search this blog

Saturday, April 30, 2016

Y-hg J2 cannot be a Proto-Indo-European marker

The claim that the Proto-Indo-Europeans came from West Asia and largely belonged to Y-haplogroup J2 seems to be popular online nowadays. I won't discuss here in detail the reasons why, but suffice to say it has a lot do with aggressive lobbying on several online forums and blogs by a few people of Southern European extraction, like Dienekes Pontikos.

It was always a shaky proposition, but difficult to debunk thoroughly. Until now.

Thanks to recent advances in both modern and ancient DNA research, we can now safely say that Y-haplogroup J2 was not involved in any rapid, large scale population expansions during the Late Neolithic/Early Bronze Age (LN/EBA), the generally accepted Proto-Indo-European time frame.

It thus fails to meet even the most basic criteria of a Proto-Indo-European diagnostic marker. The Proto-Indo-Europeans, after all, were surely highly patriarchal and patrilineal, and therefore expected to have left a clear signal of their migrations in the Y-chromosomes of many present-day Indo-European speakers.

For instance, an analysis of data from the deep sequencing of human Y-chromosomes as part of the 1000 Genomes Project suggests that not a single major subclade of J2 began expanding even roughly close to the LN/EBA. See here.

In the plot above three lineages jump out at you. E1b, R1a, and R1b. The first is associated with the Bantu expansion, that occurred over the last 4,000 years. The second two are likely associated with Indo-Europeans in both Asia and Europe, respectively. The timescale is on the order of 4 to 5,000 years in the past. The association between culture and genes, or the genetic lineages of males, is rather clear, in these cases. In other instances the growth was more gradual. For example, the lineages likely associated with the first Neolithic pulses, J and G.

Moreover, not a single instance of J2 has been reported from remains classified as belonging to the Andronovo, Battle-Axe, Corded Ware, Khvalynsk, Poltavka, Potapovka, Sintashta, Srubnaya and Yamnaya archaeological cultures. In other words, Kurgan and Kurgan-derived groups generally accepted to be early Indo-European, whch is a view that now has very strong support from ancient genomics. See here and here.

To date, most of these samples have probably come from elite burials. So at some point, when many more non-elite samples are sequenced, we are likely to see J2 among a few supposedly early Indo-European individuals. But so what?

There might be a couple of ways to salvage the Proto-Indo-Europeans = J2 theory. We'd have to argue that...

- the Proto-Indo-European time frame was actually the early Neolithic


- the Proto-Indo-Europeans were a small group that Indo-Europeanized the steppe Kurgan people, perhaps mainly via female migrations, and then did not partake in the main early Indo-European expansions

But the former is not particularly clever when viewed in the context of historical linguistics data. See here.

For instance, almost all IE language branches testify to a word designating ‘wool’. Since archaeological evidence suggests that wool sheep did not exist until the beginning of the fourth millennium BCE, the existence of the word in PIE would indicate that the disintegration of the proto-language could not have taken place before this date. Similarly, words for concepts such as ‘wheel’, ‘yoke’, ‘honey bee’ and ‘horse’ may be correlated directly with concrete, datable archaeological evidence.

And the latter isn't very parsimonious, and to me looks like special pleading. Why even bother?


Gökhan said...

Anatolian hypothesis is an out of date hypothesis which is based on subjective prejudges of some dinosourous european archeologists who thinks that "indo european" expansion is expansion of the civilization, so the agriculture. Thanks genetic studies that disproved such a fictional hypothesis forever. Same thing also valid for racist authors who claims that so called "white race" come through Caucaus. All of those fictions based on the idea of "indo europeans are the best race" which have brought civilization and/or "whiteness" to all around the world. Genetic studies are disproving that too.

On the other hand, its also possible that some J2 clades like J2b might participated to indo euopean expansion. Whether elite or not they might be the part of such an expansion.

George Okromchedlishvili said...

