November 05, 2009

Finally, structure in haplogroup R1a (Underhill et al. 2009)

I have lobbied for more structure in R1a to be discovered since the early days of this blog, and finally the R1a monolith seems to be cracking.

(Watch this space for my comments once I read the paper).

UPDATE (Nov 05):

This paper makes a very important contribution by studying the diversity and distribution of Y-chromosome marker M458 defining the new haplogroup R1a1a7.

R-M458 reaches high frequency and diversity in central and eastern Europe. It is virtually absent in northwestern Europe, the Near East, and Asia east of the Urals. The maximum frequency is reached in south (36.4%) and central (33.3%) Poland.

The earliest expansion time for R-M458 is found in Poland (10.7ky), but since the paper uses the effective mutation rate that I criticized elsewhere, this date should be divided by a factor of 3 giving an age of 3.6ky. This matches quite well the age for the Balto-Slavic split according to Gray and Atkinson. As with the recent paper on J-P58, adopting the germline rate makes excellent sense.

If R-M458 had started expanding 10.7ky ago, then by the time of the early dispersals of Kurgan groups east, it would have been present among them, and we would expect to find it east of the Urals and in the Near East/Central/South Asia. To reconcile this age with the archaeological picture of west-east movements across the steppe seems impossible. However, the situation resolves itself neatly when we realize that J-P58 is only 3-4 thousand years old, and was not in existence at the time of the Kurgan expansion.

However, the paper also is inconsistent with an origin of R-M17 either in a glacial refugium or with the expansion of the Kurgan culture, as the diversity of R1a1a*(xR1a1a7) is not particularly high in either Russia or the Ukraine and is much higher in India and Pakistan:
Analysis of associated STR diversity profiles revealed that among the R1a1a*(xM458) chromosomes the highest diversity is observed among populations of the Indus Valley yielding coalescent times above 14 KYA (thousands of years ago), whereas the R1a1a* diversity declines toward Europe where its maximum diversity and coalescent times of 11.2 KYA are observed in Poland, Slovakia and Crete.

Hopefully we will not have to wait another decade for the R1a1a* chromosomes to be further resolved, and thus yield signal(s) specific to the archaeologically attested trans-Ural spread of the Kurgan culture and/or the Indo-Iranians.

There are also other useful conclusions that we may draw from this study:

1. The presence of R1a in Scandinavia does not reflect any sort of Balto-Slavic admixture, as R-M458 is lacking in them. This, coupled with the discovery of R1a chromosomes in ancient remains from Eulau and Lichtenstein in German territory suggests that R1a was present among ancient Germanic speakers. However, there is also structure in R-M458 frequency in Germany itself, with lowest percentages reached in the north and west; this suggests an admixture with Baltic and/or Slavic elements in present-day Germany itself.

2. In several Balkan samples the frequency of R-M458 ranges between 0-12% which is at most a third of its maximum frequency, suggesting that the extent of Slavic admixture in the Balkans is upper-bounded by about 1/3. The complete absence of R-M458 in Italy and its sub-1% representation in Anatolia further support the idea that R-M458 in the Balkans is of medieval and later origins.

3. The samples from Greece (N=263), Macedonian Greeks (N=57), and Crete (N=361) have R-M458 frequency of 4.2, 8.8, and 2.2%. Assuming a frequency of R-M458 at 36.4% in ancestral Slavs, as in south Poland, the admixture estimates are 11.5, 24, 6%. These should probably be interpreted as upper limits (plus statistical margins) because the highest present-day frequency of R-M458 is probably lower than that in early Slavs.

4. Interestingly, the Slavs of FYROM have an R-M458 frequency of 3.8%, barely different from that of Greeks at large, suggesting that (i) the claims of some Greek nationalists that they Slavs of FYROM are newcomers to the Balkans are wrong, just as (ii) the claim of some FYROMian nationalists that they are markedly different from Greeks are wrong. The actual truth is that the Slavs of FYROM are largely of old Balkan (pre-Slavic) stock who adopted a non-Balkan Slavic language, just as the modern Turks are largely of old Anatolian (pre-Turkish) stock who adopted a non-Anatolian Turkic language.


European Journal of Human Genetics doi:10.1038/ejhg.2009.194

Separating the post-Glacial coancestry of European and Asian Y chromosomes within haplogroup R1a

Peter A Underhill et al.

Abstract

Human Y-chromosome haplogroup structure is largely circumscribed by continental boundaries. One notable exception to this general pattern is the young haplogroup R1a that exhibits post-Glacial coalescent times and relates the paternal ancestry of more than 10% of men in a wide geographic area extending from South Asia to Central East Europe and South Siberia. Its origin and dispersal patterns are poorly understood as no marker has yet been described that would distinguish European R1a chromosomes from Asian. Here we present frequency and haplotype diversity estimates for more than 2000 R1a chromosomes assessed for several newly discovered SNP markers that introduce the onset of informative R1a subdivisions by geography. Marker M434 has a low frequency and a late origin in West Asia bearing witness to recent gene flow over the Arabian Sea. Conversely, marker M458 has a significant frequency in Europe, exceeding 30% in its core area in Eastern Europe and comprising up to 70% of all M17 chromosomes present there. The diversity and frequency profiles of M458 suggest its origin during the early Holocene and a subsequent expansion likely related to a number of prehistoric cultural developments in the region. Its primary frequency and diversity distribution correlates well with some of the major Central and East European river basins where settled farming was established before its spread further eastward. Importantly, the virtual absence of M458 chromosomes outside Europe speaks against substantial patrilineal gene flow from East Europe to Asia, including to India, at least since the mid-Holocene.

Link

155 comments:

Gioiello said...

Then Indo-European languages are more ancient than many thought and the presence of R1b1b2 in Western Europe is more ancient than many, except me and a few others, are thinking.
I have said many times on this forum, and on others, that a very ancient R1a peopled India and Europe, and the same for R1b1* (in India R2 or others). But if the Indian R1a1 has no link with East European one, at least since the Middle Holocene, the presence of IE languages is older, as I have always theoricized. Good, very good! And I am waiting for your comments, being the paper a fee-paper.

Bryaxis said...

I'm a bit surprised at the mention of no patrilineal gene flow from eastern europe to India in the last 5000 years, I'd have though that the macedonian armies of Alexander and the veterans of Seleucide's colonies would have left a bigger mark in the north-western corner of India... With some thousand males soldiers without their wives in the area one might have though that the gene would have been more present. Or did the study take too small a group from that area ?

Ge3k0s said...

@Dienekes : Hi, I've knew your website since today and I ask myself : Are you Greek (and live in Greece) or are you from Greek origin (Greek-American ?) ?

In any case congratulations for your amazing blog.

Jean said...

"Then Indo-European languages are more ancient than many thought.."

I strongly suspect the usual problem with dating (3 times too old), so let's wait and see.

Jean said...

"mass of morons"

Love the neat alliteration, but morons - no. The authors haven't yet seen a good reason to stop following the herd on the dating system. I am rather surprised that they haven't in this particular case, because several papers have linked R1a1 with the spread of Indo-European languages.

But patience is a virtue. At some point enough aDNA will have built up that geneticists can use together with radiocarbon dates to provide a reality check on the theory-based calculations.

ashraf said...

indo-european languages are linked with J2 hg.
R1a peoples used to speak borean agglutinative languages but were assimilated by the proto-indoeuropeans.

Gioiello said...

Which is this, a Semite or specifically an Arab science?
Being Semite languages very recent(less than 6kyBP), probably I'll be able to demonstrate how they have of Indo-European. Semite roots presuppose roots shorter, with one or two consonants, that Indo-European have.

ashraf said...

lislakh(j1 proto semites+j2 proto indo-europeans)

http://www.nostratic.ru/books/(202)The%20implications%20of%20lislakh%20for%20nostratic.pdf

http://www.nostratic.ru/books/(88)prepositions.pdf

RJ said...

agglutinative? Wow! I love these new adjectives!

Maju said...

Very interesting. I'd love to find out more details.

For instance: does the sentence "Marker M434 has a low frequency and a late origin in West Asia bearing witness to recent gene flow over the Arabian Sea" means a flow from India to West Asia or the opposite?

The Aryan Invasion theory laid to rest?...

Not really because this theory is built on linguistic and archaeological data, rather than genetics (which are too young to really matter so much). What does alter radically is the perception that it had such a massive genetic impact in South Asia. South Asian R1a seems local and not of European origin.

I am rather glad about this because I always found very hard to explain how a bunch of nomadic raiders could cause such a massive impact in an agricultural society. At least for me it makes easier to explain Indoeuropean expansion (in India or elsewhere) as a political, cultural and linguistic process but in most cases not a demic one (exceptions would be the then ill populated areas of Baltic Europe).

Maju said...
This comment has been removed by the author.
Gioiello said...

Many thanks, Ashraf, for your elencation, but we Italians know these things from at least a century: Alfredo Trombetti, Elementi di glottologia, Bologna 1923 and the other monumental works. How ancient is your "shariba"? "Nel Georgiano si ha spesso per caratteristica un suffiso in -m, -b o -v, per es. su-: tema del presente swam- bere" p. 234. If you are a linguist you certainly know all the gradation of these root in Indo-European.

Anatole Klyosov said...

I have no doubt that the paper contains beautiful 2000 or so haplotypes along with absolutely wrong chronology and even more wrong conclusions.

Taking into account that Zhivotovsky always wrong with dating off by 300%, his "mid Holocene", that is about 6-7,000 years bp, will turn into some 2000-2500 years ago, that is WAY after the Aryan (R1a1) coming to India.

The thing is that the Russian R1a1 haplotypes are practically identical with the Indian "Indo-European" R1a1 haplotypes. The Russian R1a1 have a common ancestor 4800+/-500 ybp, while the Indian IE R1a1 have a common ancestor 4050+/-450 ybp. Their ancestral (base) haplotype is identical with the Russian R1a1 one in the 25-marker format, down to single allele. My own R1a1 67-matker haplotype is VERY close to a bunch of Indian R1a1 haplotypes, and we share the same branch on a 67-marker haplotype tree.

I can continue with linguistic as well, that the Russian language is pretty close (linguistically) with Sanskrit, along the the Lithuanian language which is also close to Sanskrit, that Russian language shares 28% of the basic 100-word Swadesh list with the Persian language (also IE), which gives 3500 years after the split, as well as that with Sanskrit, etc. etc., but what is the point? To ridicule one more paper with a wrong data? There were many, there will be more.

By the way, I said "IE R1a1", since there are another Indian R1a1, which came apparently from South Siberia between 8-12,000 ybp. They are NOT the "Indi-European" haplotypes.

RJ said...

There really is NO archaeological evidence to back up the AIT Maju. So now we'll just have a spate of linguistic scholar mafioso and their clueless cronies emptying out their dying supply of ammo before they change their minds.

Maju said...

A lot of people think that the archaeological evidence is there and is clear enough. There is even a "Vedic age" that shows the typical barbaric pattern of the usual IE Kurgan invasion, just like Corded Ware in Central Europe.

But the key issue is linguistic. Tell me: which is your explanation for Portuguese and Bengali being such similar languages? There is no other model than the Kurgan theory: Indocentric "sceptics" are so self-focused that do not even bother proposing an alternative model to discuss.

It is an ideological matter. For what I can see those "sceptics" are more or less like Christian literalists who enjoy looking for apparent contradictions (often not such) but are never able to propose anything that is both different and better.

terryt said...

"several papers have linked R1a1 with the spread of Indo-European languages".

I think that is the most likely connection as far as IE languages go. However, as Maju wrote, 'it makes easier to explain Indoeuropean expansion (in India or elsewhere) as a political, cultural and linguistic process'. So the relationship is not simple.

"if the Indian R1a1 has no link with East European one, at least since the Middle Holocene, the presence of IE languages is older".

Not necessarily so old. R1b is not associated with IE languages except in so far as many males with R1b Y-haps adopted the languages.

RJ said...

Let's let empirical evidence be the deciding factor Maju.

pconroy said...

Ronjoy,

You've commented here a number of times about your seeming belief that Indians are a relatively pure people/race with no European/Aryan input, and you are vehemently against the Indo-European languages being spread from Europe to India.

Yet your style of argumentation is simplistic, mostly just trying to shout other people down, but presenting very little in the way of facts or theories - just refutations. Why won't you propose an explanation for the current spread of Indo-European languages, who or what carried them from where to where??

RJ said...

I've never suggested that Indians are a 'pure people/race'. In fact my arguments are quite often directed at people who seem to have a notion that their 'race' is 'pure'. I also don't understand what you mean by 'European/Aryan' input - that's an oxymoron at best.

As far as the spread of IE languages goes, I'd prefer that more research be done before subscribing to any theory or proposing my own.

German Dziebel said...

"The Russian language is pretty close (linguistically) with Sanskrit, along the the Lithuanian language which is also close to Sanskrit."

They are all IE languages - true, but some of the more specific similarities between Slavic, Baltic and Sanskrit (e.g., palatalization) are not thought by Indo-Europeanists to result from common descent but rather from independent innovations or early areal influences.

"which is your explanation for Portuguese and Bengali being such similar languages?"

