

How come a protohistorical or historical Greek-spread gene can be almost as common in Britain as in Anatolia? It is prefectly known that Greek colonization in Anatolian coasts was very intense, but it's quite doubtful that ancient Greeks ever reached Britain at all, much less colonized it. Again the pattern is not consistent with a historical Greek-vectorized dispersal. In fact, the high density in places like Kosova, totally unrelated with metal ages Greek colonization (but not with Neolithic one also originated in Greece) and other inland Balcanic spots (rather than the coasts) strongly suggests a different dispersal pattern.

He also confirms Pericic suggested pattern: ... similarly to the results reported by Peričić et al. (2005) for E-M78 network α, the dispersion of E-V13 and J-M12 haplogroups seems to have mainly followed the river waterways connecting the southern Balkans to north-central Europe.... Pattern that doesn't look at all like Greek colonizaton, either Mycenaean or classical.

Dienekes said...

>> Pericic: ... point to the Vardar-Morava-Danube river system as one of major routes for E3b1, in fact E3b1 α, expansion from south and southeastern to continental Europe.

>> How can this fit with Mycenaean or classical Greek expansion is a an absolute mystery to me, honestly.

Well, if you bothered to read what you are commenting on, it wouldn't be a mystery.

Pericic et al. use the Z.U.F. rate, therefore their conclusions are wrong.

>> How come a protohistorical or historical Greek-spread gene can be almost as common in Britain as in Anatolia?

E-V13 isn't common in Britain; it could very well have been introduced by assorted "Romans" there. As for Anatolia, it was colonized predominantly by Ionians who were of "Pelasgian" origin, not by Dorians. This also explains neatly the higher frequency of E-V13 in Southern Italy and Sicily which was mainly colonized by Dorians, as well as in Macedonia, as the Dorians and Macedonians were related. And, of course, we shouldn't forget that Anatolia is inhabited by Turks who are genetically a composite of various peoples, and only partly of the Greeks.

>> In fact, the high density in places like Kosova, totally unrelated with metal ages Greek colonization

Albanians aren't native to Kosovo, they are mostly post-medieval immigrants there.

There are probably three sources of E-V13 north of Greece: Epirotic E-V13 (E-V13 is high in Epirus, the region identified by Aristotle as ancient Hellas), Macedonian E-V13; it is well known that the Macedonians dominated and Hellenized the barbarians to their north; as well as southern Greek, especially Corinthian/Dorian from the coastal colonies of what is now Albania.

Dienekes said...

I should have written "NW of Greece". "NE of Greece": in Bulgaria: 17%, mostly south of the Jirecek line with plentiful colonies on both the Aegean and Black Sea coast in close proximity. Only 8% in Romania, north of the J.L. 9% in Ukraine with many Greek settlements, only 2-4% in Russia.

Maju said...

Pericic et al. use the Z.U.F. rate, therefore their conclusions are wrong.

It doesn't matter at all. My problem is because of geographic distribution patterns. I'm not going to discuss your hypothesis on mutation rates (just another guesstimate) but I find that the geographical patterns for E-V13 do not fit at all with any explanation that involves overseas Greek colonization of coastal areas.

Albanians aren't native to Kosovo, they are mostly post-medieval immigrants there.

That's a valid objection, ok. But still Albania proper was not a place of major Greek colonization. It was still Illyrian when the Romans arrived there.

E-V13 isn't common in Britain...

Ok. Admitted: I misread Britain where it says "Western Europe". Still Greek colonization of Western Europe, excepting Marseilles, was nearly null. Nevertheless E-V13 is almost as common there as in heavily Greek-colonized Anatolia or Caucasus.

There are probably three sources of E-V13 north of Greece: Epirotic E-V13 (E-V13 is high in Epirus, the region identified by Aristotle as ancient Hellas), Macedonian E-V13; it is well known that the Macedonians dominated and Hellenized the barbarians to their north; as well as southern Greek, especially Corinthian/Dorian from the coastal colonies of what is now Albania.

Macedonians dominated the Paeonians and in Philip's time they conquered part of Thrace (southern Bulgaria and Greek/Turkish Thrace). No Greek presence ever existed north of the Jirecek Line (and south of it it was mostly cultural influence, not colonization), yet you have more E-V13 in Rumania or Ukraine than in Anatolia and almost as much E-V13 in Bulgaria than in Greece itself (your own data). The historical patterns of Greek colonization don't fit with that again.

And you cannot ignore that the main "arrow" of E-V13 into the continent, as quoted before, is via Serbia ("Vardar-Morava-Danube"). The pattern only fits with one known historical (actually prehistorical) process: Neolithic colonization beginning in Thessaly (Balcanic and, secondarily, Danubian Neolithic).

Dienekes said...

>> but I find that the geographical patterns for E-V13 do not fit at all with any explanation that involves overseas Greek colonization of coastal areas.

It was explained to you that E-V13 does not involve only Greek colonization of coastal areas, but also Epirotes/Macedonians. Greeks moved freely in the region where E-V13 is frequent for 900 years between the time of Alexander and the arrival of the Slavs.

>> Still Greek colonization of Western Europe, excepting Marseilles, was nearly null. Nevertheless E-V13 is almost as common there as in heavily Greek-colonized Anatolia or Caucasus.

Marseilles wasn't the only Greek settlement in Western Europe. Also, there was plenty of opportunity in Roman or later times for a few Mediterranean migrants to reach Western Europe; it's not like E-V13 reaches any high frequency there.

No Greek presence ever existed north of the Jirecek Line (and south of it it was mostly cultural influence, not colonization), yet you have more E-V13 in Rumania or Ukraine than in Anatolia and almost as much E-V13 in Bulgaria than in Greece itself (your own data).

There were plenty of Greek colonies of the Romanian coast, and in Ukraine, not to mention the Bosporan kingdom, and medieval settlements of Greeks in the Crimea, or Phanariotes and Greek traders in present-day Moldova and Romania.

The claim that there were no Greeks north of the Jirecek line is laughable.

Dienekes said...

Neolithic colonization beginning in Thessaly (Balcanic and, secondarily, Danubian Neolithic).

Neolithic colonization is incompatible with the age of E-V13. It is based on the Z.U.F. factor of 3.6, which in turn is based on simulations where the average 300-generation (Neolithic) haplogroup has 150 living descendants, at least four orders of magnitude lower than the reality of E-V13.

Crimson Guard said...

Is there any figure percentages?

I think the Greek colonization idea is kinda a stretch.

Dienekes said...

I think the Greek colonization idea is kinda a stretch.

I am willing to consider alternative explanations that are consistent with both the molecular and archaeological-historical data.

KerryODair said...