Well, J2 association isn't as comical as G2a fantasies that a small batch of Ossetia nationalists have apparently been parading (on Caucasians are notorious for nationalistic glorification of their "great" ancestors so it's probably not wise to be too hard on these guys).

Arch Hades said...

Well the CHGs carried J2, and may have even been the originators of J2..CHG autosmal genes were in all those steppe populations that don't carry J2, but oddly their Y lineages had been mostly erased. My guess is a few patriarchal beasts carrying the R1 lineages came out on top in the steppe and the J2 lineages must of have disappeared. So no, I don't think J2 could be the signature marker of the PIEs.

Matt said...

I kind of pretty much agree with the overall argument about a star like founder effect driven expansion of R1 and apparently no such expansion in J2, and the spatial pattern is also generally better.

However looking at their graph: (link from Razib Khan's blog)

I'm not sure if I actually believe their growth estimates are going to be a smooth and continuous process for those lineages that they *do* find as exhibiting long term population growth, like G and J2.

From reading that G seems to have a pretty much continuous rate of growth since 30,000 YBP from their estimates, J2 since approx 25,000 YBP.

Which both seem kind of strange! Shouldn't the Neolithic matter a *little* bit, with population sizes for Neolithic lineages having a sudden growth spurt around then? Even if there's no star like substructure due to founder effects.

Autosomally there is a signature of the Neolithic present - (except for Finns). So should be found in the Y as well. (Also in light of Jones finding that CHG, the ancient J2 carriers we have were at pretty small population sizes during the last Glacial Maximum, which doesn't square with continuous growth of populations associated to J2).

If G or J2 doesn't have a star pattern, OK, no sudden strong founder effect, but founder effect or no, a big increase in mutational diversity downstream from the mutation that defines G or J2 would be expected to dated to the Neolithic onwards, not 30000 YBP, which is quite a strange timing and seems out of step with autosomal population size expansions.

So I'm not totally sure their method is dating population size expansion times correctly for some of these haplogroups, and the true picture is probably more complicated than a single expansion for many groups, with more dynamic history of stasis and expansion. This is something I'm dabbling in a bit and don't have a lot of knowledge of though.

andrew said...

I agree with the conclusion, but I also think that the understandable reason that people associated J2 with Indo-Europeans is that J2 was present in Anatolians and Greeks, each of whom were Indo-Europeanized ca. 4kya, and once those civilizations assimilated J2 men from the substrates that they conquered, both the Hittites and Greek speaking Greeks expanded to various places along the Northern Mediterranean coast and into the Levant, leave traces linked to Indo-European peoples in those places. I recall toying with the J2 as Indo-European marker for a couple years when the first Y-DNA hg papers starting coming out and never managing to make it fit. (A fair amount of Southern Mediterranean and Iberian J2 may be a legacy of historic era Islamic expansion arriving as a minor portion of an Arabian mix that is predominantly J1.)

The trouble is that J2 doesn't generalize widely and consistently enough across the entire Indo-European geographic range to be associated with PIE, isn't in the relevant ancient DNA, and seems to be already emerging and expanding in the Mesolithic and early Neolithic era.

An important unsolved question about J2 is when it, and its sister clade, J1 arrive in their West Asian highland and SW Asian lowland modern core territories. Given that G2 dominates EEF and is probably derived from an Anatolian EEF population, Y-DNA J may have arrived in Anatolia sometime after the first wave Neolithic that produced the LBK but before the Hittites. For example, a mixed J2/J1 population could have arrived with Eneolithic wave of highland farmer/herder/copper worker migrants from the Caucasus to Anatolia and the Zargos Mountains, with a herder population that was heavy in J1 due to random founder effects later migrating into the lowlands and becoming the founding population of modern Arabs.

I've long been deeply skeptical of the hypothesis that there is continuity between modern J1 populations of SW Asia and the Upper Paleolithic hunter-gatherers of SW Asia who eventually colonized the rest of Eurasia.

andrew said...