Tangential to the debate as it is, many living Indo-European languages would not have been considered "related" by standard comparativist criteria, if not for the underlying ancient languages preserved in written sources. French and Armenian are so vastly dissimilar (comp. Arm erku and French deux 'two', which are cognates) that only our knowledge of Old Armenian, Latin, etc. makes them part of the same family.

pconroy said...

Ronjoy,

But there has been a theory out there since the late 1700's that accounts for the spread of Indo-Europeans, and more evidence piles up all the time to support it...

The http://en.wikipedia.org/wiki/Andronovo_culture is derived from associated European cultures and the progenitor of the Aryan culture, what do you not understand about that.

Polak said...

M458 is obviously a Slavic marker. So the reason it's missing in Asia is the same reason it's largely absent from Western Europe; Slavs never migrated there in large numbers.

But Slavs weren't the proto-Indo-Europeans, who obviously carried markers ancestral to M458.

So the age estimates in this report are indeed a problem. They have to be divided by about three to match the most obvious explanation for the distribution and frequency of M458 around Europe.

Maju said...

Tangential to the debate as it is, many living Indo-European languages would not have been considered "related" by standard comparativist criteria, if not for the underlying ancient languages preserved in written sources. French and Armenian are so vastly dissimilar (comp. Arm erku and French deux 'two', which are cognates) that only our knowledge of Old Armenian, Latin, etc. makes them part of the same family.

I don't think you are right because we know virtually nothing of old Albanian or old Germanic or even old Celtic and yet all these languages groups are acknowledged beyond doubt within the IE superfamily, while it was quite difficult to prove descent of living Indo-Aryan languages from Sanskrit.

But, anyhow, it would only make the question to be re-formulated: what's the explanation for Latin, Gaelic, Albanian, Tocharian and Sanskrit to be so closely related?

Let's let empirical evidence be the deciding factor Maju.

Empirical evidence? Sure. But it needs to be put together in a theory. Otherwise is like watching apples fall from the tree and knowing nothing of the law of gravity.

Once you have the model based on empirical knowledge, you go again and test it against further empirical data. But without any model... it's like looking at the moon - you may even believe it's made of cheese.

In fact my arguments are quite often directed at people who seem to have a notion that their 'race' is 'pure'.

Actually that is the ideological issue behind this matter: when the Indoeuropean language family was discovered in the 19th century, Eurocentrism, colonialism and racism caused this to be viciously deformed into a theory of superiority of the Germanic peoples.

Now we know that Germanics have nothing to do with the original IEs. We know that the IE homeland was at the very fringes of Europe, in what used to be Asia until some Czar redraw the borders and that Southern/Central Asian IEs have almost nothing to do with European IEs (at least out of Eastern Europe).

But for Indians it's about the same: because India is the "holy land" and any bunch of barbarians that come from outside cannot be said to be "Aryans" that religiously is meant to mean "pure". Would they be from Congo or Mongolia it'd be the same.

That's why I say it's and ideological, religious, matter, rather than empirical. If Indocentrists could at least draw a model that made any sense... but they are too busy licking their colonial wounds and building the house beginning from some aprioristic ceiling.

At the moment, the Kurgan model is the only one that proposes a coherent scheme, that fits with archaeology and linguistics.

Sincerely, I'd be happier if IEs would have stayed at the Urals and never migrated anywhere. But reality is different.

Maju said...

But there has been a theory out there since the late 1700's that accounts for the spread of Indo-Europeans, and more evidence piles up all the time to support it...

No. The Kurgan model only has half a century. Pre-existent IE origin hypothesis were very confuse and quite irrelevant: at the time all what was known was that IE languages were philologically related.

eurologist said...

Now we know that Germanics have nothing to do with the original IEs. We know that the IE homeland was at the very fringes of Europe, in what used to be Asia until some Czar redraw the borders

Even that we don't know - it is still a theory. The fringes of Europe are not the center of IE diversity - central eastern Europe is, more so if you look about 3,000 years back.

Of course IE must have come to India from the west, and the Kurgan model (in its non-invasive, forget the horsies, cultural variation) seems a very viable one. But that absolutely does not mean it is also correct for Europe.

Quite the contrary - IMO, IE is way too old for that, the Kurgan model only applies to the Satem languages, and the Kurgans originally were receptors of IE from the west.

About the estimated ages, we will likely see more surprises in the future. I don't think they are off...

Maju said...

Even that we don't know - it is still a theory.

Everything is "a theory" in science. Newton's physics is a theory, but one that works pretty well (even today for most purposes).

In science, a theory is something quite solid and proven. What we call "theory" in colloquial language would be better described as "hypothesis", "conjecture" or "speculation" - which are different from a scientific theory.

and the Kurgan model (in its non-invasive, forget the horsies, cultural variation) seems a very viable one. But that absolutely does not mean it is also correct for Europe.

Actually the Kurgan model works best (it has been more throughly studied and tested) in Europe. If the Kurgan model would not work even for Europe, then I'd give up on it.

the Kurgan model only applies to the Satem languages.

Satem/Centum is not a valid linguistic division. Satem is just a "Scythian" pronunciation fashion and has little or nothing to do with linguistic phylogeny. Balto-Slavic is satem but it's first and foremost Western IE (all others are Centum).

Polak said...

^ The ages are way off.

R1a1a7 corresponds to the N and P clusters in the Polish R1a1 project. These are Slavic and Polish respectively, and quite young.

The likely ancestral cluster, K, does not correspond to this new R1a1a7. And this is the cluster we should be looking at when trying to track proto-Indo-European migrations.

Polak said...

Here...

http://www.gwozdz.org/PolishClades.html

ashraf said...

First attested indo-european speaking folks are:
1/Anatolian Hittites
2/Indo-Aryan mittanis of east anatolia
3/Mycenean Greeks

All these folks rose in area where the dominant haplotype is J2,also according to their self depictions these folks were of the mediterranean oriental phenotype and not of the nordic slavic one,also they shared many grammatical and lexical items(and even deities and mythologies)with their southern afrasian neigbours.

You could look to Dolgopolsky's nostratic dictionnary where he lists some 3000 common nostratic roots.

There are even some roots that are shared by semitic and some indo-european languages but not by all indo-european languages.(example:semitic/ie
ardh/earth,abalu/apple,ishtar/star,khom/home,qart/grad,aku/egho,atta/tu,ma,moist.....)

http://www.dspace.cam.ac.uk/bitstream/1810/196512/49/00ND_ALL.pdf

Gioiello said...

The fact that "qart" is from IE *ghort-" I think that nobody doubts. Cortona, Cortina, Cordova, Gortyna etc. are overall diffused in the European world and probably before that something like a Semitic language has arisen. I hypothesized that *Cort- was the form of the Proto-Proto-IE, like the Rhaetic-Etruscan. In Latin we have "hortus" but also "curtis", etc. etc.

ashraf said...

It couldnt be this way because afrasian is 10 k years phylum(according to Diakonoff)whereas ie is much more younger(8-6 k years).

Also first attested towns were built in fertile crescent and not in anatolia(as a homeland of ie languages).

AWood said...

"All these folks rose in area where the dominant haplotype is J2,also according to their self depictions these folks were of the mediterranean oriental phenotype and not of the nordic slavic one,also they shared many grammatical and lexical items(and even deities and mythologies)with their southern afrasian neigbours."

Actually all depictions and descriptions of Greeks are post-Dark Ages, or from Minoan pottery which, as we know, were most definitely different people. There is Homer's work which many on this board would say is "Nordic hogwash", but I still feel the Greeks were multi-ethic, especially early on. Greeks were comprised of many different ethnic groups, and various groups held power at different stages.

Greek and Albanian, as well as the western Anatolian languages are also the oldest IE languages, thus I feel the root starts there, possibly considerably earlier than predicted. J2a4, or J2b may indeed be common in Greece but the former is most likely a result of interaction with Semitic speaking people who expanded out of various Mesopotamian empires over the last 4,000 years. The earlier people, likely R1b (possibly R1a too?) and E3b date back to the Neolithic or earlier.

Anonymous said...

Afrasian phylum maybe has 15k years...

http://img396.imageshack.us/img396/7186/cushiticstratigraphy.jpg

German Dziebel said...

"we know virtually nothing of old Albanian or old Germanic or even old Celtic and yet all these languages groups are acknowledged beyond doubt within the IE superfamily, while it was quite difficult to prove descent of living Indo-Aryan languages from Sanskrit.

But, anyhow, it would only make the question to be re-formulated: what's the explanation for Latin, Gaelic, Albanian, Tocharian and Sanskrit to be so closely related?"

Indo-European languages are undoubtedly related but some of them are so divergent from others that only the privilege of having ancient written sources allowed for the establishment of "kinship" between them. I don't know what Indo-Aryan languages you have in mind, but, as for Albanian, sound correspondences with proto-Indo-European are still very problematic for this language (there're at least 4 ways in which PIE *s is reflected in Albanian in seemingly the same phonetic environment). Germanic and Celtic languages have plenty of ancient sources, not as old as Hittite, of course, but still their existence has made a comparativist's job much easier.

This is only a footnote to the discussion.

"Satem/Centum is not a valid linguistic division. Satem is just a "Scythian" pronunciation fashion and has little or nothing to do with linguistic phylogeny."

Yes, the centum/satem division, which refers to the absence vs. presence of special reflexes of PIE palatalized consonants (Sanskrit sata/Avestan satem vs. Greek kata/Latin centum 'hundred'), is an areal grouping and doesn't represent genetic subgroups within Indo-European. "Western" languages tend to be centum, while Eastern languages tend to be satem, with the exception of the easternmost Tocharian language, which is centum, and Armenian and Albanian, for which evidence is controversial.

One of the interesting features of IE languages is the fact that very few phonetic commonalities observed between branches stem from common descent. Mostly, these commonalities are the result of either convergence or diffusion. Hence, we don't know much about the real branching structure of Indo-European. See Garrett, "Convergence in the formation of Indo-European subgroups: Phylogeny and chronology", in Phylogenetic methods and the prehistory of languages, ed. by Peter Forster and Colin Renfrew (Cambridge: McDonald Institute for Archaeological Research), pp. 139-151

AWood said...

"It couldnt be this way because afrasian is 10 k years phylum(according to Diakonoff)whereas ie is much more younger(8-6 k years)."

taken from lowly Wikipedia:
"The earliest attestations of a Semitic language are in Akkadian, dating to ca. the 23rd century BC (see Sargon of Akkad) and Eblaite, but earlier evidence of Akkadian comes from personal names in Sumerian texts. Researchers in Egypt also claim to have discovered Canaanite snake spells that "date from between 2400 to 3000 B.C."

Good golly...it's the empire over time known as Akkadia/Assyria/Babylon...who spawned the Egyptians, Semitic tribes and the powerhouse Mesopotamia..... So how does this group have anything to do with the Aegean? An influence and some spread yes...but language? Not so sure.

Anonymous said...

what unites IE as a valid language family? sorry for my bad english.I think I understandable. :)

Dienekes said...

Actually all depictions and descriptions of Greeks are post-Dark Ages, or from Minoan pottery which, as we know, were most definitely different people.

That is inaccurate, there are plenty of depictions of humans in Mycenaean iconography.

Anonymous said...

Ashraf : "also they shared many grammatical and lexical items(and even deities and mythologies)with their southern afrasian neigbours."

You're referring to the major thunder deity Indra/Thor/Perun/Teshub (etc...) fighting the giant snake, that is comparable to Seth (Main god of Nubt, before 3,000 BC - north of south Egypt) fighting the giant snake Apopis (reminds of vritra, jörmungandr, zmey, Illuyanka, etc...) that the Hyksos and the Persians perceived as their god (hence Set being nickanmed the god of foreigners, despite being a major ancient Egyptian God) and probably to the semitic Hadad (-> Baal), right ?

Gioiello said...

German writes: “Indo-European languages are undoubtedly related but some of them are so divergent from others that only the privilege of having ancient written sources allowed for the establishment of "kinship" between them”.

But, whilst Semitic languages are very similar then very young, Indo-European ones are very different, then very old. Not only Hittite has separated at least 6000YBP (I think much earlier), but also Albanian, with its h4 (the laryngeal of “hapet” = opened) has separated before all the other languages. It is also clear, I think, that the most ancient stratum of European names (rivers, etc.) presuppose a previous phase of IE, till what I name the Rhaetian-Etruscan. Then Indo-European is rooted in Paleolithic Europe. Semitic, by the most ancient phase of Camitic, in Africa. Then in another part of the world (and I am one who have always supported the origin of hg. E in Middle East and not in Africa).

Anonymous said...

What do you think about IE's subgroupings, and more specifically, about Indo-Iranian's subgroupings (I believe that II is most internally diverse IE's living branch).

German Dziebel said...

"Not only Hittite has separated at least 6000YBP (I think much earlier), but also Albanian, with its h4 (the laryngeal of “hapet” = opened) has separated before all the other languages."

I would add "Tocharian" but would caution against interpreting all "h" attested in extant languages as the (only) direct descendants of ancient "laryngeals." The same can be said of "s" as a marker of "satem" languages (French cent 'hundred' has s- but this s- has no historical relationship to s- in Sanskrit sata). Armenian commonly has h in "laryngeal" places (e.g., hav 'grandfather', Lat avus, Hittite huhhas) but it can just as easily be a secondary development. Albanian does show unexpected h (as in vjeher 'father-in-law' next to Sanskrit svasura) but it's relationship to laryngeals is unclear. It does display a tripartite system of stops (similar to the one attributed to PIE by some linguists) but again it may not be a retention from PIE but a later independent development. A Pamir language, Yazgulyam, has the same tripartite system totally as a result of a relatively recent historical accident.

pconroy said...