Your assumptions are based on Ken's formula on estimate of age. Can this age estimation change radically while looking at a larger set of markers. You have used 105 haplotypes from haplozone. Of that 105 there are 60+ haplotypes with 67 markers tested and verified by deep sub clade test that they are V13. There is a definite clustering of four groups in this group of 60+.

I would be interested to know if Ken's formula for age holds up with 67 markers. I think this could be a stronger case for an age of the clade. Even though the database is heavily skewed towards haplotypes from the British Isles there are many Eastern European haplotypes as well.

To take it a step further how Ken's interclade formula would calculate V13's versus V12's.

Dienekes said...

132.4968188 generations for 51 V13+ 67-marker haplotypes (52 in database - 1 labeled as modal), about 2,000-1,300BC (depending on generation length).

Dienekes said...

I click on 67 for length, V13+ for SNP and Select Panel: All, not sure why you say there are 60+ haplotypes, are there different options I should use?

Dienekes said...

255 generations for the interclade between V13+ and V12+ or 5700BC-4400BC.

KerryODair said...

Select Quick Query and select by panel all option in the first selection. Select snp for v13 and subgroup members only option. Step 4 The first sort can be done by length. All the 67 marker haplotypes should descend to the bottom of the list. This also can be done with V12's. We have 14 67 marker tests with 6 of these deep clade tested as V12. The balance of the six are with high probablity also V12 as well.

Dienekes said...

I still don't see 60+ V13+ haplotypes

Dienekes said...

With my heuristic (accumulation of variance at 0.87mu/generation and the 1/400 mutation rate, see part I), the 67-marker E-V13 haplotypes from haplozone give an age of 139 generations.

KerryODair said...

Correction 50+ I was doing a quick visual count. So if I understand you correctly the age based on the haplozone haplotypes for V13 is 2000 B.C. at 30 years a generation and 1300 B.C at 25 years a generation.

The interclade date between V13's and V12's 5700 B.C. or 4400 B.C. based on the same years per generation above.

This in affect this would be a founders date for those specific haplotypes in the haplozone database?

Dienekes said...

It's an estimate of the time to their MRCA.

KerryODair said...

Thank you for the explaination. It seems that 67 markers had no great impact on the dating. I do find it a bit odd that the 12 marker date of 2020 B.C. to be older than the 67 marker date of 2000 B.C. in regards to the V12's. On the surface I would have thought the 67 markers would generate an increase in age.

Hopefully, someone with better talents than myself can write a utility to utilize Ken's formulas for clade and interclade MRCA. We are in need of something easier to deal with like the McGee Utility. It is very difficult to follow Ken's formula's without a significant math background.

KerryODair said...

Correction to last post should read "in regards to V13's" not V12's

Maju said...

It was explained to you that E-V13 does not involve only Greek colonization of coastal areas, but also Epirotes/Macedonians. Greeks moved freely in the region where E-V13 is frequent for 900 years between the time of Alexander and the arrival of the Slavs.

I don't think I can agree with that. Alexander's empire (and its successors) never reached further north than Paeonia and southern Thrace. Just look at any history book: no Greek presence in Serbia, or Hungary ever (unless you want to believe they are all descendants from Jason himself, who's said to have sailed the Danub river). And for the rest (most of Bulgaria, Rumania, Ukraine, Albania) Greeks had some coastal colonies, ok, but they never controlled anything in the interior.

Marseilles wasn't the only Greek settlement in Western Europe.

It was the only colony worth that name. The rest were emporia: mere trading posts. And all circunscribed to the area between the Alps and the Pyrenees (Rhodes and Emporion were at the very eastern end of the Pyrenees). Greek trade existed further south but the area was under Phoenician control, and the only heard-of Greek emporium there (Hemeroskopeion) is quite doubtful and has never been found archaeologically.

Marseilles and a handful of trading posts cannot justify the claim that nothing less than 2.7% of Western European Y-DNA belongs to E-V13. Sorry, no way. Would it be 2.7% of Provenzal Y-DNA, then it could be even somewhat low, but all Western Europe?!

Such figure can only pertain to a deeper colonization in an older date when population density was lower. And that can only be the arrival of Neolithic.

A "Greek" component, yes, but in a different context and timeline.

Also, there was plenty of opportunity in Roman or later times for a few Mediterranean migrants to reach Western Europe; it's not like E-V13 reaches any high frequency there.

We are talking of one in every 37 men. It's not that low either.

And it's not low at all if compared with Anatolia, where Greek penetration was widespread. They were Ionians? Ok. But Chios and Lesbos show at least as much haplogroup E (your data in a previous post) as Greek Macedonians, to whom you want to attribute a massive colonization of the inland Balcans and southern Central Europe.

There were plenty of Greek colonies of the Romanian coast, and in Ukraine...

But not in the interior. And not in Serbia or Hungary at all.

not to mention the Bosporan kingdom

The hellenized realm of the last Cimmerians...

... and medieval settlements of Greeks in the Crimea, or Phanariotes and Greek traders in present-day Moldova and Romania.

I wouldn't like to revive the past discussions you had with other people on Slavic or Turkish (or Roman or Jewish or Venetian or even Catalan, I could add) presence in Greece and their possible genetic effects. But I fear you are mirroring their attitude: trying to build on what could well be considered anecdotic and peripherical historical events (in some cases not even historical) a model where reality doesn't really fit.

I would probably be among the last ones to deny the historical importance of Greece but I know its crucial role started much earlier (Neolithic) and I also know pretty well the historical extent of Greek colonization and influence. And the map of Greek colonization (or the Alexandrian empire if you wish) dosn't fit well with the patterns of distribition of E-V13.

I agree that it's of Greek origin (having arrived there probably via West Asia - founder effect) but then it basically moves northwards through the central Balcans, what is very different than the coastal colonization of historical and Mycenaean Greeks.

Sorry but the facts fon't fit with your model. It's Greek, yes, but much older than what you say.

Dienekes said...

>> And it's not low at all if compared with Anatolia, where Greek penetration was widespread. They were Ionians? Ok. But Chios and Lesbos show at least as much haplogroup E (your data in a previous post) as Greek Macedonians, to whom you want to attribute a massive colonization of the inland Balcans and southern Central Europe.

The original population of Chios was wiped out by the Turks during the Greek War of Independence. As for Lesbos, it was colonized by Aeolians not Ionians.

In any case, it's far easier for Greeks of any kind to make a demographic impact in islands than in Anatolia where the population was far larger, and only certain coastal areas were colonized.

>> But I fear you are mirroring their attitude: trying to build on what could well be considered anecdotic and peripherical historical events (in some cases not even historical) a model where reality doesn't really fit.

That is the only model that fits if one considers the point of origin and age of E-V13.

Maju said...

In any case, it's far easier for Greeks of any kind to make a demographic impact in islands than in Anatolia where the population was far larger, and only certain coastal areas were colonized.

