search this blog

Sunday, May 13, 2018

Hittite-era Anatolians in qpAdm


The apparent lack of steppe ancestry in five Hittite-era, perhaps Indo-European-speaking, Anatolians was interpreted in Damagaard et al. 2018 as a major discovery with profound implications for the origin of the Anatolian branch of Indo-European languages.

But I disagree with this assessment, simply because none of these Hittite-era individuals are from royal Hittite, or Nes, burials. Hence, there's a very good chance that they were Hattians, who were not of Indo-European origin, even if they spoke the Indo-European Hittite language because it was imposed on them.

Moreover, I am actually seeing a minor, but persistent, signal of steppe ancestry in one of the two Old-Hittite Period (~1750–1500 BCE) samples: Anatolia_MLBA MA2203. Indeed, I can put together very coherent, chronologically sound models using a couple of different methods to demonstrate this. Below is a fairly decent qpAdm model.

Anatolia_MLBA_MA2203
Anatolia_EBA 0.794±0.073
Ukraine_Eneolithic_I6561 0.206±0.073
tail: 0.400704
Full output

Obviously, these numbers aren't exactly impressive. But if the signal is real, then it might be an indication of things to come when someone manages to sequence at least a few genomes from confirmed Hittite remains. None of the other Anatolia_MLBA individuals, three of whom are from the Assyrian Colony Period (~2000–1750 BCE), show such obvious steppe ancestry.

Anatolia_MLBA_MA2200
Anatolia_EBA 1.000
Ukraine_Eneolithic_I6561 0.000
tail: 0.449485
Full output

Anatolia_MLBA_MA2205
Anatolia_EBA 0.983±0.069
Ukraine_Eneolithic_I6561 0.017±0.069
tail: 0.618499
Full output

Anatolia_MLBA_MA2206
Anatolia_EBA 0.868±0.089
Ukraine_Eneolithic_I6561 0.132±0.089
tail: 0.708811
Full output

Anatolia_MLBA w/o MA2203
Anatolia_EBA 1.000
Ukraine_Eneolithic_I6561 0.000
tail: 0.286377
Full output

In any case, apart from all of that, Damagaard et al. do take a measured and sober approach to interpreting their archaeogenetic data in the context of the Indo-European homeland debate. The paper also includes a very thorough linguistic supplement, freely available here, which reveals that there is Eastern European Hunter-Gatherer (EHG) ancestry in soon to be published Maykop culture samples. From the supplement (emphasis is mine):

Despite a general agreement on a Pontic-Caspian origin of the Anatolian Indo-European language family, it is currently impossible to determine on linguistic grounds whether the language reached Anatolia through the Balkans in the West (Anthony 2007; Mallory 1989: 30; Melchert 2003; Steiner 1990; Watkins 2006: 50) or through the Caucasus in the East (Kristiansen 2005: 77; Stefanini 2002; Winn 1981). From their earliest attestations, the Anatolian languages are clustered in Anatolia, and if the distribution reflects a prehistoric linguistic speciation event (as argued by Oettinger 2002: 52), then it may be taken as an indication that the arrival and disintegration of Proto-Anatolian language took place in the same area (Steiner 1981: 169). However, others have reasoned that the estimated period between the dissolution of the Proto-Anatolian language and the attestation of the individual daughter languages is extensive enough to allow for prehistoric mobility within Anatolia, theoretically leaving plenty of time for secondary East-to-West dispersals (cf. Melchert 2003: 25).

Whatever the case may be, there are no linguistic indications for any mass migration of steppe-derived Anatolian speakers dominating or replacing local populations. Rather, the Anatolian Indo-European languages appear in history as an organically integrated part of the linguistic landscape. In lexicon, syntax, and phonology, the second millennium languages of Anatolia formed a convergent, diffusional linguistic area (Watkins 2001: 54). Though the presence of an Indo-European language itself demonstrates that a certain number of speakers must have entered the area, the establishment of the Anatolian Indo-European branch in Anatolia is likely to have happened through a long-term process of infiltration and acculturalization rather than through mass immigration or elite dominance (Melchert 2003: 25). Furthermore, the genetic results presented in Damgaard et al. 2018 show no indication of a large-scale intrusion of a steppe population. The EHG ancestry detected in individuals associated with both Yamnaya (3000–2400 BCE) and the Maykop culture (3700–3000 BCE) (in prep.) is absent from our Anatolian specimens, suggesting that neither archaeological horizon constitutes a suitable candidate for a “homeland” or “stepping stone” for the origin or spread of Anatolian Indo-European speakers to Anatolia. However, with the archaeological and genetic data presented here, we cannot reject a continuous small-scale influx of mixed groups from the direction of the Caucasus during the Chalcolithic period of the 4th millennium BCE.

...

Under the “Steppe Hypothesis,” the Indo-Iranian languages are not seen as indigenous to South Asia but rather as an intrusive branch from the northern steppe zone (cf. Anthony 2007: 408–411; Mallory 1989: 35–56; Parpola 1995; Witzel 1999, 2001). Important clues to the original location and dispersal of the Indo-Iranians into South and Southwest Asia are provided by the Indo-Iranian languages themselves.

The Indo-Aryan and Iranian languages share a common set of etymologically related terms related to equestrianism and chariotry (Malandra 1991). Since it can be shown that this terminology was inherited from their Proto-Indo-Iranian ancestor, rather than independently borrowed from a third language, the split of this ancestor into Indo-Aryan and Iranian languages must postdate these technological innovations. The earliest available archaeological evidence of two-wheeled chariots is dated to approximately 2000 BCE (Anthony 1995; Anthony and Ringe 2015; Kuznetsov 2006: 638–645; Teufer 2012: 282). This offers the earliest possible date so far for the end of Proto-Indo-Iranian as a linguistic unity. The reference to a mariannu in a text from Tell speakers. Leilān in Syria discussed below pushes the latest possible period of Indo-Iranian linguistic unity to the 18th century BCE.