Another strongly suggestive piece of evidence against J2 as an Indo-European marker from the Wikipedia article on the hg which has data on frequencies in linguistically Dravidian v. linguistically Indo-European South Asians:

"J2-M172 was found to be significantly higher among Dravidian castes at 19% than among Indo-European castes at 11%.

J2-M172 and J-M410 is found 21% among Dravidian middle castes, followed by upper castes, 18.6%, and lower castes 14%.[21] Subclades of M172 such as M67 and M92 were not found in either Indian or Pakistani samples which also might hint at a partial common origin.

Haplogroup J-P209 was found to be more common in India's Shia Muslims, of which 28.7% belong to haplogroup J, with 13.7% in J-M410, 10.6% in J-M267 and 4.4% in J2b (Eaaswarkhanth 2009)."

Kristiina said...

Matt, where do you get the idea that Finns lack Neolithic ancestry? Where do they say so in that paper you refer to in your post? In this figure Finns have the same components as the British (

My Eurogenes K8 results show that I am 26 % Near Eastern (WHG is at 50% and ANE at 17%).

This figure from the Kotias paper ( shows that Finns are as close to Kotias as Spanish.

This figure from the Aigean farmers paper ( shows that Finns are 25% Barcin8.

Aram said...


What You think about J2b branch? It has a rather atypical distribution for Neolithic demic expansion. Also Hamilton thinks that it's real TMRCA in Europe is younger than proposed by other methods. He proposes 5200 ybp for the European J2b.

Aram said...


**Same thing also valid for racist authors who claims that so called "white race" come through Caucaus. **

I don't know why for some people the skin color is so important. For me it is not.
But I think Caucasus (CHG) as one of sources for white skin genes is a plausible theory.
Anybody living in South Caucasus and Eastern Anatolia can testify that in that region ( especially in mountain parts ) there are very light skin people with blue eyes without having a recent Northern or European ancestry. Also as far as I remember one of CHG samples did have quite light skin.

Aram said...

This guy has no any recent Slavic and European ancestry as far as I know. He is a wrestler. He has no typical Armenian look but in mountainous regions one can find 1-2 cases in each village. I guess the same is true also for neighbouring countries where the CHG level is particularly high.

Davidski said...

What You think about J2b branch? It has a rather atypical distribution for Neolithic demic expansion.

You mean M12? Do you have a good map?

Isn't that a Near Eastern marker? Can't really see it having anything to do with the Proto-Indo-Europeans.

Onur said...


In this figure Finns have the same components as the British (

No, they have some polymorphic variants private to themselves, you should read the description and look at the colors carefully (the same population-specific colors are used in the parts b and c of Figure 1 too).

FrankN said...

On the linguistic side:
Honey (actually mead and the Graeco-Italo-Celtic mel) is quite a Wanderwort: The mead also occurs as S.Drav *mat, Semitic *mays, and Chinese 蜜 (Old Chinese "mit"); M. Witzel has somewhere presented a linguistic chain into SEA (Thai mthu, Malay madu, etc.). PIE *mel(it) is shared with Uralic, Altaic and Dravidian, in the sense of honey, liquor, sap, or milk. The bee finds close parallels in *pera/*pora as a root shared by Uralic, Altaic, N.Caucasian, Afro-Asiatic, Dravidian, Sino-Tibetan, Korean and Japanese (c.f. Lat. fucus "bee drone" vs. Gondi phuki "bee, honey", Latv. biti vs. O.Jap. patí, Germ. Biene vs. Burush. phen, OEngl. beo vs. Lao pheo, etc.). Looks like paleolithic terminology or Wanderwörter; not particularly well suited to identify the PIE homeland/markers.