Understood and fully agreed. But for the same reason it should have been very difficult to make an impact in the inland Balcans, right? Yet you have rates of E-V13 that are many times those of Anatolia there.

Why? There's only one explanation: because it's Neolithic and while Thessalian Neolithic had a major impact in the Balcans, and to a lesser extent in all Europe, it did not in Asia Minor.

Dienekes said...

According to your theory, people to the north of Greece got E-V13 from them in the first 3-4ky since the beginning of the Neolithic and not in the next 3-4ky since the emergence and expansion of the recognizable Greek world.

This theory is not very plausible, and fails completely once we consider that E-V13's STR variation is completely incompatible with the Neolithic. This was made manifest by Cruciani et al. even though he used Z.U.F.-style correction factors for the mutation rate (2.4 and 2.6); he still placed the E-V13 expansion in the Bronze Age.

One can make E-V13 Neolithic only by discounting the STR evidence for its age.

Maju said...

One can make E-V13 Neolithic only by discounting the STR evidence for its age.

Sure. I consider age estimates based in the MC hypothesis for what they are: educated guesses. I do pay some attention to them but I do not consider them "evidence" of any sort, just abstract mathematic models based on other estimates such as the Chimp-Human divergence date, hunches on effective population sizes, and so on. There's no way to create "evidence" with that: it's just a good hunch, like equations suggesting how many planets may be inhabited in the Galaxy and the like: we just don't have enough knowledge of the actual variables involved to make accurate estimates. It is not C-14.

In any case, any model should be as consistent as possible with the available data, which is mostly modern distribution patterns and internal clade structure. IMO, it is when you ignore them when you are rejecting the evidence. If the age estimates do not fit with clade structure/distribution and well known prehistory/history, then it surely means that age estimates are the ones wrong.

Why? I'm not sure but the known geological dates for the formation of the Congo river and therefore the most likely scenario of separation of chimps and bonobos is quite older than the date achieved via the fossil record. At least 15% older, and maybe quite more. Also mutation rates seem to be the subject of contention, as well as other variables involved in the calculations. In general the whole TRMCA model seems in need of some very good fine tuning before it can produce reliable estimates.

In any case, I quote from your original post:

Pericic et al. (2005) give a 7.3kya estimate for the expansion of E-M78α (almost perfectly equivalent to E-V13) for Southeastern European populations north of Greece.

That is roughly 5,300 BCE, not much more recent than the actual archaeological dates for the expansion of Balcanic Neolithic. Surely the error margins allow for something like 6000 BCE, and that would fit perfectly. 5000 BCE would fit with the Dimini-Vinca wave, that also expanded northwards via Serbia (but that affected a much smaller territory anyhow).

I know you don't agree but Pericic's estimate seems to fit with the evidence.

Dienekes said...

>> n any case, I quote from your original post:

>> >> Pericic et al. (2005) give a 7.3kya estimate for the expansion of E-M78α (almost perfectly equivalent to E-V13) for Southeastern European populations north of Greece.

>> That is roughly 5,300 BCE, not much more recent than the actual archaeological dates for the expansion of Balcanic Neolithic. Surely the error margins allow for something like 6000 BCE, and that would fit perfectly. 5000 BCE would fit with the Dimini-Vinca wave, that also expanded northwards via Serbia (but that affected a much smaller territory anyhow).

>> I know you don't agree but Pericic's estimate seems to fit with the evidence.

In other words, you discount the molecular clock hypothesis, but are perfectly willing to use a miscalibrated one if it agrees with your pet theory.

Maju said...

In other words, you discount the molecular clock hypothesis, but are perfectly willing to use a miscalibrated one if it agrees with your pet theory.

I am not "using" it, just noticed that it fits much better with the evidence. That could be a way to calibrate MC estimates: to contrast their results with the pre-/historical scenarios and structural/geographic evidence. In fact it is probably the only way to calibrate it properly (assuming it can be calibrated at all).

And it's not my pet theory anyhow. It's the only thing that makes any sense and therefore is quite mainstream. I did not invent such theory, I just agree that it's the only one that makes any sense.

Dean said...

The last part of the blog entry is interesting. Whose idea is it that modern Albanians of E3b type are largely descended from Greek founders?

Maju said...

Whose idea is it that modern Albanians of E3b type are largely descended from Greek founders?

Dienekes'.

Of course they would still be "Greek" descendants with the Neolithic-spread model too - though surely calling "Greeks" to the Neolithic peoples of Greece is a loose use of the ethnonym.

Dienekes said...

dean, there were many Greek colonies in what is now Albania in antiquity, especially from the Peloponnese where E-V13 is very frequent and old, whereas Albanian E-V13 seems much younger. There were also Epirotes and Macedonians near by.

So, it makes excellent sense that the arrow of transmission was Greeks=>Albanians, since E-V13 is more diverse in Greeks than in Albanians, and its origin outside Europe makes it likely that it moved from the Aegean to the Balkans.

Andrew said...

Dienekes,

At least in terms of writing up project updates for genealogists, I really like your speculation about a Greek semi-mythical patriarch being "Mr European V13". But can I tempt you into an admittedly un-scientific speculation about which semi-mythical figure he might have been? :) Of course I would never claim that this was anything other than playful speculation.

Maju,

Steven Bird must be watching, and I feel sure he'd be thinking in a direction which might be compatible with what you are saying, but also with the basic facts Dienekes is presenting: could Dienekes be right about the age, but only roughly right about the ethnicity: The Thracians were a major nation in the Hellenic world, though not quite Greek in most senses of the word. And their links to northern Europe may have been stronger, and even strengthened during Roman times when they were conquered and spread around.

Andrew

Maju said...

could Dienekes be right about the age, but only roughly right about the ethnicity

Dienekes has recently developed a new theory of his own on genetic age estimates. As I have commented before, I think that age estimates are basically educated guesses because even if they usually take form of sophisticated mathematical equations, the variables in them are mostly guesses themselves.

So for me Prehistory, specially when well studied, as it's the case for Europe, with its C14 dates and its archaeo-cultural sequences, is a reference that cannot be ignored.

The pattern of E-V13 fits best with that of Balcanic Neolithic. I am not aware of Thracians ever reaching into Northern Serbia or Hungary, while their Phrygian relatives did penetrate in Anatolia. I don't see that Thracians may explain the Vardar-Morava-Danub axis either.

Dienekes said...

But can I tempt you into an admittedly un-scientific speculation about which semi-mythical figure he might have been? :)

Pelops, Danaus, Deucalion, Hercules?

The first two were foreign and with Peloponnesian connections, the first one naming the land, the second one naming the Danaans.

The last two were "divine-born". Deucalion has "pan-Hellenic" connections, whereas Hercules had his Heraclids who led the Dorians into the Peloponnese, while the Macedonian royal house were also Heraclids, more precisely Temenids.