...

The traces of early Indo-Aryan speakers in Northern Syria positions the oldest Indo-Iranian speakers somewhere between Western Asia and the Greater Punjab, where the earliest Vedic text is thought to have been composed during the Late Bronze Age (cf. Witzel 1999: 3). In addition, a northern connection is suggested by contacts between the Indo-Iranian and the Finno-Ugric languages. Speakers of the Finno-Ugric family, whose antecedent is commonly sought in the vicinity of the Ural Mountains, followed an east-to-west trajectory through the forest zone north and directly adjacent to the steppes, producing languages across to the Baltic Sea. In the languages that split off along this trajectory, loanwords from various stages in the development of the Indo-Iranian languages can be distinguished: 1) Pre-Proto-Indo-Iranian (Proto-Finno-Ugric *kekrä (cycle), *kesträ (spindle), and *-teksä (ten) are borrowed from early preforms of Sanskrit cakrá- (wheel, cycle), cattra- (spindle), and daśa- (10); Koivulehto 2001), 2) Proto-Indo-Iranian (Proto-Finno-Ugric *śata (one hundred) is borrowed from a form close to Sanskrit śatám (one hundred), 3) Pre-Proto-Indo-Aryan (Proto-Finno-Ugric *ora (awl), *reśmä (rope), and *ant- (young grass) are borrowed from preforms of Sanskrit ā́ r ā- (awl), raśmí- (rein), and ándhas- (grass); Koivulehto 2001: 250; Lubotsky 2001: 308), and 4) loanwords from later stages of Iranian (Koivulehto 2001; Korenchy 1972). The period of prehistoric language contact with Finno-Ugric thus covers the entire evolution of Pre-Proto-Indo-Iranian into Proto-Indo-Iranian, as well as the dissolution of the latter into Proto-Indo-Aryan and Proto-Iranian. As such, it situates the prehistoric location of the Indo-Iranian branch around the southern Urals (Kuz’mina 2001).

Citation...

Guus Kroonen, Gojko Barjamovic, & Michaël Peyrot. (2018). Linguistic supplement to Damgaard et al. 2018: Early Indo-European languages, Anatolian, Tocharian and Indo-Iranian. http://doi.org/10.5281/zenodo.1240524

Update 14/05/2018: I managed to, more or less, reproduce my qpAdm models with qpGraph. This is never a simple and easy task, so I'm now more confident that Anatolia_MLBA MA2203 really does harbor ancestry from the steppe.




See also...

Likely Yamnaya incursion(s) into Northwestern Iran

Graeco-Aryan parallels

Late PIE ground zero now obvious; location of PIE homeland still uncertain, but...

193 comments:

Dmytro said...

"However, with the archaeological and genetic data presented here, we cannot reject a continuous small-scale influx of mixed groups from the direction of the Caucasus during the Chalcolithic period of the 4th millennium BCE."

Does this also work for them if one substitutes "the Balkans" for "the Caucasus"?

Davidski said...

Probably not, otherwise they would've said.

Open Genomes said...

MA2200 and MA2203 both date to the same stratum at Kalehoyuk, Stratum IIIb, the Old Hittite Period.

The test here is against Bronze Age and earlier populations, excluding the Bronze Age Anatolians, Minoans, and Mycenaeans.

Here's the restricted nMonte3 for MA2200, which shows continuity with the earlier populations. Notice the Armenia EBA, Hajji Firuz (modal) Chalcolithic, and Balkan Neolithic alongside early Anatolians. There seems to be post-Neolithic admixture from further east:

Sample ID: Anatolia_MLBA:MA2200
nMonte3 with restricted nMonte3
Source data table: Global25

Total number of samples compared: 1006

Number of cycles: 1000
Batch size: 500
Distance penalty: 0.001
Restricted nMonte minimum percent limit 1%

Full nMonte populations >= 1% limit: 10

Restricted nMonte distance: 1.09930

Restricted nMonte percentages:
24.6% Anatolia_ChL
20.8% Armenia_EBA
14.6% Tepecik_Ciftlik_N
10.6% Peloponnese_N_o
10.2% Peloponnese_N
 9.8% Hajji_Firuz_ChL
 9.4% Krepost_N

Nearest Item Distances:
1.8819 Anatolia_ChL:I1584
3.1146 Tepecik_Ciftlik_N:Tep003
3.1537 Krepost_N:I0679_d
3.2309 Peloponnese_N_o:I3920
3.2543 Armenia_ChL:I1632
3.6110 Hajji_Firuz_ChL:I4351
3.7596 Vucedol:I3499
3.7626 Armenia_ChL:I1407

Here's the exact same test for MA2203.
An exploratory analysis suggests that MA2203 is of Balkan origin. N. Italian Beaker may be a proxy for Balkan admixture. Further studies and samples can test this hypothesis:

Sample ID: Anatolia_MLBA:MA2203
nMonte3 with restricted nMonte3
Source data table: Global25

Total number of samples compared: 1006

Number of cycles: 1000
Batch size: 500
Distance penalty: 0.001
Restricted nMonte minimum percent limit 1%

Full nMonte populations >= 1% limit: 7

Restricted nMonte distance: 1.83790

Restricted nMonte percentages:
30.8% Tepecik_Ciftlik_N
25.6% Peloponnese_N_o
24.4% Balkans_BA
14.6% Maros
 4.6% Beaker_Northern_Italy

Nearest Item Distances:
2.6691 Peloponnese_N_o:I3920
2.9989 Tepecik_Ciftlik_N:Tep003
3.0287 Anatolia_ChL:I1584
3.0531 Balkans_BA:I4331
3.1745 Krepost_N:I0679_d
3.2949 Maros:RISE374
3.3177 Beaker_Northern_Italy:I1979
3.4368 Hungary_BA:I7041

Rob said...