Ashraf says:
First attested indo-european speaking folks are:
1/Anatolian Hittites
2/Indo-Aryan mittanis of east anatolia
3/Mycenean Greeks

All these folks rose in area where the dominant haplotype is J2


This is incorrect!

The Hittites were an Indo-European speaking group that migrated/invaded south and took over the Hattic Empire.

The Mittani were an Indo-European speaking group that migrated/invaded south and took over the Hurrian Empire.

The Myceneans are a little more complicated, with a few theories of where they came from, but also a North to South invasion/migration.

So none of them originated in the areas they were first attested.

Gioiello said...

German, of course French “cent” presuppose Latin “centum (kentum)”, then it is secondary, but laryngeals in Albanian are very rooted and I think sure. But there is other: in a paper (not published: then I was the unique follower of Alfredo Trombetti in Italy and perhaps in the world) of more than 30 years ago, I demonstrated that IE *swe- presuppose *dhle-, and Albanian has something of this ancient phase. For this IE *sweks “6” I linked with Sino-Tibetan *druk, Chinese “liu”, Cantonese “le”. This explains Albanese gj- where other IE languages have s-.

AWood said...

"That is inaccurate, there are plenty of depictions of humans in Mycenaean iconography."

There is some but not a lot, at least not where you can see clear phenotypes. Perhaps you have a source where I could take a gander.

AWood said...

"The Hittites were an Indo-European speaking group that migrated/invaded south and took over the Hattic Empire."

Many people say that, but I don't see why they couldn't have just been a local group of Aegean people? I don't see a need for them to migrate vast distances.

pconroy said...

Aaron,

But think about this for a second - did the Hittites develop their horsemanship on some island in the Aegean - where horses are not native - or was it from the Steppes, where horsemanship first arose, and also probably the famous chariot??

Though the latter may have spread have been an adaption of a 4-wheeled vehicle from Sumeria, that spread to the Steppes.

Unknown said...

Dienekes got some company in Ashraf,

One is a Greek nationalist and the other seems to be newly educated Pakistani or middle eastern. Not sure he tested his haplogroup or not.

Before discussing the origin and divergence of IE languages
Dienekes and Ashraf's agenda is to push shamelessly that J2 is the source of knowledge. One is greek and other is Muslim fanatic.

Archeology, history or genetics doesn't matter.

German Dziebel said...

Gioiello, I would be interested in reading your paper. There're some puzzles in IE historical phonology that need to be resolved, and I'm open to ideas. For some of my ideas, see http://kinshipstudies.org/?page_id=6

German Dziebel said...

Gioiello, by the way Hittite shows an interesting correspondence of s to IE *d (as in siu 'god', Skrt dyavus 'same'), which is reminiscent of Albanian gj ~ IE *s (gjashte ~ *s(w)eks 'six'.

Gioiello said...

German, my paper was written in Italian and I have said you the essential. I have a letter of that period of the great Italian linguist Vittore Pisani, who said me he admired my knowledge but he hadn’t my faith in the monogenesis of the language. After I did other: I earned a living by teaching in the High School, I published books of poetry and criticism, but I am more and more convinced to have been then in the right. A few days ago I have ordered from Amazon “The Invention of the Jewish People” (Eikh u-matai humtza ha-am ha-yehudi?) ( I have been banned from two forums for supporting this) and “On the Relations between Chinese and the Indo-European Languages”. Perhaps I’ll take again my researches. In the meanwhile Russians (from Illich Svytich to Starostin etc.) and Americans have done many progresses. First for me it would be important to see how Trombetti had already discovered. For instance he had already understood the theory of Tesnière and of the Structural Grammar.

German Dziebel said...

Gioiello, I read Italian and I've read some stuff by Pisani in the original. IE numerals and kin terms, although complex morphologically, lack etymologies, which makes me think that there may be some important sound laws that are still missing from our IE portfolio. Once these additional laws are discovered, connections with other language families may be in order. I'm very critical of Nostratic/Eurasiatic and other superfammilies but I don't doubt that IE is related to other language families.

Gioiello said...

About Hittite “siu” I think it is probably a assibilation of t/d before –i-, diffused all over the world: Ancient Green “filusin” from “filonti-n”, Finnish etc. etc. Nothing to do probably with my theory, where s- developed from dh- and –w- from -l- ( see Polish “l”). Your difficulties I think are due to the fact that only a large comparison can give us the knowledge, for instance Caucasian Languages in the researches of Starostin. There was a “w” and a “j” that affected consonants and played a role of vowels or semivowels. Being you a Russian, you know well the role of palatalization.

Gioiello said...

Another element of my theory was the link between -n- and -i-: from Trombetti: Bantu "tjantjatu" to "sheshatu" from "sha-i-shatu". See Albanian "muaj" = month etc.

Maju said...

what unites IE as a valid language family?-

Nice question. You'd need to read some books or at least articles to grasp the answer. What is clear is that philologists have a massive consensus on this matter and that it's been going on for more than a century now.

What do you think about IE's subgroupings, and more specifically, about Indo-Iranian's subgroupings (I believe that II is most internally diverse IE's living branch).

There are two main branches: Western and Eastern (Indoiranian). Both are diverse but I'd say Western IE is more, probably because it's older. II has two/three sub-branches: Indo-Aryan, Iranian and the diffuse group of minor Afghan languages, while Western has at least Germanic, Italic, Celtic and Balto-Slavic in the group. I'd say that the Western branch is more diverse but that the greatest historical diversity lies in neither group but towards the middle, where several languages that belong to none of this branch exist or have existed: around the Black Sea.

Dexippus said...

Dienekes, I am a bit perplexed by some of your conclusions:

"In several Balkan samples the frequency of R-M458 ranges between 0-12% which is at most a third of its maximum frequency, suggesting that the extent of Slavic admixture in the Balkans is upper-bounded by about 1/3. The complete absence of R-M458 in Italy and its sub-1% representation in Anatolia further support the idea that R-M458 in the Balkans is of medieval and later origins."

How does any of that prove that R-M458 is of medieval and later origins? Do the geneticists you've cited specify this?

"The samples from Greece (N=263), Macedonian Greeks (N=57), and Crete (N=361) have R-M458 frequency of 4.2, 8.8, and 2.2%. Assuming a frequency of R-M458 at 36.4% in ancestral Slavs, as in south Poland, the admixture estimates are 11.5, 24, 6%. These should probably be interpreted as upper limits (plus statistical margins) because the highest present-day frequency of R-M458 is probably lower than that in early Slavs."

Assuming that R-M458 is a medieval Slavic genetic marker, then wouldn't this conclusion technically contradict the anthropological evidence you've cited in one of your articles indicating a very limited input of Slavic genetic material in the overall Greek population?

Wouldn't this particular conclusion somewhat prove the weaker form of Fallmerayer's thesis regarding the biological composition of the Greeks (i.e. significant northern admixture)? If so, then how do you reconcile this study along with other genetic analyses (Y-chromosome, autosomal DNA, mtDNA) that you've presented in your articles indicating a very limited Slavic genetic impact on Greek populations?

"Interestingly, the Slavs of FYROM have an R-M458 frequency of 3.8%, barely different from that of Greeks at large, suggesting that (i) the claims of some Greek nationalists that they Slavs of FYROM are newcomers to the Balkans are wrong, just as (ii) the claim of some FYROMian nationalists that they are markedly different from Greeks are wrong. The actual truth is that the Slavs of FYROM are largely of old Balkan (pre-Slavic) stock who adopted a non-Balkan Slavic language, just as the modern Turks are largely of old Anatolian (pre-Turkish) stock who adopted a non-Anatolian Turkic language."

Even if the FYROMians are mainly of pre-Slavic stock, how can they have a lower R-M458 frequency than the Greeks if they were historically under the prolonged ethno-cultural influence of the Slavs (i.e. Bulgarians)?

I hope you can answer my questions.

All the best.

Ebizur said...

Gioiello said,

"For this IE *sweks “6” I linked with Sino-Tibetan *druk, Chinese “liu”, Cantonese “le”."

The Cantonese word for "six" is luk6 (pronounced in a manner similar to English "look"). "Le" is certainly an inaccurate transcription.

Other pronunciations of the morpheme that is represented in writing by the Chinese character 六:

Mandarin: liù
Sino-Korean: ryuk ~ yuk
Sino-Japanese: roku, riku
Sino-Thai: hok
Sino-Vietnamese: luc

As you may see, the "northern" reflexes of this etymon (Mandarin and Korean) exhibit a palatal segment that the "southern" reflexes (Thai and Vietnamese) lack. As usual, Japanese dictionaries reflect both the northern (what the Japanese call "kan-on," i.e. "Han pronunciation") and the southern ("go-on," i.e. "Wu pronunciation") forms, but only the southern form (/roku/) is used colloquially in Japan.

Dienekes said...

How does any of that prove that R-M458 is of medieval and later origins?

If R-M458 was of ancient origin in the Balkans, then it would have been present in Italy and Anatolia that were colonized from the Balkans.

Assuming that R-M458 is a medieval Slavic genetic marker, then wouldn't this conclusion technically contradict the anthropological evidence you've cited in one of your articles indicating a very limited input of Slavic genetic material in the overall Greek population?

No, because I never claimed that Greeks don't have any Slavic admixture. An upper limit of 1/10 for the general Greek population is not very difficult to imagine, and if as I suspect R-M458 was even higher in early Slavs (because all modern Slavs must have become admixed to some extent with pre-Slavic populations), then the true figure is probably lower still.

Even if the FYROMians are mainly of pre-Slavic stock, how can they have a lower R-M458 frequency than the Greeks if they were historically under the prolonged ethno-cultural influence of the Slavs (i.e. Bulgarians)?

They don't have a lower R-M458 frequency than the Greeks as the frequency difference is not significant. I really see no difficulty here; the early Slavs that settled in the Balkans had a high R-M458 frequency but were dissolved genetically in the pre-existing populations. North of Greece they were able to spread their language, while in Greece they weren't able to wipe out the Greek-speaking element which reasserted itself with time. The situation parallels that of the Romance speakers, with Latinity taking root in the more rustic north and Greekness persisting in the south, even though there were very few Italian settlers who were the agents of Latinization.

Dexippus said...

"If R-M458 was of ancient origin in the Balkans, then it would have been present in Italy and Anatolia that were colonized from the Balkans."

Sounds pretty logical. Thank you very much for clarifying that.

"No, because I never claimed that Greeks don't have any Slavic admixture. An upper limit of 1/10 for the general Greek population is not very difficult to imagine, and if as I suspect R-M458 was even higher in early Slavs (because all modern Slavs must have become admixed to some extent with pre-Slavic populations), then the true figure is probably lower still."

True, but your articles on the Greeks somewhat imply that the Slavs had such a negligible genetic impact on the Greeks that they might as well have been nonexistent. Such an implication does correlate quite nicely with information from the Oxford Dictionary of Byzantium and the archaeological record whereby the presence of "Slavic" wares and settlements in Greece are extremely limited (much to the chagrin of the ideological Pan-Slavists).

By the way, what does the upper limit of 1/10 for the general Greek population mean exactly? (1 out of 10 Greeks has Slavic ancestry, 10% of Greeks have Slavic ancestors, or that Slavic admixture makes up 0.1% of Greek genetic make-up?)

"They don't have a lower R-M458 frequency than the Greeks as the frequency difference is not significant. I really see no difficulty here; the early Slavs that settled in the Balkans had a high R-M458 frequency but were dissolved genetically in the pre-existing populations. North of Greece they were able to spread their language, while in Greece they weren't able to wipe out the Greek-speaking element which reasserted itself with time. The situation parallels that of the Romance speakers, with Latinity taking root in the more rustic north and Greekness persisting in the south, even though there were very few Italian settlers who were the agents of Latinization."

Makes sense, but isn't a 3.8% frequency in FYROMians mathematically lower than a 4.2% frequency in Greeks and an 8.8% frequency in Macedonian Greeks? Yes, you are absolutely correct in that the differences are negligible. But do the geneticists provide an explanation as to why the frequencies of R-M458 are lower in FYROM than in Greece if we know from history that the FYROMians were the ones who were linguistically and ethno-culturally "Slavicized" during the Middle Ages?

All the best.

Dean said...

As far as Greeks are concerned, the Slavic genetic contribution seems a little complicated. If it's accurate that 10% of Greeks are descended from Slavs, and a little over 5% of these Slavic descendants are R-M458 (per this study), who are the other 5%? Can we assume that these are people with haplogroup I? Vlachs could have brought some of what could be considered Slavic into Greece.

Vlachs could have brought some R-M458 into Greece. Even if these Vlach-descendants are in part descended from Slavs, it would be inaccurate to say that the R-M458 Vlach descendants were involved with Slavic invasions. I always want to know where the subjects are from. If they are from a big city, and their geographic background was not investigated, they could be lumped into the wrong group. If they are R-M458 and were from a Vlach part of Greece, that would skew the results, though perhaps by not that much. I still want to ensure as much accuracy as possible.