This is all speculative, of course. If we ever manage to Y-SNP-test ancient bones, we will be able to tell whether or not the different Greek tribes had different patrilineal origins.

could Dienekes be right about the age, but only roughly right about the ethnicity: The Thracians were a major nation in the Hellenic world, though not quite Greek in most senses of the word.

Given that Thrace was colonized by Greeks but not the Peloponnese by Thracians, I think the weight of evidence is in favor of a Greek rather than a Thracian origin. The Thracian theory also doesn't explain the strong association of E-V13 with Greek colonization, e.g. in Cyprus or Magna Graecia.

Andrew said...

Maju: their Phrygian relatives did penetrate in Anatolia.

Good point. You are referring to the fact that V13 is not thought common there?

Dienekes: The Thracian theory also doesn't explain the strong association of E-V13 with Greek colonization, e.g. in Cyprus or Magna Graecia.
...and similarly Maju: I am not aware of Thracians ever reaching into Northern Serbia or Hungary

On this type of reasoning things are difficult. We can not ignore that the Roman and Byzantine empires mixed the peoples of these areas to at least some extent, as did the collapse of the latter.

Maju said...

Maju: their Phrygian relatives did penetrate in Anatolia.

Good point. You are referring to the fact that V13 is not thought common there?


Exactly.

On this type of reasoning things are difficult. We can not ignore that the Roman and Byzantine empires mixed the peoples of these areas to at least some extent, as did the collapse of the latter.

We already discussed the issue of possible Roman effects and, IMO, their effect should have been low and in any case should not have expanded Greek genetic influence north of the Greco-Latin linguistic divide, much less via marginal inland routes.

The pattern in the Balcans does not show a dominant coastal distribution, as would be expected from historical Greek colonization and later Greek trade within the Roman Empire. It shows a clear inland axis that can only be associated to Neolithic expansion - of all the historical and prehistorical phenomenons we know about.

This doesn't mean that Cyrpiot genetics are not of Mycenaean age. Cyprus and Southern Italy, as well as Anatolia must have suffered quite strong Greek demic influx in the Mycenaean, Classical and even Hellenistic periods.

The low presence of Greek-specific genetics (and of E-V13 specially) in Anatolia can be partly (but only partly) explained by the expulsion of ethnic Greeks by Turkey in 1920. The main explanation though must be that the region was already densely populated and that Greek (and Phrygian) colonizaton therefore had only a limited impact in the genetic pool.

Dienekes said...

Maju: their Phrygian relatives did penetrate in Anatolia.

The Phrygians migrated to Asia Minor in the 12th c. BC, so we wouldn't expect them to carry E-V13 there under the scenario of a Bronze Age Greek origin of E-V13. Under a Neolithic scenario, E-V13 would already be frequent among them, and they would have introduced it to Asia Minor.

John said...

No Greek presence ever existed north of the Jirecek Line (and south of it it was mostly cultural influence, not colonization)

Wow. Just popped in from a google.
On the European side of the Black Sea, Greek colonization along the west coast was quite heavy up through the Danube delta and beyond. We are talking major settlements.

To the south of the line, in the coastal area of Thrace, the population was almost entirely Greek and were so from antiquity through the early 20th century. These were quite large populations compared to the interior.

I am no expert on genetics but know the history and archeology of this region quite well.

Maju said...

That doesn't contradict what I said. I'm talking about the mainland, not the coastal strips. In fact the main issue is that the principal axis of E-V13 spread is along the Vardar-Morava-middle Danub axis, which obviously has nothing to do with maritime Greek colonization (but correlats well with Neolithic expansion).

Dienekes said...

>> the principal axis of E-V13 spread

You keep repeating that and yet Albania and Bulgaria have more E-V13 than the countries in your "principal axis".

Maju said...

You keep repeating that and yet Albania and Bulgaria have more E-V13 than the countries in your "principal axis".

I follow Pericic in that (see above). Should I think he is wrong? Why?

Cruciani 2007 also mentions some oddballs for your Greek maritime theory:
- Slovaks: 8.33% E-V13
- Hungarians: 9.43% E-V13

No mention of Serbs but the Republic of Macedonia has as much E-V13 as mainland Greece (17%), while Albanians double that figure (32%).

All this is absolutely inconsistent with an explanation that only accounts for historical and proto-historical Greek maritime colonization. There's no way that there is more "Greek" genetics in Albania than in Greece itself... unless it means a founder effect at a time when the country was virtually desertic. That time can only be the Neolithic.

There is no way to explain the high levels of E-V13 in the middle Danub if this clade has a mythical Greek founder of Mycenaean times. But it is well explainable by Neolithic colonization instead.

Btw, the highest concentrations of E-V13 west of Sicily are among Atlantic Iberians (Portuguese, Asturians), where historical Greek colonization was zero. This must mean (again) some older flow of Neolithic or maybe Chalcolithic age. There are other significative ammounts of this clades in most unlikely places like Denmark (3%), Germany (4%). All that can only be explained with Neolithic founder effects or something of the like. Ukranians (strong in E-V13) may have affected Northern Europe genetically... but at a time when the Greek ethnicity did not yet exist as such.

Dienekes said...

There's no way that there is more "Greek" genetics in Albania than in Greece itself

First of all, E-V13 isn't a quantifier of Greek genetics, it is one of many Greek founding lineages.

Second, I explained to you that diversity in the Peloponnese is higher than in Albania. This is consistent with the foundation of Greek colonies in the Adriatic by Peloponnesians who carried a subset of the genetic diversity which had accumulated in the source region. By contrast, E-V13 diversity is lower in Thessaly than in the Peloponnese, a contradiction of your "Thessalian Neolithic" thesis.

Repeating Pericic isn't going to solve the problem of E-V13 being more frequent in Bulgaria and Albania than in the supposed corridor leading to Central Europe.

In any case, there is no study of the geographical distribution within the studied countries in Pericic to substantiate the claim of concentration close to rivers. Using a wrong mutation rate => wrong date => wrong archaeological correlates.

Maju said...

By contrast, E-V13 diversity is lower in Thessaly than in the Peloponnese, a contradiction of your "Thessalian Neolithic" thesis.

From your own data of May 21, Sesklo/Dimini showed the highest apportion of haplogroup E (40%), which I assume is largely E-V13. The stats you are managing now must have been watered down by Larissa which instead shows a very marked "Cretan" signature in their Y-DNA (coincident with the mythical origins of this city).

Larissa was not there in the Early Neolithic but Sesklo was and is generally considered the seed of most European Neolithic.

Dienekes said...

From your own data of May 21, Sesklo/Dimini showed the highest apportion of haplogroup E (40%), which I assume is largely E-V13.