As great as the paper is, it’s still working within the framework of the flawed steppe model
Hence their explanation of “mysterious wafting language expansion” to explain Anatolian - instead of following the data.
Btw Dave, how much do you want to bet that future Hitties also don’t have steppe ?

Rob said...

Btw, look at the mtDNA of the chiefs - C1. Probably minor at best steppe admixture from the east, not Ukraine eneolithic.

Arza said...

At least in G25 MA2203 takes also a lot of Central European neolithic ancestry:

_Anatolia_MLBA:MA2200
Anatolia_EBA_av 98%
LBKT_MN 2%
Ukraine_Eneolithic:I6561 0%
Distance 2.3423%

_Anatolia_MLBA:MA2203
Anatolia_EBA_av 74%
LBKT_MN 17.6%
Ukraine_Eneolithic:I6561 8.4%
Distance 3.0597%

_Anatolia_MLBA:MA2205
Anatolia_EBA_av 100%
Distance 3.0855%

_Anatolia_MLBA:MA2206
Anatolia_EBA_av 98.6%
LBKT_MN 1.4%
Distance 2.6512%

An angle worth exploring IMHO.


@ Rob
It's the PIE reconstruction that is flawed, not the steppe model.

Arza said...

@ Rob
Btw, look at the mtDNA of the chiefs - C1

Oh come on, this would be _eventually_ valid if he would have 50% steppe.

gR1andpa comes from the steppe, marry your C1 lady and has a daughter. Daughter marry local J-I Joe and voila J-C1 combo with steppe ancestry transmitted via R1 guy.

Open Genomes said...

Here's the same test for MA2205, from Kalehoyuk Stratum IIIc, the Assyrian Colony Period. MA2205 died in the battle for the pubic building in 1750 BCE. Here we can see that like the later MA2200, MA2205 has some "Eastern" post-Neolithic admixture alongside the native earlier Anatolian ancestry. Significantly, both MA2200 and MA2205 are in the same low-level Y haplogroup J-Z6271 (J2a) and appear to have the same ancestry, regardless of the fact that MA200 is from the later Old Hittite Period:

Sample ID: Anatolia_MLBA:MA2205
nMonte3 with restricted nMonte3
Source data table: Global25

Total number of samples compared: 1006

Number of cycles: 1000
Batch size: 500
Distance penalty: 0.001
Restricted nMonte minimum percent limit 1%

Full nMonte populations >= 1% limit: 10

Restricted nMonte distance: 1.33130

Restricted nMonte percentages:
28.6% Tepecik_Ciftlik_N
20.2% Hajji_Firuz_ChL
15.0% Armenia_EBA
12.6% Peloponnese_N_o
12.2% Anatolia_ChL
 6.6% Levant_BA
 4.8% Armenia_MLBA

Nearest Item Distances:
2.6397 Anatolia_ChL:I1584
3.1976 Tepecik_Ciftlik_N:Tep003
3.4254 Armenia_ChL:I1632
3.4390 Hajji_Firuz_ChL:I4241
3.5355 Hajji_Firuz_ChL:I4351
3.6898 Armenia_ChL:I1407
3.7583 Peloponnese_N_o:I3920
3.8279 Armenia_EBA:I1658

Open Genomes said...

MA2206 was a woman who died in the attack on the public building at the end of the Assyrian Colony Period. She was found beneath MA2208 who is Y G-Z41194 (in G2a2b1-M406). They didn't perish in the fire, and two apparent attackers were found dead in the doorway of the same room. Gedmatch shows she isn't related to MA2208.
(Gedmatch vs. Z643830=MA2206 vs. Z282612=MA2208)

She also appears to have similar ancestry to MA2200 from the Old Hittite Period, and MA2205 who died in another room of the public building, but perhaps even more "Eastern" ancestry.

Are MA2200, MA2205, and MA2206 Hurrians, or even Assyrians?

Sample ID: Anatolia_MLBA:MA2206
nMonte3 with restricted nMonte3
Source data table: Global25

Total number of samples compared: 1006

Number of cycles: 1000
Batch size: 500
Distance penalty: 0.001
Restricted nMonte minimum percent limit 1%

Full nMonte populations >= 1% limit: 14

Restricted nMonte distance: 1.21970

Restricted nMonte percentages:
18.6% Hajji_Firuz_ChL
14.8% Tepecik_Ciftlik_N
12.4% Anatolia_ChL
11.2% Levant_BA
10.8% Armenia_EBA
10.0% Peloponnese_N_o
 8.6% Peloponnese_N
 8.0% Krepost_N
 5.6% Armenia_ChL

Nearest Item Distances:
2.4156 Anatolia_ChL:I1584
3.0405 Tepecik_Ciftlik_N:Tep003
3.2222 Armenia_ChL:I1632
3.3722 Peloponnese_N_o:I3920
3.4299 Krepost_N:I0679_d
3.5855 Hajji_Firuz_ChL:I4351
3.6059 Armenia_ChL:I1407
3.7130 Hajji_Firuz_ChL:I4241

Open Genomes said...