The horse is related to the hart, apparently HG terminology. Of course, there is PIE *hekwos, shared with N. Caucasian, and possibly connected to Proto-Altaic *èk‘á ‎(“to move quickly, to rage”). Most likely also HG terminology, borrowed rather than invented by PIE, and not very "posh" as a term. For horses as status symbols and war animals, Germanic and Celtic went for the mare (originaly "stallion", c.f. "marshal"), apparently a borrowing from Mongolic. The widespread Slavic Kherevec "Stallion" contains the Georgian male specifier Khary "bull" mounted on a root that appears related to Turkish beygir "horse". Imports from Central Siberia, with some Caucasian mediation, as would be expected from the archeological record. But, even if borrowed, isn't PIE *hekwos giving an
indication that PIE only broke up after they received the domesticated horse? Actually not. Such technical terms can spread more or less uniformly at any point in time, as, e.g., demonstrated by the Arabic alcohol.

Wool is etymologically related to the (spinning) wheel and as such rather EEF (EN) than pastoralist (LN) terminology. NE.Cauc *pal and Tibet. *pal/*pul look noteworthy.

That leaves us with wagon, wheel, and axle. Spinning wheels and straight wooden shafts mounted to axes are EN technology. In fact, already the Mesolithic (Kongemose/Ertebolle) had wheel drills and axes/adzes for boat building. As to the wagon/cart, archeological evidence so far places its origin with Funnelbeakers. There are older examples of children toys on wheels from CT, possibly also Kiziltepe in SE Anatolia, but note that Amerindians (Olmecs) also had wheeled toys without ever developing the cart/wagon. Thus, the PIE terminology related to wheel and wagon may have originated anywhere between Anatolia and the SW Baltics.
Anyway, FB was involved in the spread of vehicles, either as inventor
or as early adaptor. So, let's look at FB yDNA from around 3,500 BC to see if any putative PE marker shows up there. We don't have much in this respect:
- R1 (xR1b1a2, R1a1a) from Quedlinburg (I0559, 3645-3537 BC);
- G2a2a from Salzmünde (I0551, 3400-3025 BC)
- I2a1b1a from Esperstedt (I0172, 3360-3086 BC)

EEF, Mesol. HG, and R1 of unknown origin, all possibly involved in the spread of PIE. Make your choice. J2, however, is missing, as is sizeable CHG admix in the a/m samples (which, however, do neither cover Polish nor Nordic FB).

Nirjhar007 said...

Something big coming up in some hours, are you ready?. :D

Matt said...

Hi Kristiina, I'm not really saying what you think I am.

When I mentioned the Finns, its that the Finns in the 1000 genomes don't show in their genomes the same sign of increase in population size around the Neolithic as everyone else.

That's the "signature of the Neolithic" in population size increase I'm talking about.

See this Figure 2b -

Instead they seem to stay a genetically small population, with more similar genetic population size to paleolithic Eurasians. That really all I'm referring to. I'm not talking about the sources of their ancestry.

(This is all in connection with my not really buying into the idea that the Y-dna haplogroups G and J2 have really been expanding at a continuous growth rate since around 30,000 and 25,000 YBP respectively, as depicted by the graph I linked upthread).

However, looking at it in more detail, there are big differences between Iberian, British, CEU, Tuscan as well, so I might have overstated the differences in Finnish population size history from other Europeans.

wagg said...

FrankN: (about ekwos) "Such technical terms can spread more or less uniformly at any point in time, as, e.g., demonstrated by the Arabic alcohol."

Ridiculous. The root ekw- comes from a PIE root: all the IE words are derived as expected in every languages by the rules of change determined by the experts.
Anyway, how could anyone believe that the same root is found from the eastern part of the IE world - i.e. ancient north-western China (Tokharian B yakwe) and ancient India (Sanskrit asva) - to the western part (old Irish ech) as a post-PIE wanderwoort (sometimes through large non-IE territories)....

Gill said...

Some branches of J2 probably got "caught up with" the IE expansion, but we don't know which ones yet. Clearly J2 expanded with the Neolithic revolution which probably brought it all the way to Central Asia as well as the Caucasus, Europe, and South Asia.

And the biggest impact was on J2, not by J2. Meaning we see differentiation of the branches at around the IE expansion timeframe (Easily seen on YFull for example). The surviving branches which differentiated at that time would have been people assimilating into or integrating into/with the IE expansion. Those stories are probably interesting on their own.