Maju said...

Makes sense, but isn't a 3.8% frequency in FYROMians mathematically lower than a 4.2% frequency in Greeks and an 8.8% frequency in Macedonian Greeks? ...

Good point.

Another issue I was pondering is that the frequencies for this lineage are pretty low in most of Ukraine and Belarus and instead rather high (though not as much as Poles, Czechs or Slovaks - or even some Germans) among ethnic Russians.

This seems inconsistent with a Slavic spread pattern.

I am not wholly excluding a partly "Slavic" spread of this lineage but I don't see clear that it fits with this pattern better than, say, early Indoeuropeans (at that time limited to Eastern, East-Central and East-SE Europe).

The core area of NE Central Europe certainly resembles the spread of Globular Amphorae, the culture that preceded Corded Ware. Instead the ethnic Russian spread appears so markedly ethnic (to the exclusion of even Ukrainians and Belorussians) that would seem to have spread just a few centuries ago, as Russia itself expanded.

Sincerely, no idea. But the lineage as such cannot be as young as only some 1300 years ago. Even if it has been spread partly by Slavic migrations, it needs to have been common before those in the core Slavic area (Poland?).

terryt said...

"First attested indo-european speaking folks are:
1/Anatolian Hittites
2/Indo-Aryan mittanis of east anatolia
3/Mycenean Greeks"

Expanding on Pconroy's comments: that's just the written languages. We have no idea of what was happening further afield (which goes for the Semitic languages you mention too). Besides which the three IE languages in question were already very different, probably indicating a long previous period of diversifation, and subsequently three separate 'migrations' into what were the adjacent geographic regions. The three separate routes from the north are easily imagined although perhaps difficult to prove.

"All these folks rose in area where the dominant haplotype is J2"

Perhaps. But we know that particular languages are not usually closely correlated with particular haplogroups. So the languages could well have been introduced by what is now a very minor haplogroup in the region.

"indoeuropean apollo/semitic baal
indoeuropean ahura-mazda/semitic atra hasis
indoeuropean zeus/semitic dhusara
indoeuropean diana/semitic tanit"

Religions and names for gods are even less reliably connected to haplogroups. In fact very often borrowed across cultures.

"By the way, what does the upper limit of 1/10 for the general Greek population mean exactly?"

I think we'd have to assume that the various independent groups that have been involved in the development of Greek-speaking people have become thoroughly (or nearly so) mixed by now. So 10% of the whole genome of all Greek-speaking people combined derives from groups originaly Slavic-speaking from the north. Individuals might today have anywhere from none to 100% from the north, although either extreme is pretty unlikely.

Gioiello said...

Ebisur says: “The Cantonese word for "six" is luk6 (pronounced in a manner similar to English "look"). "Le" is certainly an inaccurate transcription”.
I have in my Prima classe of “Liceo Scientifico” a Chinese boy who comes from Fu-Jian: he pronounces “le” (“e” as in French). I thought that the dialect was Cantonese, being Canton close to Fu-jian. Probably it isn’t true, but that some Chinese pronounce so be sure: I have heard it from a direct voice of one of them.

Gioiello said...

If R-M458 was of ancient origin in the Balkans, then it would have been present in Italy and Anatolia that were colonized from the Balkans.
What a surprise if we’ll ascertain that Balkans were colonized from Italy, at least in the first phase after the Younger Dryas!
The situation parallels that of the Romance speakers, with Latinity taking root in the more rustic north and Greekness persisting in the south, even though there were very few Italian settlers who were the agents of Latinization.
Rumanian, with 90% of Latin roots, wouldn’t have survived without a massive genetic component: Rumanians, after the defeat and probably the annihilation of the previous population, derive from Italian and celt-romanized people of the Rhine Valley. Then they too had some mixture with Slavs and others. But how many genes had they given to Greeks, Albanians and Slavs by Vlachs and Aromunians?

Anonymous said...

Sorry I haven't read all the posts. So much! All for the Aryan haplogroup R1a. Hard to think Poles and Russians as Aryans.

Dienekes has a hate of Near Easterners, probably as he is Greek, and Greeks are not much different from Lebanese. Culturally Greeks are Middle Eastern. Their version of Christianity is Eastern and African, straight out of Egypt.

I think his dividing dating by three is stupid. Not logical at all. He only does that because J1 comes out older than he likes, and indicates it proceeded the more Whitebread haplogroups into Europe barring haplogroup I, which is of east Anatolian in origin. We all know the R group is Asian in origin, specificially Central Asian/Siberian. The entry of the R group into Europe mostly after the LGM. R1a entered with the I.E languages. The Proto I.E language is 10 ky old at a stretch, probably younger, and entered Europe from the east in the Bronze Age along with weapons. It is in the archeological records in Europe. A peaceful group of farmers from the Neolithic with no weapons, no fortifications followed by aggressive immigrants with weapons, fortified villages in the Bronze Age. Even Greek "history" says the Hellenes, an I.E speaking group from Asia with weapons and an aggressive attitude overwhelmed the peaceful Pelasgians and forced their language upon them. The Celtic speakers in the historical era pushed out Iberian speakers, and the later Roman occupation killed off what was left of Iberian. Spain and Portugal are good examples. For hundreds of years many Iberians spoke Andalusian Arabic before the reconquest of Iberia by Castilian speaking Christians. Languages change without much genetic changes occurring.

The age of the Polish R1a I have no qualms with, but it must be realised that 10 kya there were no Poles, there were hardly any native Europeans in the part of Europe, and the ancestors of those people whose language became Polish were in Asia 10 kya. They just brought their haplogroups with them with no "Polish" haplotypes left in Asia. Archeology and aDNA of various remains found in NE Europe have shown the male remains to be R1a. The remains are from about 5 kya. Obviously R1a entered Europe long after it formed in Asia.

I am J1e, I am European. My haplotype, 67 markers, matches with one person, an American whose European ancestor comes from another country to mine and spoke a different language to mine. This Yank is 63/67 match. According to FTDNA he is related to me from 12 to 24 generations. I have checked my paternal line back fifteen generations, and my line comes from the one country and my surname is mentioned back in the 1400s. It is unlikely that the American is related to me for many generations more than 24, in the pre surname period of Europe. The Yank has a version of my surname applicable in his ancestral country. Using Ysearch, the closest to my haplotype is a Spaniard, genetic distance of 25 in 67 markers. The furthest from me is a person with a Jewish surname who has a g.d of 55 out of 67. I am closer to other J1 subclades, even some subclades of J2 than this Jew with the same haplogroup as me, J1e. I only use 67 markers to work out age, anything else, less than 37 markers, is just wasting time. When you check out whether your son is your son, I would go for 67 markers particularly if you are common R1b.

Dating haplogroups is not a science, just based on statistics. It may not be accurate but it is what the experts use. Dienekes is not an expert or a geneticist. Using prejudice, hate of Middle Easterners, is not a way to age haplogroup.

Gioiello said...

First of all we are guests of Dienekes and, please, we cannot agree with him, but a due respect!
It would be very heavy if your aggressiveness was due to your "semitic" haplogroup! Many of us would be justified ad abundantiam.
You are clearly of Italian descent, also if it dislikes you, whichever is the early origin of J1e, but probably your haplotype is in Italy from many thousands of years, certainly before someone used the name "Semite".
If your son/daughter matches you on 84% in the 23&me sharing, he/she is your son/daughter, without 67 markers: other Europeans match you on about 74%, Chinese on 71% and an Yoruba at 68%. All Ashkenazim are European, though they are within 74,10 with me. I'd be curious to test Ahmadinejad, probably a true Jew. Iranians are among 70,00 and 70,10%.

Polak said...

Here's the write up about this on my blog. Very similar to what Dienekes' has done here I guess, but we seem to agree on this issue so it's to be expected.

http://polishgenes.blogspot.com/2009/11/r1a1a7-signal-of-slavic-expansions-from.html

Further work with R1a1 and ancient DNA, as well as improved haplogroup age estimates will fully back up what I wrote, to the letter.

eurologist said...

Actually the Kurgan model works best (it has been more throughly studied and tested) in Europe. If the Kurgan model would not work even for Europe, then I'd give up on it.

Then so be it - people should start preparing for that...

Satem/Centum is not a valid linguistic division. Satem is just a "Scythian" pronunciation fashion and has little or nothing to do with linguistic phylogeny. Balto-Slavic is satem but it's first and foremost Western IE (all others are Centum).

Of course it is one of the valid distinctions (out of several), but likewise there is of course a reason for Slavic languages being Satem despite their proximity to Germanic.

Just forget about Baltic, Polish, (Northern) Russian, White Russian, Slovakian, Czech, Serb, even Bulgarian/"Macedonian" - all extremely recent (without ancient homes at all) compared to the Satem-Centum separation (or West-East, if that suits you better). Once you go back to Slavic <= Eastern Ukraine, it is clear that the relation to the other Satem languages is originally ~3,000 - 5,000 years older than the millenia of mixing with Germanic and Baltic that characterizes most of the current Slavic languages.

Again, don't look at the current Slavic languages as being the source of anything - they are a product. Yet, IMO, whatever moved East and then to Iran, Afghanistan, and India clearly originated fairly close to the proto-Slavic homeland, and like the proto-Slavic, got its language originally from farther West.

Once you dial back things a couple of millenia, IE idiosyncrasies of the Caucasus/Black Sea area are easier to understand as a consequence, not a source.

Unknown said...

All this dating stuff is getting very distracting. Nothing can be discussed without one group arguing the dates are too early or late.

Dienekes, can you please provide a summary document stating exactly what you think the recalculated dates for All the Haplogroups should be, with a timeline and correlation to known archaeology. If you are right you should be right for the whole picture and not just a sub part.

eurologist said...

what unites IE as a valid language family?

As others have said, there are many text books written on this.

Let me explain it to you in relative and quite anecdotal terms.

Someone growing up in Egypt can readily tell that Hebraic is very, very similar, without much work. That person, when spending a couple of weeks in Iran, can also figure out hundreds of Arab loan words, easily, but still will be at a total loss as to Farsi and the underlying language structure, in general.

Conversely, when one of my Iranian friends for fun asked me to identify a particular modifier, I immediately guessed right that it was a deminutive.

I grew up with one parent fluent in the old "Plattdeutsch", the other from the Bavarian/Austrian group, yet growing up right next to Denmark. To this day, I can read and pretty much fully understand Dutch (never learned it in school), and some Danish. Of course, I learned English in school, and then Latin, and later voluntarily took on French and Spanish.

After Latin, French was a strange language that seemed written in a weird Latin, with relatively little German/Gallic contribution other than the very "colorful" pronunciation. I can also read and understand a lot of Italian, but clearly, Spanish is a whole lot closer and much easier for me given its relation to the 2,000 years old Latin I learned in school.

All of them are familiar, with quite similar grammatical structures, and any deviations are not too surprising. Know one, know them all...

Coming to Greek and Slavic languages, it gets a bit more convoluted. With much work, I can recognize similarities to Germanic roots in Slavic about as often as in Gaelic or Greek, Yet, I can read Greek better, and understand more of it. There is something ancient about Slavic languages that goes beyond the loan words and Sprachbund and makes it much harder.


The same thing applies, even more so, to Iranian, the close Afghan, and then the Indian IE languages. However, they are just different in a different way, and I would not judge that solely from the Germanic/Roman closeness perspective.

Maju said...

@Eurologist:

Then so be it - people should start preparing for that...

Meh. I said it works well, wouldn't be the case...

Once you go back to Slavic <= Eastern Ukraine...

Last I read (from super-expert Polak) it was Western Ukraine. But AFAIK the issue is far from clear.

... it is clear that the relation to the other Satem languages is originally ~3,000 - 5,000 years older than the millenia of mixing with Germanic and Baltic that characterizes most of the current Slavic languages.

I find this very unorthodox (re. linguistics). Nearly all philologists consider Balto-Slavic a real subfamily and about the same number think that Balto-Slavic is part of Western Indoeuropean (probably originated with Corded Ware and maybe consolidated with the Urnfields Complex). Slavic satem is surely the effect of Scythians, Cimmerians and similar Eastern Indoeuropean groups in Eastern and Central-Eastern Europe.

...

@Polak: Nice info at your blog entry. However you have to admit that the given age estimates match quite closely with Corded Ware, regardless of whatever later movements associated to Slavic or Russian-specific expansion.

The objections you make re. Scandinavia can be explained via founder effects; after all IE migration into the North doesn't seem too intense - the influence of Pitted Ware seems much more important for those Eastern clades you mention.

Inversely it can be argued that the "blank" in Ukraine and Belarus clashes with the Slavic hypothesis. Also these areas would be the ultimate origin of Eastern lineages in Scandinavia if they arrived with Pitted Ware.

Vincent said...

Annie Mouse wrote: Dienekes, can you please provide a summary document stating exactly what you think the recalculated dates for All the Haplogroups should be, with a timeline and correlation to known archaeology?

This seems to be an overreaching request: asking Dienekes for a complete synthesis of Y-DNA population genetics, extended to cover unstudied haplogroups is a bit much.