Anatolian Early Bronze Age (c. 2500 BCE) MA2212 from Ovaoren (Nevesehir) appears to be a straight-up mix between Early Neolithic of the Anatolian-Balkan variety and various "Eastern" populations, which also apparently includes something related to BMAC. This may not be a coincidence, because his Y J-Y37496 (in J2a1-L26, like the others except for M2208 in G-M406) is in J-Y7687, which has another subclade, J-M47, which is very common in South Asia.

Sample ID: Anatolia_EBA:MA2212
nMonte3 with restricted nMonte3
Source data table: Global25

Total number of samples compared: 1006

Number of cycles: 1000
Batch size: 500
Distance penalty: 0.001
Restricted nMonte minimum percent limit 1%

Full nMonte populations >= 1% limit: 9

Restricted nMonte distance: 1.56620

Restricted nMonte percentages:
22.6% Krepost_N
12.6% Peloponnese_N_o
11.8% Hajji_Firuz_ChL
10.4% Barcin_N
 9.4% Armenia_EBA
 9.4% LchashenMetsamor
 9.0% Levant_BA
 8.0% Dzharkutan2_BA
 6.8% Tepecik_Ciftlik_N

Nearest Item Distances:
3.1370 Krepost_N:I0679_d
3.3115 Peloponnese_N_o:I3920
3.6377 Anatolia_ChL:I1584
3.7939 Hajji_Firuz_ChL:I4241
3.8271 Tepecik_Ciftlik_N:Tep003
4.0815 Beaker_Northern_Italy:I1979
4.0954 Balkans_BA:I4332
4.1264 Armenia_ChL:I1632

Samuel Andrews said...

Armenians have basically no Steppe ancestry, according to G25. That backs up the idea PIE emerged in the Caucasus.

But I don't trust these new Anatolian genomes were IE-speakers. We'll have to wait for confirmed IE Anatolians.

The Steppe ancestry in Mycenenans is very real and super significant. I think it confidently confirms Greek is from the Steppe aswell as Celtic, Germanic, Italic, Balto-Slavic, and Indo Iranian.

It's hard to imagine IE expansion moved a tiny-winy bit in 4500 BC from Caucasus to Steppe, then independently went all over Eurasia in 2600-1600 BC. Still, think it's very possible Hittite derives from the same widespread Steppe movements.

epoch2013 said...

@Rob

Ma2200 has mtDNA K1a and M2203 has mtDNA J1c. It is in Supp Inf, page 79, table S15.

John Johnson said...

Interesting though not surprised 'Anatolian' would cause such a problem. Just read what Kroonen wrote. Here is what's key:

'Our findings corroborate the Indo-Anatolian Hypothesis, which claims
that Anatolian Indo-European split off from Proto-Indo-European first and that Anatolian Indo-European represents a sister rather than a daughter language. Our findings call for the identification of the speakers of Proto-Indo-Anatolian as a population earlier that the Yamnaya and late Maykop culture.'

Anthony 2007 of course takes a stab at explaining Hittite's and related IE-Anatolian language's peculiar linguistic position to PIE in the section entitled 'Migrations into the Danube Valley: The Suvorovo Novodanilikovka Complex'. Accordingly, the IE's that became 'Anatolian' break away from Sredny Stog c. 4200 BCE and Pre-Anatolian languages would then be introduced into the lower Danube Valley at around 4200-4000 BCE. Partly explains some of the linguistic peculiarities within the IE components of the Anatolian family because the break was so early in time as Yamnaya is usually considered the actual time for PIE.

Aside from the aDNA complications, these two samples do indeed appear to be, based on the archaeology,'commoners', as has been mentioned here but also on other boards. Considering that the Hittites/Nes represent an example of elite, coupled with Anthony's hypothetical early split from the Pontic Caspian Steppes, with the fact that Anatolia was prior significantly populated with G2a males stemming from the ENF days then its perhaps not that surprising that these two samples would yield next to no evidence of steppe admixture.

But again, given Anthony's hypothetical early split, what from a genetic standpoint would or should we expect to see during the Hittite period among the elites if we ever get such a sample? Of course, this with consideration that a scenario similar to Anthony's is correct.

Samuel Andrews said...

Something to keep in mind. If PIE originated in the Caucasus then moved up into the Steppe, you'd expect the Steppe to have a diverse collection of IE speakers. Instead, it looks like the R1a M417 and R1b L23 groups derive from the same source. Not just because their Steppe ancestro had the exact same ratio of CHG/EHG but drift not explained by current EHG-CHG genomes as well as shared mtDNA.

Ric Hern said...

We see Steppe people moved early on into the Northern Caucasus Fortresses prior to Yamnaya. Which Steppe Culture could this be ? We see the domesticated horse arrive in Azerbaijan during this period. We see a Westward, Southwestwards spread from there into Anatolia. Then we see R1b in Hajji Firuz...

So I still think Proto-Hittites followed this route maybe establishing a trade route and not expanding far from the traderoute during the early stages....

Maybe the Steppe people who took over the Fortresses in the Northern Caucasus pushed those "Hittite"(Minoan Relatives) people steadily Southwards and they eventually ended up in Central Anatolia with actual Hittites established to their East Northeast.

Davidski said...

@All

I added a couple of qpGraph trees to the post. They seem to match the qpAdm output and do make sense, so I don't think this is all just a fluke.

Ric Hern said...

@ Davidski

Thank you.

Huck Finn said...

@Davidski: Very good point. Genetic studies simply can't debunk linguistic results, it is not possible.

Open Genomes said...