That is the frequency of DE not E-V13. The frequncy of E-V13 is 35.1% but its estimated age (by King et al.) is 4.3kya, long after the Neolithic, even using the slow evolutionary rate. Peloponnesian E-V13 is more than twice as old.

Maju said...

That is the frequency of DE not E-V13. The frequncy of E-V13 is 35.1%...

What I was assuming more or less (I said: which I assume is largely E-V13. It is in any case one of the highest concentrations of E-V13 in the planet.

... but its estimated age (by King et al.) is 4.3kya, long after the Neolithic, even using the slow evolutionary rate. Peloponnesian E-V13 is more than twice as old.

I don't give much value to these estimates, as you know, but certainly it is interesting that Peloponesian E-V13 appears to be older by that method. Maybe it is Mesolithic and not Neolithic after all?

What doesnt make any sense is that some "hero" appeared suddenly (from where?) and managed to make his rare clade widespread, not just in Greece and its area of influence (what in itself sounds impossible) but as far away as Slovakia. IMO if hat would be the origin of any actual haplogroup, the most likely result is that it would have been reabsorbed by the dominant ones soon after by mere drift - at most he could have left some erratics.

Dienekes said...

You have lots of opinions but no hard data to back them up except your blanket skepticism of the molecular clock. But even the slow molecular clock of Z.U.F. leaves 4,000 years between the Thessalian Neolithic and the age of E-V13. Perhaps the molecular clock ticks at a custom-made "Maju rate" in Thessaly. But even with your "Maju clock" you can't deny that E-V13 is older in the Peloponnese. So, to substantiate your "Thessalian Neolithic" hypothesis, you need your Thessalian clock to be real slow, and the Peloponnesian clock to be real fast.

Maju said...

Are you calling the MCH "hard data"?

AFAIK you have no model to explain E-V13 in Slovakia being more important than in Anatolia and as important as in Sicily. As I see it, it is all wishful thinking based 100% on a handful of TRMC estimates.

I have hard data: real distribution patterns and quite well estabilished archaeological knowledge. You have mathematical card houses. Sorry but it is that way. Not for being more mathematical it is more scientific necesarily.

Besides, my opinions, for the most part, are not anything that I have built out of the blue, the same model has been suggested once and again by many many scientifical publications.

Dienekes said...

AFAIK you have no model to explain E-V13 in Slovakia being more important than in Anatolia and as important as in Sicily.

Right... 2 out of 24. Talk about a mathematical house of cards. But, I forgot, you don't like math or statistics.

Maju said...

Dienekes: you are an intelligent person, you surely can see perfectly that the mainly Balcanic spread pattern of this clade does not correspond with Greek historical colonization, which was focused mainly in other areas and always limited to the coasts.

Guess you know the tale of the Emperor's New Clothes, right?

Andrew said...

Coming back to this old one, Dienekes I am surprised that a person of your interests has not remarked on the possible role of "Arvanites" in the Greek distribution of V13. Might the higher variety of V13 in Greece be because different clades mixed there? (Similar to the way in which Sicily has high variety of M78.)

Dienekes said...

I am surprised that a person of your interests has not remarked on the possible role of "Arvanites" in the Greek distribution of V13

This is unlikely to have made an impact, for several reasons:

1) Arvanites settled in very specific Greek locations and quite recently (less than 1000 years for sure), but E-V13 is found all over Greece, in Greek testees from Asia Minor, in Crete, in Cyprus. Indeed, it is one of the most uniformly distributed haplogroups in Greece, which is inconsistent with it being the result of recent admixture.

2) To reach Albania, E-V13 must have passed through Greece. It is less parsimonious to suggest that E-V13 flew over Greece to its north and then back to Greece

3) I already mention in my post the evidence from Calabria where the Albanian minority does not possess it in any great degree, while regular Calabrians (who are significantly of Greek descent, both ancient and medieval) do.

4) Peloponnesians have both more of it and it is more diverse in them than in Albanians

5) Admixture can sometimes lead to increased diversity, but this happens when differentiated undetected subclades of a haplogroup mix. On the other hand, E-V13 has a very clear starlike phylogeny and thus this is not the case. The mix of pre-Arvanite and Arvanite E-V13 would simply not increase STR variance by much.

6) There is good evidence that the Albanians did not inhabit the region they now inhabit originally. Albanian lacks sea-related vocabulary and Greek loan words, which would be strange if they were the next-door neighbors of the Greeks since prehistory.

The E-V13 in Albania is probably the result of absorption of pre-Albanian speakers, particularly Epirotes and the descendants of Greek colonists from the Peloponnese.

Andrew said...

d: 1) Arvanites settled in very specific Greek locations and quite recently (less than 1000 years for sure), but E-V13 is found all over Greece, in Greek testees from Asia Minor, in Crete, in Cyprus. Indeed, it is one of the most uniformly distributed haplogroups in Greece, which is inconsistent with it being the result of recent admixture.

Answer: The proposal I am asking to be considered is whether different branches of V13, mixed in Albanian speaking areas of Greece, this increasing diversity locally. So this would of course not be inconsistent with V13 being widespread over the Balkans including Greece.

2) To reach Albania, E-V13 must have passed through Greece. It is less parsimonious to suggest that E-V13 flew over Greece to its north and then back to Greece

First, see your own point 6 below. Second, fact is that no one knows where the Albanian language comes from, and more generally how people moved around the Balkans over history is pretty mysterious, and that includes the population of modern Greece. How on earth can we decide which of these populations stems mainly from male lines closer to Asia than the other? Also see my response to your point 1.

3) I already mention in my post the evidence from Calabria where the Albanian minority does not possess it in any great degree, while regular Calabrians (who are significantly of Greek descent, both ancient and medieval) do.

If we talk about small emigrant communities I think founder effects limit what conclusions we can draw surely?

4) Peloponnesians have both more of it and it is more diverse in them than in Albanians

And this part of Greece had Albanian speaking areas?

5) Admixture can sometimes lead to increased diversity, but this happens when differentiated undetected subclades of a haplogroup mix. On the other hand, E-V13 has a very clear starlike phylogeny and thus this is not the case. The mix of pre-Arvanite and Arvanite E-V13 would simply not increase STR variance by much.

Maybe I misunderstood you, but I thought you were arguing that there is diversity detectable in V13, which was your argument that it is older in Greece than Albania. If V13 is just does not show "structure" so to speak then I do not understand how you intended to argue that.

6) There is good evidence that the Albanians did not inhabit the region they now inhabit originally. Albanian lacks sea-related vocabulary and Greek loan words, which would be strange if they were the next-door neighbors of the Greeks since prehistory.

This is very interesting of course, but firstly we have to distinguish the Albanian people from the Albanian language, which seems to have been a local language in a Christian enclave in Ottoman times. Both Greece and Albania seem to have gone through periods when they were less populated? Second we have to suggest what this means concerning your hypothesis that Greek V13 is older than Albanian V13.