Bottom line, we need a lot more NGS level Y data. Ancient and modern.

Gökhan said...


You are right. I just crtized the attitude on human history. Those racist "ideas" have ruled the scientific world for decades. But as you state white skin and light haired colours are the so frequent along CHG dominated populations.

capra internetensis said...


Words that spread after the break-up of a proto-language can in principle be distinguished from words present in the proto-language (whether borrowed into it or not) because they will not have undergone all the characteristic sound changes of each branch from the earliest stage. There is the possibility of etymological nativization, or the word may lack suitable phonemes to track its early history. But we can at least see that 'ek'wos was present before satemization of velars, is present in almost all branches, and is usually regular, so to suggest that it spread as a loanword after the break-up of PIE seems quite far-fetched. Of course the horse need not have been domestic at the proto-language stage.

What is the etymological relationship between wheel and wool? Generally they do not seem very similar; Greek lenos and kuklos, Sanskrit ūrṇa and cakra show velar stops in "wheel" which have softened to in English and eventually been lost (in most dialects) but were never present in "wool".

MfA said...

PIE wool is *h₂wĺ̥h₁neh₂, it doesn't have any connection to wheel.

proto-aryan *herine-
Avestan *verina
Sanskrît ऊर्णा (ū́rṇā)
Hittite hulana
Armenian գեղմ (geġm)
Russian волна (volna)
Old Greek λῆνος (lē̃nos)
Kurdish(Kurmanji) hirî

Kristiina said...

Thanks Matt for the explanation! I can buy that! So it was in particular about this graph:

I see that the most spectular growth has taken place in Bengali. Chinese are on the second place. The population growth in Finland has been very moderate and not very high in England and Scotland either. Very understandable. In Finland a population expansion started only at the end of the 18th century and did not result to a really high density.

Fanty said...

Has anyone readt that study that suposedly found what populations are the source for some traits?

I read it said that milk lactose tolerance trait originated in neolithic farmers, but was rare (possibly because there was no need for it) and steppe people would have needed it because of their culture, but did not possess it. Only the combination of steppe peoples food culture with farmers lactose tolerance genes made them explode and spread. blabla.

Well, the same study claimed the fancy pigmentations of Europeans (fair hair and eyes) originate in the Hunter Gatherer cultures. Did the study itself actually say WICH hunter gatherer cultures these fair traits suposedly originate? After all there is WHG, SHG, EHG and CHG involved.

Krefter said...

"Something big coming up in some hours, are you ready?. :D"

Ancient DNA from India :). Bet you $500 R1a-Z93 pops up in the Bronze age and there's lots of Andronovo ancestry in South Asia today.

Davidski said...

Nah, not from India just yet, but it should be pretty good.

Give it a few more hours though.

la señora bibiloni said...

Rakhigarhi DNA results ae out? I hope it's true...

Krefter said...


Iain Mathieson blog, "Lactase persistence and ancient DNA"

The earliest confirmation of LP-allele is from Bell Beaker. It also pops up in Bronze age Hungary, Unetice, Srubna, Andronovo, Late Neolithic Sweden, and Late Neolithic Ireland. It doesn't matter which ancestors(WHG or EEF or CHG or EHG) is the ultimate source. What matters is that it was very strongly selected for in Europe between 3000 BC and 2000 AD.

I've looked over the evidence. This LP-alele really does have a huge impact on whether or not someone is lactose intolerant. Most Europeans without it are lactose intolerant. So, what we have is from 6000 BC-3000 BC, essentially everyone in Europe was Lactose Intolerant. Then suddenly in 2300 BC we start to see around 1-5% of the people are lactose tolerant. Then by the Iron age in Poland(we have hardly any data from anywhere else), over 60% of people are Lactose Tolerant. And we can see today, in Iberia and Northern Europe, well over 70% of people are Lactose Tolerant.

Krefter said...


Forgot to mention. A very high percentage of people in SC Asis have the same LP-allele Europeans do. It's possible between 2000 BC and 2000 AD, like in Europe, it was selected for in SC Asia. It's also important to note the LP-allele wasn't selected for in all of Europe. SE Europe doesn't have a lot.