@John Johnson

What we might expect to see from David Anthony's hypothesis about pre-Yamnaya Sredny Stog migrating to the Lower Danube Valley around 4200-4000 BCE to form Anatolian is what we actually see with MA2203, a Hittite Era Anatolian who appears to be mostly Balkan, and who has extremely distant connections to the steppe via the Balkans, but no direct ancestry from Yamnaya or from anything around the Caucasus. He's substantially Neolithic because his steppe ancestry is so distant and diluted with Balkan Neolithic.

It's pretty easy to prove that MA2203's ancestry isn't directly from Yamnaya. In this restricted nMonte3, MA2203 doesn't come up as having any direct steppe ancestry, as compared with Anatolian. The run showing
Balkan ancestry is a better fit, as well, with a distance of 1.83790 vs. 1.98050 for this run:

Sample ID: Anatolia_MLBA:MA2203
nMonte3 with restricted nMonte3
Source data table: Global25

Total number of samples compared: 60

Number of cycles: 1000
Batch size: 500
Distance penalty: 0.001
Restricted nMonte minimum percent limit 1%

Full nMonte populations >= 1% limit: 11

Restricted nMonte distance: 1.98050

Restricted nMonte percentages:
76.4% Anatolia_EBA
17.4% Barcin_N
 6.2% Tepecik_Ciftlik_N

Nearest Item Distances:
2.5465 Anatolia_EBA:MA2213
2.6985 Anatolia_EBA:MA2210
2.9989 Tepecik_Ciftlik_N:Tep003
3.0739 Anatolia_EBA:MA2212
4.3664 Tepecik_Ciftlik_N:Tep002
4.4642 Barcin_N:I0708
4.5799 Barcin_N:I1103
4.5987 Barcin_N:I1098

Full nMonte distance: 1.81820

Full nMonte percentages:
57.6% Anatolia_EBA
24.0% Barcin_N
 9.8% Tepecik_Ciftlik_N
 3.8% Poltavka
 2.8% Srubnaya_MLBA
 2.0% Yamnaya_Samara

Source populations:

Barcin_N
EHG
Poltavka
Srubnaya_MLBA
Tepecik_Ciftlik_N
Ukraine_N
Yamnaya_Samara
Anatolia_EBA

Simon_W said...

I wouldn't say the Hittites imposed their language onto the Hattians. Because the Hittites held Hattic in high regard, they even wrote Hattic texts for ritual purposes, to venerate Hattic gods. So the mechanism behind the extinction of Hattic must be somewhat different.

Davidski said...

@Open Genomes

So realistically, what are the chances that all of these cultures met in Central Anatolia during the Early to Middle Bronze Age?

Barcin_N
EHG
Poltavka
Srubnaya_MLBA
Tepecik_Ciftlik_N
Ukraine_N
Yamnaya_Samara
Anatolia_EBA

I'd say none.

When modeling, it's useful to be as realistic as possible, both in terms of the genetics and chronology, and not mix distal and proximate sources too much, unless absolutely necessary.

old europe said...



Open genomes

Now Sredni stog has more or less 30% EEF and now we are told that is considered the PIE homeland........we have only ANF in Hittite ( to be confirmed) we have 25% EEF in Andronvo.....mainland Europe remained overall with an EEF WHG majority....considering the fact that SS is mainly a cultural offshot of Cucuteni ( highly likely they spoke a farmer related language) it is not weird to think that PIE emerged inside farmers culture. Taking in consideration that archeology fully support this thesis now even genetics is showing that it is possible.

Davidski said...

@Samuel Andrews

Armenians have basically no Steppe ancestry, according to G25. That backs up the idea PIE emerged in the Caucasus.

https://drive.google.com/file/d/1gqCmmLc4Ll080VCV8-MNDMfKI3k1oSV_/view?usp=sharing

The new Iron Age Armenian samples from Lchashen Metsamor also show steppe ancestry.

@old europe

We have only ANF in Hittite.

I call bullshit on that, but I'm sure you'll just believe what you want to believe.

Seinundzeit said...

Anthro Survey,

I know you'll get a kick out of this...

These models belong in the previous thread. Unfortunately, now that I finally have time to fiddle around, I see that the Graeco-Aryan comment stream has become an unnavigable behemoth. So, I guess it’s best to continue discussion here.

Anyway, long story short:

The peoples of Greater India, the ancient samples from the Swat valley, and the Kalasha (an Indo-Aryan relic) all prefer Sintashta, with a very consistent Botai signal (probably an indication that Steppe_MLBA_East is the real source. But in the case of the Swat valley ancients, and with the Kalasha, the dynamics might be different).

By contrast, Pashtuns of various stripes, Pamiris, Yaghnobis, and Tajiks all prefer the Kushan-related or White Hun-related (Wusun and Kangju) ancient populations, with some Sintashta_MLBA (most likely a remnant of the older Indo-Aryan layer). (I might be wrong about whom the Wusun and Kangju were, in terms of historical identification. Even if so, the point remains unchanged)

All of this ties in beautifully with the linguistic situation, and seems to prove the age-old speculation that Pashtun ethnogenesis was intimately connected to the later Iranic populations of the steppe.

Furthermore, and as I expected, the Swat Valley folk have much more Botai-related than any population in South Asia proper.

For example, Arkotkila_IA shows 10%! The Sintashta signal is still there, and is obviously real, but the excess of West_Siberian_N (noted by RK and Matt) is also real.

In addition, noticeable BMAC and Jiroft-related admixture is confirmed for the ancient Swat valley samples.

Finally, I get a pretty cool model for myself.

Output is posted below...

Simon_W said...

In the entire northwestern quarter of Anatolia there is no evidence for Anatolian IE:
https://de.wikipedia.org/wiki/Datei:Anatolisch.png

Hence the site of Kaman-Kalehöyük, although in central Anatolia, is in fact at the border, at the fringe of the distribution area of Anatolian IE languages. This might explain the presence of individuals without Steppe ancestry in the context of the Steppe theory.