Dienekes said...

The proposal I am asking to be considered is whether different branches of V13

Covered in #5

2) To reach Albania, E-V13 must have passed through Greece. It is less parsimonious to suggest that E-V13 flew over Greece to its north and then back to Greece

First, see your own point 6 below


If my point #6 is valid, and Albanian-speakers didn't inhabit Albania in the past, then it would be even less likely that E-V13 would have reached them bypassing Greece, since no one seriously suggests that Albanians originated to the south or east of Greece (from where E-V13 must have come from).

If we talk about small emigrant communities I think founder effects limit what conclusions we can draw surely?

A fair number of Albanians settled in Calabria. If E-V13 was a high-frequency haplogroup in early Albanians then it is highly unlikely that so little of it would be found in their descendants.

And this part of Greece had Albanian speaking areas?

Sure.

Maybe I misunderstood you, but I thought you were arguing that there is diversity detectable in V13, which was your argument that it is older in Greece than Albania. If V13 is just does not show "structure" so to speak then I do not understand how you intended to argue that.

Diversity does not equal structure. E-V13 haplotypes in Greece, Albania, and indeed the rest of Europe belong to a single star cluster. It's simply the case that they are more diverse in Greece than in Albania. This has been ascertained both by Cruciani who mentions Greece, FYROM and Apulia as areas of high E-V13 Y-STR diversity, and is evident also in the separate Albanian sample of Bosch.

but firstly we have to distinguish the Albanian people from the Albanian language

The Albanian people are the people who speak the Albanian language. The genetic composition of this people (and indeed all non-isolated peoples) has varied over time.

concerning your hypothesis that Greek V13 is older than Albanian V13.

Not a hypothesis: a fact.

Andrew said...

>If my point #6 is valid, and Albanian-speakers didn't inhabit Albania in the past, then it would be even less likely that E-V13 would have reached them bypassing Greece, since no one seriously suggests that Albanians originated to the south or east of Greece (from where E-V13 must have come from).

Surely Thrace is one of the theories that goes around, which means East, and includes areas very close indeed to Asia (closer than the high V13 parts of Greece?) including parts of Greece according to whatever historical definition you use?

>>If we talk about small emigrant communities I think founder effects limit what conclusions we can draw surely?

>A fair number of Albanians settled in Calabria. If E-V13 was a high-frequency haplogroup in early Albanians then it is highly unlikely that so little of it would be found in their descendants.

I'd like to hear more about this. What are the relatives percentages, and how big was the study?

>>Maybe I misunderstood you, but I thought you were arguing that there is diversity detectable in V13, which was your argument that it is older in Greece than Albania. If V13 is just does not show "structure" so to speak then I do not understand how you intended to argue that.

>Diversity does not equal structure. E-V13 haplotypes in Greece, Albania, and indeed the rest of Europe belong to a single star cluster. It's simply the case that they are more diverse in Greece than in Albania. This has been ascertained both by Cruciani who mentions Greece, FYROM and Apulia as areas of high E-V13 Y-STR diversity, and is evident also in the separate Albanian sample of Bosch.

Can you explain the difference you are making between diversity, what you say is found in Greek V13 more than in Albanian V13, and what happens when different parts of a population move into one area? To me at first sight it does not seem to matter for example if there is a star shaped cluster or "detected subclades". Any population has some level of diversity and concentrating that population geographically must increase the diversity relative to a geographical area?

Dienekes said...

Surely Thrace is one of the theories that goes around, which means East, and includes areas very close indeed to Asia (closer than the high V13 parts of Greece?) including parts of Greece according to whatever historical definition you use?

I have not seen any evidence that E-V13 in the Balkans originated in Thrace. This hypothesis has more to do with a romantic fascination with the Thracians than with any actual evidence.

With that said, a possible origin of Albanians in Thrace solves none of the problems that I mentioned (sea terms and lack of Greek loan-words). Indeed, one would expect Albanian to have even more Greek loan words if it originated in Thrace, as Thrace was even more heavily colonized by Greeks than the Adriatic coast.


I'd like to hear more about this. What are the relatives percentages, and how big was the study?

It's in the bottom of the blog entry

To me at first sight it does not seem to matter for example if there is a star shaped cluster or "detected subclades"

Admixture doesn't always increase variance. The necessary condition is for the two sets of Y-chromosomes that mix to be differentiated.

If, e.g. a bunch of Bosnian I men moves to Sweden, and you measured the variance of haplogroup I in Sweden before and after this move, then you would see an increase, the reason being that Bosnian and Swedish haplogroup I are differentiated from each other.

This does not apply to E-V13 which doesn't have such differentiated clusters. All E-V13's in Europe belong to the single star phylogeny, not to any differentiated groups, and certainly not to any differentiated groups centered on different countries.

John said...

>>> "There are probably three sources of E-V13 north of Greece: Epirotic E-V13 (E-V13 is high in Epirus, the region identified by Aristotle as ancient Hellas),"

You sir are WRONG. E-V13 is mostly NORTH of Albania. Search for "Steven Bird E-V13" and you will see the map. The Kosovars were thrown out of Kosovo by Dusan and took to the mountains. However, populations moved over the course of centuries, like the Suli from N Albania, Arvanites etc.

dienekesp said...

You sir are WRONG. E-V13 is mostly NORTH of Albania. Search for "Steven Bird E-V13" and you will see the map.

No need to rely on an outdated map that completely ignores E-V13 in Greece.

John said...

http://www.jogg.info/32/bird.pdf

Look on page 11, right side. Kosovo is the center of it and Steven suggests that this was spread to rest from the Kosovo /FYROM border. IIRC it was started at about 8000 years ago.

North of Albania is full of it. And no one can say that Arvanites 500-1000 years later are still in a specific area. Many people also lost the "Arvanite" connection centuries ago and simply don't know it.

FYI: Malcolm Noel places the Albanian start somewhere in Kosovo /North FYROM, where Ev-13 is oldest according to Bird.

Last point: If modern Greece is based on Orthodox religion (Orthodox stayed or came in; 'Turks' were thrown out), how can you be sure of who is Greek and who is, say an Orthodox Kurd? Were all those Anatolian Greeks, really Greek?

John said...

No need to rely on an outdated map that completely ignores E-V13 in Greece.

Actually he mentions and shows it, albeit in colors. From Arta and 'up' is 0.21 whereas in Kosovo is 0.43. The next round is 0.18 in Greece. In Malesia (North Albania) is about the same as Kosovo, 0.39. I agree with Kosovo's founder effect, many found refuge in Malesi and then left, but why the North specifically?

I find it odd that Greeks usually claim that S Albania is Greek (to get around the Arvanite /Hydra and Suli factor) but the North has the what you say is 'Greek.'