Also, there is an embedded assumption that genetic history MUST correlate with archaeological history that is empirically undemonstrated and logically troubled.

But in the end as long as people use a consistent methodology the results they get will be consistent also. If Dienekes uses the same mutation rate to estimate the TMRCA for R1a that he uses to estimate the TMRCA for R1b, then the two TMRCA estimates will be consistent.

That a group of authors continue to publish TMRCA estimates that are too old by a factor of 2x to 3x is indeed frustrating. That there error is so blindingly obvious is even more frustrating.

Anonymous said...

@ Anatole Klyosov : "since there are another Indian R1a1, which came apparently from South Siberia between 8-12,000 ybp"

Can we have more informations about it, please ?

AWood said...

Interestingly, we still do not have an explanation how the Slavs blessed us inferior archaic humans in the rest of Europe with their beloved language. The homeland of the Hittites and Armenians is deficient of R1a and I2a2. Not only that, with M458+ believed to have come from north of the Black Sea, and at a much later time than R1a1 in general, it seems the rest of R1a1 really comes into question... because surely not every migrant into S. Europe would have been carrying M458+, or even been R1a1. We just know there is a presence of M458+ which is good enough for a movement of people. It does not give us an accurate statistic, especially when it peaks in Poland, which is not likely the source of immigrants. (rather Ukraine would have been)

Unfortunately the lack of R1a in Italo-Celtic languages also is a problem and a elite dominance model in Halstatt is not feasible...as Celtic was not derived from Slavic, even if Slavic speakers had entered the region at that time. (which is false anyways)

I would say another model needs to be used for the transfusion of language, probably earlier, and from a different source from R1a1. Or we can just agree that the oldest speakers (Hittite, Albanian, Greek) and the Centum branches were not R1a1. Take your pick.

Gioiello said...

Aaron writes: “I would say another model needs to be used for the transfusion of language, probably earlier, and from a different source from R1a1. Or we can just agree that the oldest speakers (Hittite, Albanian, Greek) and the Centum branches were not R1a1. Take your pick”.
I have always supported they were R1b1b2 and that the “transfugion” was “earlier”. If we watch to the first expansion from the Italian refugium, we hadn’t subclades behind R-L23+/L150+, and where we find this haplotype we have (or had) Indo-European languages. I think having demonstrate to someone who now is silent that Semitic languages took from Indo-European and not the contrary. Of course the expansion of Indo-Iranian is more recent and can have been carried by other haplogroups, certainly by the more recent R1a1 and not the previous one of 12000YBP present in India.

Dexippus said...

"Sincerely, no idea. But the lineage as such cannot be as young as only some 1300 years ago. Even if it has been spread partly by Slavic migrations, it needs to have been common before those in the core Slavic area (Poland?)."

That's actually a very good point Maju (never thought about it that way). But assuming that the R-M458 lineage is a 1300 year old Slavic marker, the frequencies supposedly present in Greeks seem a bit incredulous.

After looking at the map Polak provided at Polish Genes, I only saw two data points in Greece (one in Attica and one in Crete). The other two Balkan data points appear in south-central Albania and Vardarska (i.e. FYROM).

From what I can tell, the only Balkan data point surrounded by a visible color shade frequency of R-M458 is in south-central Albania (there is no data point in Bulgaria even though there is a visible color shade frequency of R-M458).

The near absence of a color shade frequency in the Greek data points appears to virtually coincide with the very sparse archaeological evidence involving Slavic settlements in Greece (the Oxford Dictionary of Byzantium makes this point abundantly clear despite medieval literary attestations depicting "Slavic tides").

All in all, whatever degree of Slavic admixture occurred in Greece during the Middle Ages was so limited that it was practically nonexistent (I think that in reality, the upper limit for Slavic admixture in the Greek population is much less than 1/10).

Of course, I'd avoid putting too much faith in genetic determinism given that genetics is still a young science and you just never know when another team of geneticists decides to "crack" the "R-M458 code" even further to discover that it may be something more than merely a Slavic genetic marker. Always keep an open mind (for what it's worth).

Maju said...

This seems to be an overreaching request: asking Dienekes for a complete synthesis of Y-DNA population genetics, extended to cover unstudied haplogroups is a bit much.

Obviously it is too much, but she has a lot of reason anyhow when asking for it. It is only when we can compare the overall claimed dates for all (or many) different genetic processes (in the various versions) with history or prehistory, when we can decide if they make sense or not.

Also, there is an embedded assumption that genetic history MUST correlate with archaeological history that is empirically undemonstrated and logically troubled.

A demographic expansion, such as the kind expected behind a haplogroup expansion should be traceable in the archaeological record, at least to a great extent. Certainly it should be the case in Europe where we enjoy the best accumulated archaeological research of all the planet.

Disconnecting genetic archaeology from material archaeology sounds to me as some sort of fanatic stubbornness, proper of religious freaks, not scienctifically-minded people.

Maju said...

Aaron:

Unfortunately the lack of R1a in Italo-Celtic languages also is a problem and a elite dominance model in Halstatt is not feasible...as Celtic was not derived from Slavic, even if Slavic speakers had entered the region at that time. (which is false anyways).

WTF?! Celtic is not derived from Slavic.

The low levels of R1a in West and SW Europe are obviously because the Indoeuropeans that invaded these areas in the Iron Age were already nothing like the original thing. Still, IMO, the local apportion of R1a, when compared with Poland or Russia makes for an easy estimation of the overall genuinely original male IE apportion in this area (i.e. if R1a is 2%, then the overall Kurgan-IE input is like 4%, always males only). However if you want to estimate the apportion of Urnifield/Hallstatt/La Tène IEs, then Austria or Germany are a better reference (the factor would be x4 or x5 instead of x2).

This may not strictly apply to places like Scandinavia because of the Pitted Ware complex migrations, which originated also in Eastern Europe.

Maju said...

Dexippus:

That's actually a very good point Maju (never thought about it that way). But assuming that the R-M458 lineage is a 1300 year old Slavic marker, the frequencies supposedly present in Greeks seem a bit incredulous.

Thanks. I agree that the apportion in Southern Macedonia being more than twice that of Northern Macedonia are rather odd and seem to ask of some other source, possibly older. There is also another local high in Albania (compare with Montenegro or even Serbia or Macedonia Republic) that seems to disagree with the Slavic hypothesis as well.

The details on the data ara freely available at the supplementary material. But you got the mapping points correctly, I believe (no data for Bulgaria for example). The R1a1a7 apportions for the Balcans are generally low (under 5%, often 0% or almost), excepting Southern Macedonia, Albania (but not Kosovo), some Dalmatian sites (Krk, Split), Bosnia (but lower among Bosnian Croats and Bosnian Serbs) and Romania.

This Balcanic pattern can hardly be associated with "Slavs". However I can't think of other processes either except diffuse ideas of local founder effects at different times.

I wonder if it's possible that Slavs and other invaders of the region could have been in fact low in R1a1a7, what would have made this clade behave locally as a receding one, finding refuge in the highlands like Bosnia or Albania, or "protected" coastal areas like Southern Macedonia and Krk.

Dean said...

"All in all, whatever degree of Slavic admixture occurred in Greece during the Middle Ages was so limited that it was practically nonexistent (I think that in reality, the upper limit for Slavic admixture in the Greek population is much less than 1/10)."

If this is the case, there is a probably a disproportion of Slavic influence to Slavic presence. Reasonably large parts of Greece still have Slavic-influenced toponyms, and even more of Greece had these toponyms in earlier maps. Toponyms seem to linger in Greece, for example, a few places might still be informally called by their Turkish names much after Turks were expelled and killed. One town is called Tsipiana (Turkish?), but its formal name is Nestani, which may or may not be a Slavic name.

We have to remember that if historical sources are accurate, Slavs were also expelled from large parts of Greece, so that could be a reason the Slavic invasions seem smaller. A bunch of them were reportedly moved to central Anatolia, but I have no idea as to what central Turks' genes look like.

Dienekes said...

If it's accurate that 10% of Greeks are descended from Slavs, and a little over 5% of these Slavic descendants are R-M458 (per this study), who are the other 5%? Can we assume that these are people with haplogroup I?

I-M423 seems to be represented among Slavs but lacking in Scandinavians and underrepresented in other Germanics. I suspect that the early Slavs were a mix of I-M423, R-M458, some underived R-M17, with a few other haplogroups thrown in.

pconroy said...

I think the maps here are relevant - look at Maykop (aka Maikop) as a possible ultimate source of R1b in Western Europe:

http://www.eupedia.com/europe/neolithic_europe_map.shtml

Dean said...

"I-M423 seems to be represented among Slavs but lacking in Scandinavians and underrepresented in other Germanics. I suspect that the early Slavs were a mix of I-M423, R-M458, some underived R-M17, with a few other haplogroups thrown in."

I read that there is a tenuous split between haplogroup I Dinaric North and Dinaric South. I think these lines have markers in common except for two. It seems hard to assign these clades to migrations unless something is known about people's geographical origins. Haplogroup I Dinaric North in a Greek could be mistaken to involve Slavs when it could have come from Vlachs, if it was found in people with ancestry from historically Vlach parts of Greece, like parts of northern Greece. This could happen if the Greek is from a large city, since this seems to be where most gene tests are done.

Unknown said...

I do not think it is too much to ask. I am thinking of a summary like the Europedia summary with Dieneke's theory. It is hardly a lifes work, or rocket science.

Personally I have doubts about the date calculations but not the same ones as Dienekes. The older dates match the known pre-History of the world better than Dieneke's theory of 1/3. And the whole Slav-Balkan-Greek argument may be the most politically skewed set of arguments in modern genetics. As it is, this debate is like watching two groups of people discussing two completely different books with exactly the same words.

The truth is from what I can see that the patterns observed are entirely consistent with with the older dates. A lot can happen in 6000 years, especially in the wake of glaciation.

Dienekes said...

Haplogroup I Dinaric North in a Greek could be mistaken to involve Slavs when it could have come from Vlachs

Yes, or even pre-Vlach populations. While I believe that the early Slavs had some "Dinaric" I, I do not consider it a Slavic marker since it probably did not originate in the regions usually considered Slavic homelands, but in the Balkans.

ashraf said...

mr Gioiello
R1b is very probably connected with vasco-caucasian languages.
R1b has its pick amongst Basks.
The Anatolian R1b is perhaps connected with caucasian speaking hurrians.(R1b is very commong amongst Armenians whose language have a caucasian substratum)

" I think having demonstrate to someone who now is silent that Semitic languages took from Indo-European and not the contrary"

What are those proofs as:
1/agriculture and towns were first attested in fertile crescent and south anatolia some 10-8 k years ago very distant(both spatially and temporaly)from your suggested ie homeland(only temporaly albeit if you accept the anatolian hypothesis)

2/first intrusions of ie speakers to fertile crescent are very late

3/these afrasian roots are attested in african branches of afrasian(berber,egyptian...)languages and have solid etymologies in afrasian whereas even today linguists could not give an acceptable ie etymology of many ie roots.

For example:ie septa/afrasian sab'a
afrasian sab'a came from the semitic root for the index finger(arabic sabbaba)

ie lick/afrasian lahs
afrasian lahs came from the afrasian root for tongue "lis"

arabic thuraya,berber ithri,akkadian ishtar,hebrew esther with ie star and you know that the fact that this root is present in berber means that it is at least a 7 k years old afrasian root(accordin to diakonoff dating of afrasian)so much before even the apparition of indo-europeans(if you assume that ie is 6 k years old)

you may read the topic I wrote here

http://dna-forums.org/index.php?showtopic=7715

ashraf said...

sorry I posted the wrong link
here is the correct link of the subject that I gave the title

Indo-Europeans from a linguistic point of view.
Is Indoeuropean a creole,pidgin,superstratum or sprachbund...???"

http://dna-forums.com/index.php?showtopic=7782

also you can look to this another topic
http://www.dna-forums.org/index.php?/topic/8079-possible-north-nostraticpre-proto-iesouth-nostraticpre-proto-semiticconnection/

Dean said...

"Yes, or even pre-Vlach populations. While I believe that the early Slavs had some "Dinaric" I, I do not consider it a Slavic marker since it probably did not originate in the regions usually considered Slavic homelands, but in the Balkans."

Sorry for drifting away from haplogroup R, but there are a good number of Polish people represented in a Dinaric North group via a major gene-testing company. So Dinaric North went north from the Balkans (in prehistory) but did not come back south with the Slavic invasions of the Balkans?

Dexippus said...

Dean:

"If this is the case, there is a probably a disproportion of Slavic influence to Slavic presence. Reasonably large parts of Greece still have Slavic-influenced toponyms, and even more of Greece had these toponyms in earlier maps. Toponyms seem to linger in Greece, for example, a few places might still be informally called by their Turkish names much after Turks were expelled and killed. One town is called Tsipiana (Turkish?), but its formal name is Nestani, which may or may not be a Slavic name."

According to Vacalopoulos, Slavic or Slavic-influenced toponyms are not very reliable indicators of Slavic settlement patterns. If anything, they could be attributed to other groups (i.e. Albanians and Vlachs) or, more realistically, attributed to tenants situated on monastic or lay estates. So despite Max Vasmer having "discovered" thousands of Slavic toponyms in Greece doesn't really mean that the Slavs actually settled in the territories they supposedly named given the very limited presence of actual Slavic settlements. And I don't think that the lack of physical evidence is attributed to some "Dark Age" phenomenon, because if the Slavs were dominant in Greece for 218 years (Chronicle of Monemvasia), then they would have had plenty of time to construct settlements and even statues commemorating their god Perun.