For the steppe theory itself it's certainly good to see at least two of the Anatolia_MLBA individuals having some steppe admixture.

Because the model proposed by Gamkrelidse/Ivanov isn't completely dead, it seems. It was certainly illusionary by David to call it even less likely than Renfrew's Anatolian theory. Because the Anatolian theory is as dead as it can get, like a Dodo. In fact Gamkrelidse/Ivanov suggested a homeland not in the modern state of Armenia, but rather around lake Van and lake Urmia, and that's close to Hajji_Firuz.
And as it happens... (using nMonte 1.0 with scaled data)

[1] "distance%=9.3062 / distance=0.093062"


Yamnaya_Kalmykia
"Ukraine_Eneolithic" 48.7
"Hajji_Firuz_BA" 36.95
"CHG" 14.35
"Armenia_ChL" 0
"Armenia_EBA" 0
"Hajji_Firuz_ChL" 0
"Anatolia_EBA" 0
"Trypillia" 0

We don't know for sure what type of ancestry Hattic was associated with, Barcin_N or Hajji_Firuz.

Chetan said...

@OpenGenomes Yep steppe deficient Balkan CA +Anatolia_EBA


Seinundzeit said...

Here’s the output…

Ancient Northwestern Fringe of Hind (Swat valley):

Saidu_Sharif_IA

47.8% Shahr_I_Sokhta_BA3 (60%_West_Eurasian+40%_AASI)
21.1% Shahr_I_Sokhta_BA1 (Jiroft) + 5.4% Sappali_Tepe_BA (BMAC)
15.6% Sintashta_MLBA
4.3% Botai
3.3% Hajji_Firuz_ChL
2.5% XiongNu (DA43 + DA45)

Distance=1.082

Udegram_IA
25.7% Shahr_I_Soktha_BA1 (Jiroft) + 16.5% Sappali_Tepe_BA (BMAC)
35.7% Shahr_I_Soktha_BA3 (60%_West_Eurasian+40%_AASI)
13.2% Sintashta_MLBA
6.9% Botai
2% XiongNu (DA43 + DA45)

Distance=1.0726

Arkotkila_IA

25.3% Sappali_Tepe_BA (BMAC) + 25.1% Shahr_I_Soktha_BA1 (Jiroft)
30.9% Shahr_I_Soktha_BA3 (60%_West_Eurasian+40%_AASI)
10% Botai
6.7% Sintashta_MLBA
2% XiongNu (DA43 + DA45)

Distance=3.2893

Contemporary northern Hind (Upper Caste):

Brahmin

68.2% Shahr_I_Soktha_BA3 (60%_West_Eurasian+40%_AASI) + 1.7% AASI_Ghost
20% Sintashta_MLBA + 2.2% Botai + 0.1% Sappali_Tepe_BA (BMAC)
7.8% Hajji_Firuz_ChL

Distance=1.6279

Kshatriya

72.5% Shahr_I_Soktha_BA3 (60%_West_Eurasian+40%_AASI) + 3% AASI_Ghost
16.5% Sintashta_MLBA + 2.3% Botai
5.7% Haji_Firuz_ChL

Distance=1.9318

(Unlike the ancient Swat valley, no real BMAC or Jiroft signal, and the Botai-related signal is small enough in comparison to Sintashta that one could chalk it up to Steppe_MLBA_East)

Indo-Aryan Central Asia:

Kalash

45.2% Shahr_I_Soktha_BA1 (Jiroft)
24.8% Sintashta_MLBA
21.4% Shahr_I_Soktha_BA3 (60%_West_Eurasian+40%_AASI)
6.7% Botai
1.9% XiongNu (DA43 + DA45)

Distance=2.179

(Unlike India, and like the ancient Swat valley, the Botai signal seems big enough in comparison to Sintashta_MLBA, and there is a minor East Asian signal.

Though, much lower AASI (via Shahr_I_Soktha_BA3), and higher steppe, probably because the Kalash are from just barely outside the boundaries of South Asia, while Swat is just barely inside the boundaries of South Asia)

Kho_Singanali

35.2% Wusun + 7.9% Sintashta_MLBA
24% Shahr_I_Soktha_BA1 (Jiroft) + 0.3% Sappali_Tepe_BA (BMAC)
23.6% Shahr_I_Soktha_BA3 (60%_West_Eurasian+40%_AASI) + 0.5% AASI_Ghost
4.2% Hajji_Firuz_ChL
3.4% Botai
0.9% XiongNu (DA43 + DA45)

Distance=1.7393

(Interestingly, seems that the Kho are quite distinct from the Kalasha, and much more similar to their Iranic neighbors.

No surprise though, considering that the Kalasha are an isolate, while the Kho are a very cosmopolitan population on a well-trodden path from Central Asia into northern South Asia, and thus open to steppic influences after the Indo-Aryans.

Once David has the Kohistani samples, we shall see if those Dards follow the Kalasha pattern, or are similar to Pashtuns in the same way as the Kho)

To be continued...

Seinundzeit said...

Continuing from where we left off...