Just to show how people moved, the Himariotes are originally from the North as well (Kruja, Mirdita) and found refuge after the Turks devastated their area. Someone checked the Stradiotti names and most were old Catholic one (Gjoleka, Gjipal etc)

Maju said...

The Bird maps show "clades" for Greece but do not show sample sites in Greece, what suggests that they can be subject to the same criticism of undersampling and extrapolation Bird has for Cruciani.

If you check other references (for instance this post by Dienekes (where he gathers several separate Y-DNA in Greece - I drew this map after it), you'll see that Y-DNA DE (that for Greece is almost exclusively E-V13) is pretty high (>20%) in most locations of mainland Greece, as well as the Aegean islands, reaching levels similar to those found among Kosovars (c. 40%) in several of them, including Sesklo-Dimini, the epicenter of European Neolithic.

On light of this it's very clear that E-V13 is as much Greek as Albanian signature. Would Albanians have arrived from outside the Balcans, as some claim, it seems very odd that they would have incorporated so much local Y-DNA in the very specific form of E-V13, moreso when aculturation via "elite domination" doesn't seem likely in their case (no polities for the most part).

It also seems unlikely that it can be thought original of either Greeks or Albanians as the differentiated groups we know today: while the Kosovo case strongly speaks against a "recent" Greek origin, the Chios and Lesbos case does the same against an even more recent "Arvanite infiltration".

The best explanation again seems the Neolithic dispersal model. This model does not just account well for a strong founder effect in Kosovo and Albania but also explains the rather high rates found as far north in the continental route of European Neolithic difussion (Hungary, Slovakia). I am convinced that E-V13 carriers were a major component of the peoples that spread the Balcanic Neolithic, generally understood as originated at Sesklo (modern Greece).

In this sense, I wonder if there is a gradation within Albania between the highlands (of Balcanic Neolithic background) and the coast (Cardium Pottery Neolithic instead). The highlands should show, I understand, significatively more E-V13 than the coast - while the Bird map does suggest it, it's not sufficiently clear though.

Andrew said...

>I have not seen any evidence that E-V13 in the Balkans originated in Thrace. This hypothesis has more to do with a romantic fascination with the Thracians than with any actual evidence.

Well, I was only reacting to your statement that "no one seriously suggests that Albanians originated to the south or east of Greece". My point is only that there are no really strong theories about the subject. They certainly could come from closer to Constantinople, because Christians seem to have moved both to the remoter areas where Albanian is now considered at home, as well as the islands, many of which were under Venice.

>With that said, a possible origin of Albanians in Thrace solves none of the problems that I mentioned (sea terms and lack of Greek loan-words). Indeed, one would expect Albanian to have even more Greek loan words if it originated in Thrace, as Thrace was even more heavily colonized by Greeks than the Adriatic coast.

Depends which part of Thrace?

> Admixture doesn't always increase variance. The necessary condition is for the two sets of Y-chromosomes that mix to be differentiated.

> If, e.g. a bunch of Bosnian I men moves to Sweden, and you measured the variance of haplogroup I in Sweden before and after this move, then you would see an increase, the reason being that Bosnian and Swedish haplogroup I are differentiated from each other.

Isn't it also true that if someone says that a certain population show more diversity, that this is the same as saying that individuals in that population are more differentiated from each other?

> This does not apply to E-V13 which doesn't have such differentiated clusters. All E-V13's in Europe belong to the single star phylogeny, not to any differentiated groups, and certainly not to any differentiated groups centered on different countries.

I don't see how the lack of defined clades makes a difference. If all the V13 populations of Europe came to live in one small region, would that region not have more diversity per square km than any other? And is this not the only thing you are looking at to judge age?

Andrew said...

Concerning Italian and Balkan diversity, how would we explain the high levels in the Rimini area? See "Male haplotypes and haplogroups differences between urban (Rimini)
and rural area (Valmarecchia) in Romagna region (North Italy)"
Gianmarco Ferri et al. Forensic Science International 175 (2008) 250–255.

Andrew said...

>It also seems unlikely that it can be thought original of either Greeks or Albanians as the differentiated groups we know today: while the Kosovo case strongly speaks against a "recent" Greek origin, the Chios and Lesbos case does the same against an even more recent "Arvanite infiltration".

Just in case there is a misunderstanding, my posts were not intended to claim that V13 in Greece all comes from Albanian speakers. My whole focus is upon variations in diversity because this is what Dienekes is focusing upon. What I am saying is that even if there is unambiguous evidence that there is more diversity in Greece, this diversity can easily be caused by the sorts of population movements, decreases and increases, which we know happened in the Balkans during many centuries under at least 3 multinational empires.

OTOH whether there really is higher diversity in Greece is another question. It is always a question of statistics, and when different studies conclude different things we should be at least cautious (Perecic et al finds highest diversity NORTH of the Albanian speaking area.

dienekesp said...

Look on page 11, right side. Kosovo is the center of it and Steven suggests that this was spread to rest from the Kosovo /FYROM border.

The high frequency in Kosovo is due to a founder effect, no reason to believe that E-V13 in the Balkans spread from Kosovo.

Actually he mentions and shows it, albeit in colors. From Arta and 'up' is 0.21 whereas in Kosovo is 0.43. The next round is 0.18 in Greece.

Er, he misses the Peloponnese where there is strong evidence for a high frequency, some of which was already published at the time when he wrote the article.

FYI: Malcolm Noel places the Albanian start somewhere in Kosovo /North FYROM, where Ev-13 is oldest according to Bird.

There is no evidence that E-V13 is oldest in Kosovo or North FYROM. Actually that's a nonsensical inference if he really makes it, since no study of the geographical distribution of E-V13 within FYROM has been made. Taking synthetic maps too far...

Many people also lost the "Arvanite" connection centuries ago and simply don't know it.

We don't have to rely on who "remembers" their Arvanite past today. There are maps published 200 years ago, travel reports predating that, that make it clear which locations in Greece had Arvanites in them.

dienekesp said...

I find it odd that Greeks usually claim that S Albania is Greek (to get around the Arvanite /Hydra and Suli factor) but the North has the what you say is 'Greek.'

Northern Epirus is historically Greek. and there are doubtlessly many Toscs who are Albanized Epirotes. This has nothing to do with the frequency of one particular merker, which incidentally I never claimed was "Greek", but rather an extra-European Bronze Age founding lineage that proliferated among Greeks and spread from there.

dienekesp said...

Well, I was only reacting to your statement that "no one seriously suggests that Albanians originated to the south or east of Greece". My point is only that there are no really strong theories about the subject. They certainly could come from closer to Constantinople

Again, how does proximity to Byzantium (a Greek city) help one resolve the problem that Albanian lacks Greek loan words.