There is the strong possibility that the presence of "Slavic toponyms" in Greece is attributed to the movement of Slavic martial colonists used by the Byzantines as frontier defenders sometime during the early 800s AD (this is from the Oxford Dictionary of Byzantium).

"We have to remember that if historical sources are accurate, Slavs were also expelled from large parts of Greece, so that could be a reason the Slavic invasions seem smaller. A bunch of them were reportedly moved to central Anatolia, but I have no idea as to what central Turks' genes look like."

Yes, both Emperors Constans II and Justinian II removed many Slavs from Greece. I believe, however, that the policies implemented by both emperors were mostly in response to Slavic raids rather than to Slavic settlers. We get the same kind of response from Emperor Tiberius when he sends an envoy to the Avars to attack the Slavs who were plundering Greece while Tiberius was preoccupied fighting the Persians in the east.

Dexippus said...

Having looked at the supplementary details provided by Maju, I am beginning to think that there is something peculiar with this particular study.

The map provided by Polak shows a near virtual absence of R-M458 in Greece. The supplementary data, however, states that two locations in Greece have a 9.1% frequency of R-M458A (associated with R1a1a). Yet, Attica is the only place in Greece where a data point is visible on the map. Also, the map doesn't show Attica with a color shade frequency near the 10% margin. As for Greek Macedonia, there is no data point shown on the map even though the frequency of R-M458A is at 1.8%, which is very low and congruent with the color frequency closer to 0%. Now Crete is a bit perplexing since you have a single data point that shows no visible color shade frequency of R-M458A and yet the supplementary data shows a frequency range of 7.1-7.8%.

It appears that the R-M458G (associated with R1a1a7) frequencies that Dienekes cited also do not coincide with the map. Greece has a 4.2% R-M458G frequency, but the map shows a color shade frequency much closer to 0% than to 5%. Greek Macedonia has an 8.8% R-M458G frequency, but the map again shows a color shade frequency much closer to 0% than to 10%. Crete has R-M458G frequencies of 3.6 and 1.0 from two separate samples even though only one data point is shown on the map where the island is located. If the frequencies are taken separately, then the map does correctly show a color frequency closer to 0% in Crete.

I admit that I am not a geneticist and that I am not usually critical of things that are clearly outside of my field of study. I also admit that my assessment could be erroneous. However, I believe there are inconsistencies between the map and the supplementary data.

Maju said...

look at Maykop (aka Maikop) as a possible ultimate source of R1b in Western Europe.

Maikop?! You must be kidding or something. Maikop is not the origin of Kurgans but rather local culture without further consequences (except Hittites possibly).

Eupedia is not reliable.

...

... vasco-caucasian languages...

Such thing is non-existent (at best there are a couple of hypothesis, nothing to make such affirmation as if it was a fact). And there is no particular connection between R1b and the Caucasus.

...

The map provided by Polak shows a near virtual absence of R-M458 in Greece. The supplementary data, however, states that two locations in Greece have a 9.1% frequency of R-M458A (associated with R1a1a)...

You got that wrong, I think. You are giving the data for R1a1a*(xM458). R1a1a7-M458 is at the rightmost column.

Dexippus said...

"You got that wrong, I think. You are giving the data for R1a1a*(xM458). R1a1a7-M458 is at the rightmost column."

Maybe you're right Maju. But then again, I did notice on the supplementary data how R1a1a and R1a1a7 are associated with R-M458A and R-M458G haplotypes respectively. I view (perhaps erroneously) R-M458A and R-M458G are variants of the same R-M458 haplotype.

Nevertheless, comparing the R1a1a7-M458 frequencies in Greece with what was shown on the map reveals certain straightforward inconsistencies.

Do you find the data to be erroneous or the map to be erroneous Maju (or perhaps both)? Personally, I find the latter to be more helpful than the former since it is more congruent with archaeological evidence.

All the best.

Polak said...

The relative lack of M458 in the Balkans isn't a problem. The early Slavs weren't the same, especially late on, when they crossed the Danube.

So it's likely some population movements from the north carried a lot of M458, and others much less. And then you have to factor in such things as cultural/linguistic domination without genetic influence. The modern Balkans don't exactly look like Poland from a genetic or anthropological point of view, do they?

Btw, don't mind Aaron everyone. He's some guy from DNA Forums who's having a real hard time differentiating between the early Indo-European movements (ie. the Scythians), and those of the Slavs, which are obviously a few thousand years apart.

He's still grappling with the idea how it came to be that the 3.5 thousand year old M458 wasn't present in Europe 5 thousand years ago.

Maju said...

R-M458 is the new haplogroup: R1a1a7. Not sure what you mean with the rest but I feel you're confusing the relevant SNP.

Dexippus said...

Polak:

"The relative lack of M458 in the Balkans isn't a problem. The early Slavs weren't the same, especially late on, when they crossed the Danube."

That's true, but they would still have to carry enough M458 in order to leave behind a genetic imprint that according to the map appears visible in Romania and northeastern Bulgaria.

"So it's likely some population movements from the north carried a lot of M458, and others much less. And then you have to factor in such things as cultural/linguistic domination without genetic influence."

Also true. However, the Slavic invasions in Greece were geared more towards looting rather than vast settlement-building and or cultural/linguistic domination (behaviorally similar to the Goths and Vandals).

"The modern Balkans don't exactly look like Poland from a genetic or anthropological point of view, do they?"

No, which is why the M458 frequencies in Greece derived from the supplementary data seem a bit much. Further research and larger samples are needed.

Maju:

"R-M458 is the new haplogroup: R1a1a7. Not sure what you mean with the rest but I feel you're confusing the relevant SNP."

I see. Thank you very much for clarifying that Maju.

"I wonder if it's possible that Slavs and other invaders of the region could have been in fact low in R1a1a7, what would have made this clade behave locally as a receding one, finding refuge in the highlands like Bosnia or Albania, or "protected" coastal areas like Southern Macedonia and Krk."

Possible. But wouldn't your hypothesis involving Slavs carrying low frequencies of R1a1a7 possess other dominant non-Balkan haplotypes if we argue that they expanded from some Urheimat near the Danube River?

Dean said...

"According to Vacalopoulos, Slavic or Slavic-influenced toponyms are not very reliable indicators of Slavic settlement patterns. If anything, they could be attributed to other groups (i.e. Albanians and Vlachs"

You could be right, and I see how I erred in my assumptions that the name Tsipiana was Turkish when it could instead have an Albanian element, and Nestani could be an old Greek name when I assumed it may have been Slavic just because I read about it somewhere and unquestioningly believed it.

If I was a genetic scientist, I would want some knowledge of geographical settlement history to make better theories on genes and migrations. The lack of this makes it harder to quantify, in this case, what proportion of R-M458 in the southern Balkans was brought by historical Slavic migrations, especially in places that have complex settlement histories. One village in Greece that has this, from what I read, is Valtetsi; it's locally famous because of Greek revolutionary resistance against the Turks. It was allegedly founded by Himariote Greeks, settled by Souliotes, sojourned in by Tsakonians, etc. What, I wonder, is the origin of the name Valtetsi?

Maju said...

The relative lack of M458 in the Balkans isn't a problem. The early Slavs weren't the same, especially late on, when they crossed the Danube.

Fair enough. But...

... why then the apportions of M458 seem to favor non-Slavic peoples like Greeks or Albanians or even Romanians, rather than Slavic ones like Serbs, most Croats and Slovenians? Slovenians are high in R1a1a but very low in this new subclade.

... why is there a thick strip of low M458 between Poles and Russians? This is even more curious because almost all the same areas are also lower in R1a1a (a different situation than in SE Europe).

Also, according to your map, the lineage's expansion began in Central Poland, precisely where the core of Lubon-Globular Amphorae-Corded Ware was. Never read that Slavs originated that far North instead.

Overall the spread of this lineage is almost coincident with the >40% R1a1a area. The exceptions are Czechs and two sample points at the central strip between Poles and Russians. This may suggest that M458 took active part in the European spread of R1a1a as such. It probably expanded as part of the same movement, though with diverse (or even low) success in non-core areas.

Polak said...

There were different waves of Slavs into the Balkans, by different tribes. It's likely some Slavs ended up there directly from Poland, but most didn't. Studies on mtDNA lineages in Slovenians and Bosnians supports this scenario.

Also, IMO the proto-Slavic expansions didn't start in Central Poland, but in the area now shared by Poland, Belarus and Ukraine. Central Poland was very much affected by them though.

The area between Poland and Russia, relatively barren in M458, is interesting. That's a huge area though, and only two sites were sampled there. Let's wait for more results.

Same goes for Romania. One sampled site so far doesn't say much.

Polak said...

By the way, the frequency of M458 isn't everything. Look at its diversity in the little inset map.

The diversity pattern is about as Slavic as it gets.

Polak said...

Peter Gwozdz has updated his R1a1 page again.

http://www.gwozdz.org/PolishClades.html

It seems Poles are equally divided between R1a1a7* and R1a1a*.

It's interesting that he estimates the P cluster, which apparently falls within M458, to have undergone an expansion 1500 years ago, which would not be Slavic, but rather West Slavic.

Gioiello said...

Ashraf, living you in Istanbul, you certainly know Occhiai and if you send me your email address I’ll give you any explication on languages (except on computational linguistic where I think you are better than me). Dienekes doesn’t like that we speak on linguistic on his forum.

gioiello.tognoni@gmail.com

Maju said...

The area between Poland and Russia, relatively barren in M458, is interesting. That's a huge area though, and only two sites were sampled there. Let's wait for more results.

Actually there are 9 sampling sites (6 in Ukraine and 3 in Belarus) mentioned in this paper (see supplementary material) for that area. Some have similar frequencies of R1a1a7 to their neighbours (Poland or Russia) but the central areas (6 of the 9 sampled sites) have as low levels of this lineage as almost anywhere in the Balcans (under 10%).

An area named "Ukraine Central" looks particularly low in both R1a1a (under 20%) and R1a1a7 (under 10%) and I was wondering if they may represent a pre-Kurgan remnant of sorts. After all that was the core area of Dniepr-Don culture and a province that shows late Danubian (Baden-originated?) influences already when the Kurgan expansion process was in full swing (i.e. after Seredny-Stog-II), later falling into the Lubon (>>Corded) cultural area.

This is speculative but I have always felt that Baden and Lubon cultures were acting there (in the middle Dniepr) with the same general intent: to contain the Eastern Indoeuropeans of Catacombs Culture by establishing alliance with the locals. After all, early Western IEs were for some centuries in the cultural orbit of Baden, the last great Danubian culture and probably some sort of polity (tribal federation?)

By the way, the frequency of M458 isn't everything. Look at its diversity in the little inset map.

I have the feeling that the inset is not complete: the Balcans are blank and dots showing sample sites are few. In any case, Slavic and Eastern European are very much the same thing, right? Estonia for instance is not Slavic, nor is Central Germany (or hardly was in the past).

What the inset does suggest is expansion from Poland, specifically Central Poland.

What I do have some questions is about the ethnic Russian mid-high levels of this lineage. I can hardly fathom any process other than historical Russian expansion that makes sense with it. However Karelians and Vepsians are in the very same line. On the other hand, the apportions of overall R1a1a vary a lot among these Russians and Karelians.

Polak said...

I think the relatively empty area between Poland and Russia shows Uralic and Baltic influence, as well as the results of Ukraine's history. Large tracts of that country were depopulated, and repopulated since 1000AD.

The flow of M458 into Central Germany obviously went through the Czech Republic. On the other hand, Slavs lower in M458, from western Poland, settled the German Baltic Coast IMO.

We need more markers to come in to unravel all of this. M458 is indeed a Slavic marker, but Slavs were never 100% M458. Otherwise, why does Western Poland, the core area of the early Polish state, show up as so low in M458?

terryt said...

"especially when it peaks in Poland, which is not likely the source of immigrants. (rather Ukraine would have been)".

Easily explained in theory. The peak may have originally stretched from Poland to the Ukraine but the southeastern end has been diluted by the arrival of other haplogroups since.

"as Celtic was not derived from Slavic".

I doubt if anyone is claiming Celtic is derived from Slavic. However we seem to be all agreed they both have a common source. But both will have changed greatly since they parted company.

"A demographic expansion, such as the kind expected behind a haplogroup expansion should be traceable in the archaeological record, at least to a great extent. Certainly it should be the case in Europe where we enjoy the best accumulated archaeological research of all the planet".

That's what makes New Zealand prehistory so interesting. Archeology, genetics, linguistics, bird extinction and mythology all converge and tell the same story. I believe the processes revealed there should be able to help us interpret the evidence through the rest of the world too.

"The low levels of R1a in West and SW Europe are obviously because the Indoeuropeans that invaded these areas in the Iron Age were already nothing like the original thing".

i assume that is exactly correct.

Maju said...

It seems Poles are equally divided between R1a1a7* and R1a1a*.

Upon reading this, I took some time to compare through the paper's data, because the ratio between each clade can also be informative.