Iranic Central Asia:

Tajik

43.5% Kangju + 15.5% Sintashta_MLBA
23.5% Sappali_Tepe_BA (BMAC) + 7.2% Shahr_I_Sokhta_BA1 (Jiroft)
5.4% Shahr_I_Sokhta_BA3 (60%_West_Eurasian+40%_AASI)
4.2% Botai
0.5% Karakhanid + 0.2% XiongNu (DA43 + DA45)

Distance=1.8362

Eastern_Pukhtun

22.6% Shahr_I_Soktha_BA1 (Jiroft) + 11.9% Sappali_Tepe_BA (BMAC)
18.4% Sintashta_MLBA + 7.4% Wusun + 2.7% Kangju + 2.1% Botai
28.7% Shahr_I_Soktha_BA3 (60%_West_Eurasian+40%_AASI)
3.4% Hajji_Firuz_ChL
1.8% Karakhanid + 1% XiongNu (DA43 + DA45)

Distance=1.1249

Western_Pashtun

20.9% Sappali_Tepe_BA (BMAC) + 14.9% Shahr_I_Soktha_BA1 (Jiroft)
21.5% Sintashta_MLBA + 11.4% Wusun
18% Shahr_I_Soktha_BA3 (60%_West_Eurasian+40%_AASI)
7.9% Hajji_Firuz_ChL
3.8% XiongNu (DA43 + DA45) + 1.6% Karakhanid

Distance=1.1832

SeinundZeit

31% Kangju + 5% Botai
31.2% Sappali_Tepe_BA (BMAC) + 3% Shahr_I_Soktha_BA1 (Jiroft)
26% Shahr_I_Soktha_BA3 (60%_West_Eurasian+40%_AASI)
3.1% Hajji_Firuz_ChL
0.7% XiongNu (DA43 + DA45)

Distance=1.9016

I know I’m not supposed to take this literally. But I’m only human, so naturally this model makes me want to read up on the Kangju.

Anyway, once David adds the Yusufzai, Tarklani, and Uthmankhel, we shall see how they fit within Pashtun variation.

Also, I would like to thank Alberto for sending me his Xmix files; made life so much easier.

Open Genomes said...

@David

The idea of that population list was to see if MA2203 preferred any of the steppe populations over earlier Anatolians. It's not intended to be an absolute model of MA2203's ancestry.

I did the same thing here with Hajji_Firuz_BA:I4243 (mtDNA I1b) except that the "local" ancestral populations are from Iran and Central-Eastern Anatolia. I4243 lived at least 500 years before MA2203 (2400-2100 BCE). We can see from this simple test that I4243, unlike the contemporary and later Anatolians, has very substantial steppe ancestry. We can even tell that this ancestry is more like Poltavka/Srubnaya than Yamnaya. The test works well in this case:

Sample ID: Hajji_Firuz_BA:I4243
nMonte3 with restricted nMonte3
Source data table: Global25

Total number of samples compared: 45

Number of cycles: 1000
Batch size: 500
Distance penalty: 0.001
Restricted nMonte minimum percent limit 1%

Full nMonte populations >= 1% limit: 19

Restricted nMonte distance: 1.09950

Restricted nMonte percentages:
31.4% Srubnaya_MLBA
18.0% Hajji_Firuz_ChL
17.8% Poltavka
16.0% Yamnaya_Samara
13.2% Seh_Gabi_ChL
 3.6% Tepecik_Ciftlik_N

Nearest Item Distances:
3.1458 Srubnaya_MLBA:I0422
3.2570 Srubnaya_MLBA:I0232
3.2925 Srubnaya_MLBA:I0361
3.3659 Srubnaya_MLBA:I0424
3.4725 Srubnaya_MLBA:I0234
3.5048 Srubnaya_MLBA:I0235
3.7619 Srubnaya_MLBA:I0359
3.8775 Poltavka:I0371

Full nMonte distance: 1.15800

Full nMonte percentages:
34.8% Srubnaya_MLBA
18.0% Poltavka
15.0% Hajji_Firuz_ChL
14.2% Yamnaya_Samara
13.8% Seh_Gabi_ChL
 4.2% Tepecik_Ciftlik_N

Source populations:

EHG
Hajji_Firuz_ChL
Poltavka
Seh_Gabi_ChL
Srubnaya_MLBA
Tepecik_Ciftlik_N
Ukraine_N
Yamnaya_Samara

Folker said...

David & Open Genomes

Can you try to use YamnayA Bulgarian as a source for the Steppe amdixture?

Davidski said...

@Folker

Using qpAdm, the model is good enough, but there's not much data to work with and the statistical fit is worse than in my other model above.

https://drive.google.com/file/d/10vpuscZiYXwCfM8Hcp4XP2_Dh0boA3e0/view?usp=sharing

epoch2013 said...

@David and Folker

And while you at it, use the Hajji Firuz R1b outlier as well. Just to bum around ;)

Folker said...

@Dave
Thx.
That's what I feared. Not enough data.

Rob said...

Haha yes Floker your fears came true - your outdated theory has been shown to be...outdated !
So please tell us again about your invisible elite migrations c. 2200 BC ?

Davidski said...

@epoch

Hajji_Firuz_ChL_o:I2327 doesn't work. Hajji_Firuz_BA:I4243 does work, but the fit isn't quite as good as that with Ukraine_Eneolithic:I6561, which is interesting.

https://drive.google.com/file/d/13FeGChIIAUQlsC2RR6_3gO16nv2yyvLx/view?usp=sharing

It seems to me like the key to a good fit for Anatolia_MLBA:MA2203 is gene flow from Ukraine, not just the Pontic-Caspian steppe.

What we have here are probably just Hattians, with one that has somehow ended up with minor Hittite ancestry.

epoch2013 said...

@Rob

"But Yamnaya Bulgaria does work for # 2203"

So the steppe theory is not dead. By your own data.

epoch2013 said...