Nirjhar007 said...

No Its not the Indian one.

FrankN said...

@wagg, CI, MIA

Horse: Most diagnostic sound changes date to the LBA/ IA. Proto-Celtic still had *ekʷos that only after the P-/Q-celtic splits developed into current forms. Mycenaian ikkʷos underwent the OGr. “Kw”-“P” sound shift to turn into (h)ippos. Satemisation in Indic was still in full swing between the Rigvedas and later Vedes. However, aswa "horse" already occurs in the Rigvedas, so we have a terminus ante quem around 1500 BC, which would still leave a bit of time for common borrowing after the split of PIE.
Old Armenian էշ ‎(ēš, “donkey, ass”) demonstrates use of *hekwos in the general sense of "equids". So, I wouldn't exclude the possibility that Indic aswa was borrowed, or influenced soundwise, from IVC language(s). C.f. for "donkey" Proto-Altaic *ĕĺǯu, Proto-Semitic *ʕayr- (vs. *hek[wos]).

Wool: Pokorny links "wool" to PIE *uel- "to turn, wrap" (c.f. Lat. volvere).
"Wool" would either mean "what has been turned (spun)", or "the twisted, curly one". Original use as participle (“spun”) explains the initial velar stop (c.f. Welsh gwlan "wool" vs. German ge-wellt "curly", past participle of wellen "to wave, turn around"->Welle "rotat.shaft"), adjective use the suffixed “-n” (“woolen” vs. PCelt *wlanā, Lat. lana etc.).
MIA: h₂wĺ̥h₁neh₂, comprises participle prefix and adjective suffix attached to the verb w(e)l ~ *uel-.

CI - for the initial velar stop in "wheel" terms, they are commonly derived from PIE *kwelh₁ "to turn", not from *uel-. However, both stems are so close semantically and phonetically that we might be dealing with dialectal variants of the same root. Outside of wheel terminology, *kwelh₁ has preserved the sense "to turn" only in Toch., Arm. and Alb.; it is missing from Anat. and Celtic; in IndAr it refers to movement in general (c.f. Germ *felhana "to go, proceed"); in Italic, Germanic and Baltoslavic to round/bending body parts (neck, knee); Latin also has colere, cultivare (i.e. turning over soil, e.g. by ploughing), and colus "spinning distaff".
Greek kuklos "circle, wheel" is intriguing for its irregularity. The "u" can't be reconciled with the "e" sound in *kwekwlos. Moreover IE "Kw" is expected to shift to "P” in Aeolian Greek (c.f. pente "five"), PIE *kwekwlos "wheel", have yielded OGrk. peklos/peplos instead of kuklos [Pokorny relates PIE *kwelh₁ to Gr. πέλω "in movement", πόλος "axle, ploughed land"]. C.f. on these and other problems (e.g. vocalism in Tocharian and Baltic):

Sp, the two PIE roots for "to turn", *uel- and *kwelh₁ may represent early dialectal variation within PIE. The former tends to have better preserved the original meaning, and forms the base for IE "wool" terms, while the latter, with several phonetic inconsistencies, is supposed to underlie "wheel" terminology.
O.Engl./High German Felge "felly" (spoked wheel rim) may indicate confluence of both roots in West Germanic, or alternatively a third phonetically similar PIE root *pelk̂ "to turn (plough)" that, is limited to Celtic, Germanic, Slavic, and possibly Greek (if πόλος "axle, ploughed land" isn't derived from *kwelh₁). PIE *(s)ker "to turn, bend" ("curve", "circle", "curly" etc.) adds a forth, rhotacised phonetic variant.

wagg said...

@FrankN : about *ekwos.