There may be no strong theories about Albanian origins, since Albanian as a language is attested very late, but the theory that the Albanians originated next to Constantinople can be safely rejected as their existence wouldn't have escaped notice of the many literate accounts we have of the ethnography of the region.

Depends which part of Thrace?

Not really, since there was no part of Thrace really outside of Greek influence.

I don't see how the lack of defined clades makes a difference.

It is not a matter of defined clades, it is a matter of differentiated clades, irrespective of whether they have been defined by SNPs yet or not.

It is always a question of statistics, and when different studies conclude different things we should be at least cautious (Perecic et al finds highest diversity NORTH of the Albanian speaking area.

That's inaccurate.

"In Europe, the highest E3b1a variance
is among Apulians, Greeks, and Macedonians, and the
highest frequency of the cluster is among Albanians, Macedonians,
and Greeks (table 1)."

Andrew said...

>how does proximity to Byzantium (a Greek city) help one resolve the problem that Albanian lacks Greek loan words.

It does not address it at all. The question is whether we are sure that Albanian V13 comes from areas of Greece known to have high V13 diversity. You said that no one believes Albanians could come from the East, which is an example of another direction you could get to Albania from areas closer to Asia. I just pointed out that this is not quite correct. They could have come from the East - perhaps via the north. Remember V13 had a long time to get to Albania.

>there was no part of Thrace really outside of Greek influence.

There was no part of the Balkans outside of Greek influence during some periods. Are you truly arguing that the Albanian language comes from outside all areas ever under Greek influence? And only based on loan words? Latin and Slavic also have many Greek loan words.

Genetic drift could certainly account for less diversity in Albania compared to Greece, but there are so many ways this could have happened - populations going up and down over centuries, and also emigration, an expanding dissipating population, not only immigration.

I also don't see how you've confronted my very simple question about whether concentrating any population into a smaller geographical area, will not always lead to that population looking "older" based upon normal measures of genetic diversity whenever we talk about a star shaped cluster like this.

Best Regards
Andrew

dienekesp said...

There was no part of the Balkans outside of Greek influence during some periods. Are you truly arguing that the Albanian language comes from outside all areas ever under Greek influence?

While pretty much most of the Balkans had some Greek influence, the degree of Greek influence is not the same.

Now, if the Albanians lived in their present-day location (northwest of Greece by the Adriatic) for a long time, their language would have both sea-related terms and ancient Greek loan words, since the Greeks were in that area for 3ky.

This problem would not be solved if one was simply to laterally move their origin to the east in Thrace, as the same problems would arise.

The location of the early Albanians is to be sought to the northeast of their present-day location, in the interior of the Balkan peninsula, somewhere where (i) they wouldn't be close to the sea, (ii) wouldn't be close to Greeks.

Genetic drift could certainly account for less diversity in Albania compared to Greece, but there are so many ways this could have happened - populations going up and down over centuries, and also emigration, an expanding dissipating population, not only immigration.

No, emigration can't realistically affect Y-STR variance. Read the post on how bottlenecks affect Y-STR variance. Emigration is equivalent to a bottleneck; the only difference is that the population doesn't die or fail to reproduce but moves out.

I also don't see how you've confronted my very simple question about whether concentrating any population into a smaller geographical area, will not always lead to that population looking "older" based upon normal measures of genetic diversity whenever we talk about a star shaped cluster like this.

Let a be an allele, let ma be the mean allele, let ma1,ma2 be the mean alleles of two subpopulations.

Variance is an average of terms of the form:

(a-ma)^2.

For an allele of one of the subpopulations (e.g. 1), this can be written as:

(a-ma1+ma1-ma)^2 =
(a-ma1)^2+(ma1-ma)^2+2*(a-ma1)*(ma1-ma) (Eq. 1)

(replace ma1 with ma2 for the second population)

The variances in the subpopulation is an average of terms of the form

(a-ma1)^2
(a-ma2)^2 (Eq. 2)

Now, if the two subpopulations are not differentiated, then for most loci ma1=ma2=ma. Hence, from (Eq. 1), the variance of the composite population will be an average of terms:

(a-ma1)^2 and (a-ma2)^2

If you do the math, and VAR, VAR1, VAR2 are the variance in the total, population 1, and population 2, then

VAR = f1*VAR1+f2*VAR2 (Eq. 3)

where f1,f2 are the fractions of the two populations in the composite population. From this it follows that VAR less_or_equal max(VAR1,VAR2).

In a star cluster ma1 approx_equal ma2 for most loci, hence the result of admixture will not increase the variance above that of the higher-variance population, but will result in an intermediate variance value (Eq.3)

If, on the other hand there is no star cluster but the two populations form their own clusters, then ma1 will be different from ma2 in several loci and the (ma1-ma)^2 term (Eq.1) won't be zero => increased variance.

Andrew said...

>Now, if the two subpopulations are not differentiated, then for most loci ma1=ma2=ma.

Thanks for taking your time on that. Please check my attempt to translate into English: for a truly star shaped cluster, the variance does not go up and down with the population?

I suppose that what I must really be arguing then, is that even V13 is not a perfect star shaped cluster. Surely there must be some clusters which would become more obvious if we had a better database.

I think however I can now see that measuring an age estimate from variance would not NECESSARILY go up if a single population were compounded from several closely related components.

dienekesp said...

Thanks for taking your time on that. Please check my attempt to translate into English: for a truly star shaped cluster, the variance does not go up and down with the population?

Different populations can have different variance, depending on when they started expanding and how fast. It's just that their clusters of haplotypes overlap with each other, like a small circle (younger population) within a larger circle (older population).

John said...

regarding variety:

Does it matter if the people married almost exclusively within a small group? The people from at least this region http://en.wikipedia.org/wiki/Mal%C3%ABsia only married among each other (but never within one's extended clan no matter how distanced)

We have to keep in mind that for generations that area was closed to the outsiders.

bob dok said...

V13 is Tracian.No doubt abouth that.Paleogenetic Romanian research on Tracian remains has come up with indirect proof in population of Italy(slaves)Isles(Tracians in Roman garnisons)Albanians and East Bulgarians which all come up with V13 on Y-testing.What is more V13 has been in close corelations in arival on the Balcan with Greek J2.

Dienekes said...

V13 is Tracian.No doubt abouth that.Paleogenetic Romanian research on Tracian remains

There is no "paleogenetic research" on Thracian remains that has discovered E-V13, indeed no paleogenetic evidence on E-V13 period.

John said...

Dienekes,
Albanians have 75% of mutation Delta F508, the same if not more than Northern Europeans (Greeks and Italians have less than 60%)

Google it and see what it may mean.

Kalu said...

i agree with bob-dok




documentaries

Marine Stone said...

Sorry, I am Greek and what I see from the map is that Albanians are older than Greeks, maybe they were what ancients called the Dorians. They came from nowhere else, they were always there.