I see that with R1a1a7 dominant (markedly above 50% of R1a1a) there are the following regions: Czech Republic, Central and South Poland, Central Germany and... oddly enough Greek Macedonia.

Roughly 50-50 of each clade (being generous in favor of R1a1a7 - 30-50 would also be here) I see: some other Polish and German locations (Zamosc, Wroclaw, Poland North, Germany East and South), most Slovakia, most Ukraine, Brest (Belarus' border with Poland) and Kostroma (Russia).

Certainly if this lineage has anything to do with Slavs, it would mean an important Slavic founder effect in Greek Macedonia (only comparable to the also extreme Czech case) and large parts of Germany. But hardly in the southern Slavic peoples.

Maju said...

I think the relatively empty area between Poland and Russia shows Uralic and Baltic influence, as well as the results of Ukraine's history. Large tracts of that country were depopulated, and repopulated since 1000AD.

And Belarus too? AFAIK, the area around Kiev was always more or less stable. Of course, it's impossible to be 100% certain of these things but still I'm under the impression that the Middle Dniepr had particularly dense populations since Paleolithic (and clearly since Neolithic) and that this area was actually the source (or one of the sources) of some of the flows into the Baltic region (I have already mentioned the regressive Pitted Ware complex, that clearly shows DD cultural traits in burials and pots).

I am even beginning to think of Central Ukrainians as some sort of archetype or fossil population of pre-Kurgan Eastern Europe.

...

The peak may have originally stretched from Poland to the Ukraine but the southeastern end has been diluted by the arrival of other haplogroups since.

Like what? They are surrounded by high R1a1a populations, they should have received mostly that lineage.

I doubt if anyone is claiming Celtic is derived from Slavic.

Someone was: check the context.

Polak said...

There's a lot of Baltic influence in Belarus, so the lack of M458 there isn't surprising. And Ukraine is largely a steppe zone, prone to large population movements.

Btw, did anyone notice that Poland and Slovakia have by far the oldest R1a1 in that part of Europe?

So there goes any hope for the popular theories of expansions from the steppe zone. As I've said before, the steppe was a highway. Nothing originated there.

Maju said...

While most of Ukraine can be considered steppe it is the fertile version of it, specially towards the west. Ukraine has 58% of arable land, while a truly steppary country prone to such demic swaps like Khazakstan has less than 10% arable land.

The Black Lands of Ukraine were considered the granary of Eastern Europe not so long ago. Steppe just means grassland without trees and can be semi-desertic or very fertile.

And, anyhow, considering both latitude and actual prehistory, I don't see why Ukraine's genetics should be more altered by migrations than Russia or Poland. Rather the opposite.

Polak said...

Poland and Russia both have forests where it's easy to hide. Much of Ukraine doesn't. Actually, this might also be a reason why older elements survived the M458 Slavic expansions in Belarus.

Btw, I suspect the struggles over Ukraine between the Poles, Russians, Tatars and Ottomans had a massive impact on the demography in much of the country. And that's just one of the examples of the turmoil in that part of Eastern Europe. I've always had a hard time believing there was much population continuity there, and I maintain that psoition.

Maju said...

I suspect the struggles over Ukraine between the Poles, Russians, Tatars and Ottomans had a massive impact on the demography in much of the country.

But then you should expect them to look more like Poles and Russians, not as they do: different from both but not closer to Turks or Tatars either (talking genetics, of course). And again, what about Belarus?

Polak said...

Clearly there are older, pre-Slavic elements in much of the Belarus, largely thanks to the cover available there. This study talks about Slavic movements from Central Europe into those areas, but their effect did not wipe out the Uralic influences there.

http://eurogenes.blogspot.com/2009/10/inferring-uralic-admixture-in-eastern.html

The Ukraine is a mystery at this stage. It might always be a mystery, due to the frequent population movements there. Let's see what large scale genome-wide and Y-DNA studies show about the Ukrainians.

But the low diversity of R1a1 in both Ukraine and Russia don't hold out much hope of any major population movements originating there. It looks like R1a1 entered Europe just after the Ice Age from Asia, and then Poland was a major staging point for its subsequent expansions. The spread of M458 was one of the last of these major epxansions.

Maju said...

Just for comparison (indicated as R1a1a=R1a1a*+R1a1a7 in overall rounded percentage following the supp. mat.):

Ukraine Central: 18=11+7
[nothing to compare with]

Ukraine East (Belgorod): 46=41+5
Belarus Central: 42=38+4
[very similar but:]

Russia South (Belgorod): 62=50+12
[much more R1a1a7, in line with all other Russians, who seem standarized for this marker]

[and:]

Russia Central (Kostroma): 47=35+12
Poland Central: 56=32+24
Tatars: 39=35+5
Turkey West: 6=6+1

There's nothing comparable to the Ukraine Central sample and the most alike to Ukraine-Belgorod is Central Belarus (and Belarus was never conquered by Tatars nor Turks for any long period).

Maju said...

It looks like R1a1 entered Europe just after the Ice Age from Asia, and then Poland was a major staging point for its subsequent expansions. The spread of M458 was one of the last of these major epxansions.

I can agree with that in part. However at this moment I feel that M458 expanded together with most of the other R1a1a, as a particular subphenomenon inside a larger one.

Do you have any reason to consider R1a1a(xM458) to have a Polish origin as well?

Polak said...

I think Central Belarus might be influenced by pre-Slavic elements to a great extent. Also, the Dneprovskii region of Ukraine was settled by West Slavs very late on, as far as I know. That's what Polish Wikipedia reports anyway. :p

And like I mentioned above, the Polish STR cluster termed K seems ancestral to both N and P, which correlate well with M458. K might very well have a Polish origin, or it might not. But certainly it went west and north from or via Poland, probably with I1-P, which seems ancestral to I1 up in Scandinavia.

As I don't see the Ukraine or Russia as potential sites for the various R1a1 origins, then I think Poland is the best bet. The other area that comes to mind is the former East Germany, which it seems might be dominated by a local or older R1a1 type than M458. I expect this lineage to also be found in Western Poland, and it might be as Slavic as M458, and not necessarily Germanic.

Vincent said...

Annie Mouse wrote: I do not think it is too much to ask. I am thinking of a summary like the Europedia summary with Dieneke's theory. It is hardly a lifes work, or rocket science.

If Eupedia demonstrates anything it the opposite of your contention. That site is mess of error and speculation. A synthesis that is sloppy and conjectural is less than useless.

eurologist said...

I will comment more later, but don't forget that R1a has been found at roughly equal percentage with R1b in central-East Germany 2,000 years before Slavic expansion. So it is old, and things haven't changed much there, since.

Also, many regions of Poland and much of Eastern Europe in general were extremely sparsely settled since the beginning of agriculture. So much so, that even in historic times, local powers actively sought settlers from the west. Even the area just east of the Elbe river, in the north, was originally hardly settled and so devastated after the 30-year war that it was almost completely void of people. The entire region has always been more difficult to have steady, highly productive agriculture (before industrialization of agricultue).

Much of eastern Europe was originally settled by Baltic tribes, Finno-Ugrics, eastern Germanic tribes, and Scandinavian tribes. Slavs are very, very recent, in this context. However Slavs have changed the language, and have ensured a much larger fraction of R1a in the past 2,000 years. In that region, both are relatively new, though. Poland obviously cannot be a source of anything - it is a receptor.

Polak said...

So it's a receptor that has ancestral lineages for R1a1 and I1. A receptor from areas that don't have these ancestral lineages and lower diversity in the ones that they do have.

Doesn't make any sense. Sound like the usual cliched crap to me.

Polak said...

"Even the area just east of the Elbe river, in the north, was originally hardly settled and so devastated after the 30-year war that it was almost completely void of people."

What do the events in Prussia have anything to do with the areas that remained strongly ethnically Polish?

Plz explain to us how Central and Southeastern Poland was depopulated. That should be something I've never heard before. Can't wait.

Vincent said...

Maju wrote Disconnecting genetic archaeology from material archaeology sounds to me as some sort of fanatic stubbornness, proper of religious freaks, not scienctifically-minded people.

I'm not "disconnecting" anything. I'm saying that Annie Mouse's request seems to have had an embedded assumption that is unproven and, possibly, untrue. The assumption is that instances of genetic drift (aka change in allele frequency) must leave a correlative trace in the archaeological (or linguistic) record. It is precisely this assumption that have led Zhiv, Semino, and so many others to make the blunderous conclusions that they have made.

VV

Maju said...

As I don't see the Ukraine or Russia as potential sites for the various R1a1 origins, then I think Poland is the best bet. The other area that comes to mind is the former East Germany, which it seems might be dominated by a local or older R1a1 type than M458. I expect this lineage to also be found in Western Poland, and it might be as Slavic as M458, and not necessarily Germanic.

Germanic and Slavic are just notional words when applied to haplogroups and haplotypes, unless they really originated with these peoples and/or their main spread belongs to that period, what I doubt and seems very hard to prove.

Otherwise I find very interesting the mention of a distinct "East German" cluster within R1a1a, because this would fit with the genesis process of Western Indoeuropeans (Kurgan-derived), who, as you may know are first seen (as Baalberge culture) in the southern part of East Germany, later expanding into parts of Poland and Northern Moravia, as well as colonizing the until then apparently desertic Brandenburg forest. However later on, these East German IEs became rather isolated and eventually were absorbed by their Polish brethren with the Globular Amphorae culture.

If Eupedia demonstrates anything...

... is that it is a misleading reference.

Seriously, that site is awful: it has just somebody's (more or less weird) opinions. Just because it has the "-pedia" suffix, doesn't mean it is informative.

... but don't forget that R1a has been found at roughly equal percentage with R1b in central-East Germany 2,000 years before Slavic expansion.

Actually single-handedly (Haak'08) and actually some 3300 years before Slavs are even thought to exist, within the context of Corded Ware.

Poland obviously cannot be a source of anything - it is a receptor.

No. Poland can be a local source of R1a1a variants, as it was the main core of Western Indoeuropeans (Eastern Baalberge>>Lubon>Globular Amphorae>Corded Ware) in the 3rd milennium BCE. While you do have reason on the area being sparsely populated in the Neolithic, that invalidates nothing in what regards to Indoeuropeans.

You also seem to exaggerate the importance of recent historical migrations and even the devastation caused by the 30 years war. If it would have been as extreme as you say, then countries like Brandenburg or Saxony would have just vanished. The figures are simply inflated.

I'm not "disconnecting" anything. I'm saying that Annie Mouse's request seems to have had an embedded assumption that is unproven and, possibly, untrue. The assumption is that instances of genetic drift (aka change in allele frequency) must leave a correlative trace in the archaeological (or linguistic) record. It is precisely this assumption that have led Zhiv, Semino, and so many others to make the blunderous conclusions that they have made.

You are disconnecting them and "Zhiv" and the others are at least much more correct than you are, even if they might also deserve some criticism.

A Y-DNA haplogroup is not any abstract entity that dwells in the elusive plane of mathematics but in fact it is the descendants of some specific guy. That means, at least in the beginning, a patrilineal "clan" that expands either in virgin land or at the expense of others, and that requires an archaoelogical culture that can be associated with it. Without such socio-political tools, the clans just can't expand because they are just like anybody else.

You cannot disconnect haplogroups and real people.

Vincent said...

. . . . at least in the beginning, a patrilineal "clan" that expands either in virgin land or at the expense of others, and that requires an archaoelogical culture that can be associated with it.

This view may suit your bias, but there are several problem.

At the most basic level, you have the case where there is a single culture: any changes in Y-haplogroup frequencies within such a culture will not, obviously, be detected archaeologically.

An extension of this is relevant to us because we are trying to discern the cultures 3,000 to 4,000 years later. Many differences which were clear then may be totally obscured now.

ashraf said...

mr Gioeillo here's my email

anatolian_mardini@yahoo.com

thank you

Maju said...

At the most basic level, you have the case where there is a single culture: any changes in Y-haplogroup frequencies within such a culture will not, obviously, be detected archaeologically.

Indeed. But, unless there is strong drift, there should be no changes, other than a slow accumulation of new mutations (necessarily minoritary) and the occasional minor flow with neighbors.

This in what respect to a static culture. In other cases, there's no reason for any lineage to expand geographically if there's no people expanding as well, carrying with them their cultural signature.

An extension of this is relevant to us because we are trying to discern the cultures 3,000 to 4,000 years later. Many differences which were clear then may be totally obscured now.

Psah. We are obviously unable to percieve the fine detail, as if it was a documentary or we would be living there, of course. But when people migrate they carry their culture with them, regardless that it may be altered later on, as generations pass. This signature is generally very visible in the archaeological record, provided that this record exists.

So in most cases, at least in Europe, we can discern quite clearly if there was a migration or not just from the archaeological record. And haplogroups can't travel on their own.

Unknown said...

@Vincent

There is no embedded assumption that everything will leave a record/trace. The assumption is that what has left achaelogical evidence has to be consiostant, or at least possible, within the (Dienekes')theory. Any other approach is just not scientific, it is just political posturing.

And despite Eupedias problems it is still a nice way of summarizing genetic data. Most of us can read past the politics to get to what we want to know. I expect that I will have to do the same with Dienekes version.