Old Armenian is ONE language among many others (that have the exact same meaning) so it doesn't prove anything. Celtic and Mycenian don't prove anything either (there is no particular reason why the word should have undergone a specific sound change at this point on the specific consonant at this point in time)... as for 'asva' it's a classic sound change and is perfectly in accordance with the shift changes expected in sanskrit, there is no reason at all to imagine a particular individual story for the word 'asva' or Iranic 'aspa', for that matter. You're trying to fit a square-shaped object in a circle-shaped hole. It really seems as simple as it looks.

The idea that a common root *ekw- would spread throughout Eurasia (many thousands of kilometers: Ireland - China and India !!! during a time such as Bronze Age!!) AFTER the PIE split and undergoing the same changes that the PIE words gradually underwent is just ridiculously unbelievable to me (and I'm sure to most reasonable individuals).

The distance alone is enough to disappprove the concept (unrealistic, illogical and anachronistic) but beside that, the word wouldn't have undergone the same sounds shift than the PIE words - at least in most cases. (E.g. You have latin Cord- and english (germ.) Heart - words that were separated at the PIE split and undergone a specific sound change (PIE K -> Germ. H) - but you have latin case (-us) and German käse and Dutch kaas (and English cheese) for it was a loanword from Latin in germanic (Lat. K -> Germ. K)).

Of course the same thing is true for words such as *rot-eh (=wheel/chariot : Gaulish rot-, Irish roth, Latin rot-, Lithuanian rat-, Albanian rreth, Sanskrit and Avestan rath-), or any other root that I know of (*kwekwl- (wheel), *aks- (axle), *wegheti (ride, transport?, move?), etc...): all the sound changes are just what we expect.
Same with the common stem for metal/copper? (Sanskrit ayas, Latin aes/aer-, Gothic aiz, old Norse eir, etc...).

FrankN said...

@wagg: I feel you are getting a bit lost in time and space. The P-/Q-Celtic sound shifts in question here were regular. P-Celtic evolved sometime before 600 BC (it is already evidenced in Lepontic inscriptions). The dating of Proto-Celtic is unclear. However, according to the Italo-Celtic hypothesis, both branches emerged from the Urnfield culture, which would place the split some time after 1200 BC, when the Proto-Villanova culture as presumed bearer of Italic languages emerged. The shift from Mycenaian to Ancient Greek, which involved the "kw"->"p" change, is dated to 1100 - 800 BC. The Germanic sound shift, which included "K"->"H" (centum-hundred) is roughly dated to between 600 BC (borrowing of cannabis from Skythian, which was sound-shifted into hemp) and 300 AD ("Hesse" still recorded as "Chatti" by various Roman authors), as a process that commenced in the East and gradually proceded westwards. Satemisation of Sanskrit was, as shown by M. Witzel, still ongoing between the Rigvedas and later Vedes, so we are probably talking something like 1500 - 1000 BC here.

When PIE started to disintegrate is controversial: The Anatolian hypothesis obviously promotes an earlier dating than the Steppe hypothesis. Gimbutas' approach (which is problematic as aDNA denies Steppe incursion into Transcarpathia) places it at around 4000 BC (destruction of Varna Culture). That would leave a window of some 2200 years until the first written evidence of IE (Hittite), and of some 2000 years in which already disintegrated early IE languages would still be geographically confined to North Central Europe (Corded Ware), the Western Steppe (Sintashta-Andronovo eastward expansion only by 2100 BC), and possibly the Eastern Balkans and Anatolia. No Ireland, no Tocharians, no Indoaryans yet, an area considerably smaller than is now covered by Slavic languages!

Fanty said...


Yeah, but I have no interest in the lactose tolerance allele at all. My interest was, where the fair hair/eyes alleles originate.

And I read in short summaries about that paper (wich I am too lazy to read) that Europeans have that one from "the HG", but what HG? You dont seem to know either.

Fanty said...

Oh and to make it clear why I want to know:

Because some here claimed that there is blond people in CHG related populations.
So, do they have it because CHG is the source for European fair hairs/eyes aswell, or do they have it because they have a tiny bit of WHG/EHG anchestry?

wagg said...


OK, I'm not sure the idea of an *ekwos word spreading in the earliest phase of IE spreading (but post-PIE split) is really convincing or logical but I guess, if it's soon enough, it can't be said to be impossible (even though I find it hard to admit the cultural/religious significance of the horse in IE cultures as a relatively "late" thing, and see it more of a central thing from the start, and thus I see good no reason for the replacement of an original horse word by a new one spreading throughout the whole IE area, despite the growing "originality" of the new IE cultural and liguistical identities).

Besides, the ancestor of the Tocharian language (which would apparently be the earliest PIE group to leave (after hittite, I mean, which is a special case)) is very likely to be linked with Afanasevo (so these languages developped isolated from other IE groups (as proved by the absence of othe IE influences, notably Indo-Iranian, who appeared both later and much wester originally, in all likelihood)), and it still got the *ekwos root. This strengthens the idea that *ekwos is PIE.

Pneumatikon said...

I'm not worried about the lack of ancient J2a on the Steppe. We were there. Men with the J2a Y-Chromosome in Northern India are overwhelmingly Brahmin. Their closest relatives are in Serbia and Montenegro. The Thapa surname in Nepal is common in military families. Thapa men with the J2b Y-Chromosome are almost an exact match for men living in Montenegro. Millet strains in in Mongolia are from Central Anatolia. But along the way they also show up in Kazakhstan - Andronovo territory.

There's 0% evidence your boys made it south of the Danube before the late(!!) Bronze Age. There's 0% evidence your boys made it to Italy before the Gallic invasion of the 4th Century, BC. There's 0% evidence your boys permanently settled in Anatolia before the Galatians of the 3rd Century, BC. There's 0% evidence that there ever was a "Coming of the Greeks," and international archaeologists working in Greece are stating to conclude the language is native.

More bad news. Your boys were also carrying the plague. And there's 0% evidence it was anywhere near us before... Justinian and Theodora?! Seriously. I think it was that late.

So DNA research from THREE SPECIES says the IE language is Anatolian.

Meanwhile, we've got the obviously Anatolian Cucuteni-Trypillian north of the Black Sea. They were there when wool first showed up. They produced the finest Neolithic/Chalcolithic figurines in our history. These figurines are the mark of Anatolians par excellence. They built proto-cities that dwarf the city states of Mesopotamia. Their cousins the Vinča invented metallurgy and probably the wheel. The two products...

...the Sintashta are famous for. I'd also like to point out the Sintashta are an amalgam of a Steppe population and a population from the Balkans. Or at least closely related to them.

The timing is also right. Greece/Crete about 6500 BC; when we know Çatalhöyük in Central Anatolia was overpopulated and its people were overworked. The Vinča of Serbia a thousand years later. The Hamangia on the Western shore of the Black Sea 4-500 years after that - though that's mostly a transit region. The Cucuteni-Trypillian 4-500 years after that. Boom. Boom. Boom. Slow and steady wins the race.

So we're getting an HD view of the Late Neolithic/Early Bronze Age and it's bad news all around for your view. We are on your base. You are not on ours. A non-Anatolian Steppe origin of the IE language is impossible.

And the worst news for you? The oldest IE languages with a bullet - with the exception of Tocharian - are... south of the Danube and in Anatolia. And the OLDEST is in Anatolia.

Pneumatikon said...

My DNA evidence:

My video on men with the J2b Y-Chromosome:

My video from earlier this week, where I start to link the Minoans with the Romans via the Sea People. I'm uploading a quick epilogue in a few days where I explicitly link Minoan ideology with these Neolithic/Chalcolithic cultures:

Davidski said...

Minoans weren't Indo-Europeans you idiot.

Pneumatikon said...

Really? You know that for a fact? The jury is out on that, but most professionals I've read think they were. I, on the other hand, have PROVEN that they were.

Davidski said...

I, on the other hand, have PROVEN that they were.

You're a fruitcake.

Pneumatikon said...

Hey, Davidski. Why don't you watch the GOD DAMN VIDEOS instead of bitching.