Maju said...

Not sure if you're being sarcastic: Eupedia has loads of misleading and even blatantly incorrect data. I mean in particular the prehistory/haplogroup dispersal maps.

The author doesn't even know what's the difference between Bug-Dniestr, Dniepr-Don and Kurgan cultural groups. Though in other areas he has just copied the archaeological maps I created back in time for my defunct Geocities site, some of which ended in Wikipedia, (even with the many naive errors that I can recognize now) and transposed the hypothetical haplogroups on them.

And a transposition that makes no sense whatsoever.

Can you make money doing that cr*p? Because if someone is making money of that piece of junk, then I should slam my head against the wall for not being 1% as smart (money-wise).

Hopefully it's just another amateur site.

eurologist said...

So it's a receptor that has ancestral lineages for R1a1 and I1. A receptor from areas that don't have these ancestral lineages and lower diversity in the ones that they do have.

What do the events in Prussia have anything to do with the areas that remained strongly ethnically Polish?

Plz explain to us how Central and Southeastern Poland was depopulated. That should be something I've never heard before. Can't wait.


If you go before Polish identity and even before Slavic times, you should not use the the term "Polish." By doing so you confuse the issue and add a nationalistic color to your writings that you may are not even be aware of. People say Danubian, not German or Hungarian or Austrian, and Saale River, etc., for a reason.

Sometimes we use today's geographical expressions for simplicity, but with the understanding that no connection to today's situation is implied.


Clearly, many German people of today are very, very closely related to many Slavic people of today. How could they not, sharing a vast region since LGM.

Ask an international fan in the Euro Soccer Championships to point out the Polish, Bosnian, and even Ukrainian players in the German teams. They will be at a loss.

We have also established that R1a was present in Central Germany Millenia before anything like a Slavic expansion, as was I2. Whether these people, from the Saale to the Ukraine, spoke "Baltic", proto-Germanic or proto-Slavic is another question, but given the archaeological and historic linguistic evidence, I still find it easier to believe that Slavic language elements, while developing for Millennia in the Ukraine and surrounding regions with tight Baltic and Germanic contact, entered much of Eastern Europe late, with the well-known Slavic expansion.

It is also reasonable to assume that R1a spread more widely, just then, and given the more than five or whatnot known Millennia of R1a presence from central Germany to the Steppes, there is no surprise that several distinct founder groups of much older age are reflected in this... and some ended up in Poland, as you would expect. I don't see a contradiction.

Gioiello said...

Ashraf writes: "mr Gioeillo (Gioiello, please)here's my email anatolian_mardini@yahoo.com
thank you".

Everyone can create an account on yahoo. Why haven't you asked me who is "Occhiai" and why you are an expert of "computational linguistic" rather than Turkish History?

I haven't time to loose.

ashraf said...
This comment has been removed by a blog administrator.
Maju said...
This comment has been removed by a blog administrator.
ashraf said...
This comment has been removed by a blog administrator.
Maju said...
This comment has been removed by a blog administrator.
Maju said...
This comment has been removed by a blog administrator.
Gioiello said...
This comment has been removed by a blog administrator.
Maju said...
This comment has been removed by a blog administrator.
ashraf said...
This comment has been removed by a blog administrator.
Maju said...
This comment has been removed by a blog administrator.
Gioiello said...
This comment has been removed by a blog administrator.
Maju said...

I'll take a look to it when I can but I think the truly temerary idea is to force connections that are most unlikely and that most linguist themselves take not with the proverbial pinch of salt but with a whole barrel of it. If not with mere laughs.

blogspot said...

@Anatole Klyosov "By the way, I said "IE R1a1", since there are another Indian R1a1, which came apparently from South Siberia between 8-12,000 ybp. "

is a migration from South Siberia to India confirmed by the archaeological records?

Maju said...

is a migration from South Siberia to India confirmed by the archaeological records? -

No. Absolutely not.

There are controversial indications of migration from Central Asia, not Siberia.

terryt said...

"There are controversial indications of migration from Central Asia, not Siberia".

That may have been what Blogspot meant. I guess we should specify exactly what region each name covers. Perhaps siberia east of the Urals and north of 50 degrees and Central Asia between the Caspain Sea and the Tien Shan, south of 50 degrees and north of the Iranian Plateau.

Maju said...

In common speech Siberia means Asian Russia (WP: "Siberia is the vast region constituting almost all of Northern Asia" (...) "Historically, Siberia was defined as the whole part of Russia to the east of Ural Mountains, including the Russian Far East"). Khazakstan is not Siberia and certainly Uzbekistan is not Siberia.

You may find an overlap in the Altai region which can be considered part of both regions and is in fact an important transitional node, not just between Siberia and Central Asia but also between West and East Eurasia.

terryt said...

That corresponds exactly with the regions I suggested, so we're agreed on that.

Unknown said...

I have possible explanation regarding the lack of Y-chromosome marker M458 in western Ukraine and central Belarus.

Marker M458 is not much Slavic marker as rather lechitic (polish) and was spread with lechitic tribes.

According to Nestor chronicle Eastern Slavic tribes Viatichi (Russian: вя́тичи) and Radimichs (Радимичи in Russian) were of lechitic origins.
The map of their settlements and pretensions correlates with marker M458:
http://tiny.pl/hxrz7

Viatichi were the main tribe of Moscow principolity that is why thay spread marker M458 along with their expansion of Russian territories.

And higher frequency of marker M458 north of Crimea could be explained by this fact:
Black sea steppes were not populated by Slavs till XIII century (look at the line on above map)when Russians finally destroyed Crimean_Khanate:
http://en.wikipedia.org/wiki/Novorossiya
Majority of settlers would be from Viatichi turned Ukrainian/Russian.

Maju said...

J: I can only agree with your explanation for Dnepropretovsk high
apportion of the new clade but in Russia it seems to be evenly
distributed among ethnic Russians, regardless even of their apportion
of R1a1a*. So no peculiarities for your Lechtic regions.

I am more in the line that it has Corded Ware origin and then was
re-distributed by Russian historical expansion.

And I do think that those who have argued for demic replacement at the Dniepr are very wrong. Actually, considering that most historical invaders of those areas were of Turkic or otherwise North Eurasian origin, the population would be now like that of Kalmykia or almost so.

Marjeta Vakar Manfreda Bissue said...
This comment has been removed by the author.
Marjeta Vakar Manfreda Bissue said...

Hallo,
I made some amater research about Slovenian y-DNA. It is in english (not everything)
http://www.veneti.info/en/home/articles/popular/263-poti-naih-prednikov

and in slovene (everything)
http://www.veneti.info/sl/home/articles/popular/263-poti-naih-prednikov

TODAY I found (with some more samples) that 137 Slovenians in "my" base have tested DYS447 and 57 of them are R1a.
1 has 22, 6 have 23, 46 have 24and 4 have 25.

If this: http://www.gwozdz.org/R1a.html#M458TestResults
is correct, than 37,22% Slovenes have older R1a and 4,38% have M458 (R1a1a7).

Unknown said...

Oddly, what has not been a popular discussion stemming from this post, happens to be the fact that the study suggests an Near Eastern (Turkic, Iranian, and Armenian) origin for R1a (R1a1).
A point which some have been arguing for, a very long time.

Rather, many of you are dwelling on the fact R1a1a7 is Uralic, although there is little reason to attribute this haplotype to Indo-European groups. It seems more like a later, Scythian marker. We know from history, that the Scythians inhabited the region, and later invaded various nations of the near east and india.

R1a1a7 is a another unfortunate construction of blatant Eurocentricity.

Unknown said...

This abstract is misleading as to what it implies about IEs. The full article goes into the evidence, which has again suggested, a Near Eastern origin for R1a (R1a1)

Here is a snippet from Wiki:

"R1a evolved from a male-line ancestor who was in haplogroup R1. R1 is defined by SNP mutation M173. The R1a clade can be distinguished by several unique markers including M420. R1a also has a similarly common sister-clade, called R1b, which also has M173, but is distinguished by its M343 marker. There is no simple consensus concerning the places in Eurasia where R1, R1a or R1b evolved, although Underhill et al. (2009) recently suggested that "the most distantly related R1a chromosomes [...] have been detected at low frequency in Europe, Turkey, United Arab Emirates, Caucasus and Iran" implying that R1a's origin may perhaps be somewhere near these regions".

'This' is what expresses what is the most relevant from the study, and it has been the long-standing question. Not the fact that R1a343k69b8a93.1, or whatever , came from wherever.

I can't help to feel but that the IE topic will go and forth, childishly, for eternity.

Maju said...

That would make sense, John. However, as the paper is behind a paywall, we (at least I) had to rely on what others comment. In this sense I welcome that info you provide, thanks - which would suggest a West Asian coalescence area for all R1, not just R1b.

It seems to contradict previous phylogeny-based suggestions that R1a could have an oldest depth in India, right? However I wonder if this may be influence by what you sample and analyze. I mean: Sharma did not estimate coalescence time for particular West Eurasian populations, but only for the bulk of them; he did with South Asians instead, getting deepest age for the Saharia tribe and the Kashmiri Pandits.

Did Underhill studied South Asians in any depth. How do these two papers compare?

Unknown said...

this makes sense or finally shows Slavic similarities as with the R1a thing it almost tried to show that Serbs are not Slavic and that brahmins are closer to russians or poles than serbs thats bs this has cleared it up a little bit

szopen said...

Couldn't mentioned R1a1a7 similarities in Poland in Russia be explained by expulsions of Poles into Siberia and such? Since XVII century this sort of involuntary (and sometimes voluntary) colonisation was common. I have read that at one moment Irkuck had 50% catholic population, probably most of them Poles... Heck, Poles even ended in Russians settlements in North America..

AthenianHoplite said...

I have heard that R1a in Greece came with the Mycenaean Greek invasion from modern day Ukraine. This means that R1a in Greece is also Mycenaean, and it is not only a result of Slavic admixture.

R1a is also called Balto-Slavic, Mycenaean Greek, Macedonian. this means that it is also a "Native Greek haplogroup" not only a Slavic one.

Anonymous said...

a way to find the quantity of slavic R1a in greece and other regions could be the following.Considering the Polish as 'pure' Slavs we can observe that out of the 50% R1a the 35% is R-M458 which is aproximtely the 70% of r1a.Taking as example the greek Macedonia we can conclude that 15%(8.8x1.7) of haplogroups is slavic R1a .R1a in greek macedonia reaches about 25% so 10% r1a was existant in greek macedonia before the advent of slavs.Calculating similarly the non slavic R1a in nearby regions we have about 5% non slavic R1a for greece and macedonia.So non slavic r1a drops out of greek macedonia fact that sets non slavic R1a as a possible 'genetic marker' of the ancient macedonians.

Anonymous said...

well in my previous post i did my math wrong.To calculate the slavic R1a we have to say (8.8x100)/70.(8.8 has to be 70% of total slavic r1a as 35% is for 50%.

Hellen said...

I want to present some information which could be helpful in understanding the controversies RE M458 in Russia ans Slavic expansion to the Balkans.
1. Russia was not populated by Slavs at the time, when the Slavic expansion to the Balkan began. The expansion of Slavs to the later territory of Russia began in 7th century.

2. Less density of the marker in the central Ukraine could be explained by that the territory was populated (in part?) by other tribes
http://en.wikipedia.org/wiki/Ukrainian_language

During the Khazar period, the territory of Ukraine, settled at that time by Iranian (post-Scythian), Turkic (post-Hunnic, proto-Bulgarian), and Finno-Ugric (proto-Hungarian) tribes, was progressively Slavicized by several waves of migration from the Slavic north.
3. “J” already mentioned:
--- ... higher frequency of marker M458 north of Crimea could be explained by this fact:
Black sea steppes were not populated by Slavs till XIII century (look at the line on above map)when Russians finally destroyed Crimean_Khanate:
http://en.wikipedia.org/wiki/Novorossiya
Majority of settlers would be from Viatichi turned Ukrainian/Russian.

4. ~50% of Russians and Ukranian people have R1a, so M458 is only a part of this haplogroup. So, we have to multiply the proportion in the places where Slavs settled.

5. Now, how Slavs came to their settlement to Balkan:
In 6th century, Nomadic, very brutal Avars in their expansion to Southern Europe force Slavs who were on their way to settle in the fertile Pannonia to produced agricultural products for them. These Slavs were not from modern Poland, but from Carpathian area and around. Slavs stayed in the Pannonia 1 - 1.5 centuries. During this time the exchange of population with the Balkans occur. First in their raids to the Balkans Avars were alone. Slavs were not skilled in war, but later they got this experience and contributed to the Avar's raids. Later, they were able to do this themselves, and finally (7th Century) invaded the whole Balkans. (You can read about this in the Wiki. I read more serious literature in detail).

So:
- Slavs who occupied the Balkans were not from the areas the most rich in M458.
- In the moment of the invasion, Slavs could be already pretty much “Balkanazed” .
--- The exchange of population with the Balkan could change the proportion of haplogroups at the moment of invasion.

the area of the Pannonia was re-occupied by other tribes later.
>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>
I also want to add that 3 centuries before this (3rd century), Goths were in the territory of Ukraine and around the Black Sea, they could Previosly, Tracian tribes were somewhere around Carpatians.