@David

"It seems to me like the key to a good fit for Anatolia_MLBA:MA2203 is gene flow from Ukraine, not just the Pontic-Caspian steppe"

That points to a Balkan route. If Anatolian is pre-Yamnaya, then what culture should we look at? Suvorovo? It was based at the Danube and had zoomorphic scepters.

Rob said...

@ Epoch

10% Yamnaya Bulgaria. Side flow , female mediated. Marginal . For some reason raw EHG is zero
The serious contenders at this stage are Balkans or Caucasus . EHG is out

Rob said...

There's also about 10% Yamnaya Bulgaria in some Mycenean females. Probably came with raw materials from the steppe ? Drought-stricken steppe dwellers trying to improve their lot by marrying off their daughters?

Rob said...

@ Ric Hern
You you and Anthro S managed to find that CHG refuge in the middle of the Volga yet ?

Open Genomes said...

@Folker

Here's MA2203 with Yamnaya Bulgaria and some other populations. Notice that Maros provides a very good fit:

Sample ID: Anatolia_MLBA:MA2203
nMonte3 with restricted nMonte3
Source data table: Global25

Total number of samples compared: 30

Number of cycles: 1000
Batch size: 500
Distance penalty: 0.001
Restricted nMonte minimum percent limit 1%

Full nMonte populations >= 1% limit: 7

Restricted nMonte distance: 1.65030

Restricted nMonte percentages:
43.8% Tepecik_Ciftlik_N
31.6% Maros
24.6% Hajji_Firuz_ChL

Nearest Item Distances:
2.9989 Tepecik_Ciftlik_N:Tep003
3.2949 Maros:RISE374
4.3664 Tepecik_Ciftlik_N:Tep002
4.5877 Hajji_Firuz_ChL:I4241
4.7360 Hajji_Firuz_ChL:I2323
4.7578 Maros:RISE373
4.7640 Hajji_Firuz_ChL:I4351
5.0078 Hajji_Firuz_ChL:I4349

Full nMonte distance: 1.64030

Full nMonte percentages:
44.0% Tepecik_Ciftlik_N
33.0% Maros
22.8% Hajji_Firuz_ChL
 0.2% Poltavka

Source populations:

EHG
Hajji_Firuz_ChL
Maros
Poltavka
Tepecik_Ciftlik_N
Trypillia
Ukraine_N
Yamnaya_Bulgaria

Elliv J said...

Alot of new data. Are there any clues about the formation and migration routes for the east-iranian peoples who ended up on the Steppe?

@Rob
You are very smug about Steppe but what are your own ideas for proto-Anatolian homeland?

epoch2013 said...

@David

Those D-stats in the paper, D(EHG, Mbuti; Anatolia_EBA, Anatolia_MLBA) that rendered nothing useful (Z~2) if we replaced Anatolia_EBA with MA2203..

Mbuti Yamnaya_Samara Anatolia_EBA MA2203
Mbuti Varna_o Anatolia_EBA MA2203
Mbuti MA1 Anatolia_EBA MA2203
Mbuti AG3 Anatolia_EBA MA2203

What were the Sredny Stog sampls again?

Davidski said...

@All

I updated the Botai-steppe_papers_preliminary_G25 datasheets with new samples.

https://drive.google.com/file/d/1nPoYmzhEVtDzsYaJO9NQ4qEW_EekojRD/view?usp=sharing

https://drive.google.com/file/d/1J_AexSjjz0K8pv3GHK5v06N_5TgTjsyu/view?usp=sharing

epoch2013 said...

@David

O, stupid me. I6561 *is* a Sredny Stog sample. Wow. David Anthony named the Suvorovo culture as possible expansion of Sredny Stog and possible source of the Anatolians.

Mbuti Ukraine_Eneolithic_I656 Anatolia_EBA MA2203

Rob said...

@ Ellie
Haha if I were smug I'd be obfuscating about consensus, making bold claims about subject I had no clue about, and now be dancing around the confronting reality.
Kinda sounds like all you clowns, doesn't it ?

Le skipper de Pytheas said...

I haven't seen this paper mentionned
The evolutionary history of human populations in Europe
Iosif Lazaridis
https://arxiv.org/abs/1805.01579
quotation:
"In the next few years this lingering mystery will be solved: either Anatolian speakers will be shown to possess steppe-related ancestry absent in earlier Anatolians (largely proving the steppe PIE hypothesis), or they will not (largely falsifying it, and pointing to a Near Eastern PIE homeland)"

Davidsky' point is well taken.
But Kroonen points out to evidence for Hittite speaking individuals involved in trade with Ebla (Northern Syria) around 2500 BC. So the Hittite split date from PIE is about 3000 BC at the latest, more probably around 3500 BC. And these individuals are unlikely to be of a royal background.
It seems hard to argue for a steppe origin at these early dates....

Jaydeep said...

Rob, Alberto, Ryukendo, Nirjhar & anyone else interested,

Here is the archaeological supplement A to the Daamgard et al paper,

https://zenodo.org/record/1243933#.Wvl19aSFPIU

The Conclusion of the section on South Central Asia is worth quoting in full

This survey of the archaeological and biological record of southern Central Asia yields four important findings. First, contacts between the sedentary food-producing populations of the Namazga culture populations residing in Kopet Dagh piedmont and Geokyur oasis of southern Turkmenistan who likely established the outpost at Sarazm had little to no contact with
populations residing in the southern steppe zone. Second, contacts between Bronze Age steppe populations and NMG V and BMAC populations appears to have been one in which the dynamic of cultural influence was stronger on the side of the well-established sedentary foodproducing populations, and this resulted in the partial assimilation of these initial newcomers to the region both culturally and, to a lesser degree, biologically as well. Third, not all of those who emigrated from the north turned to farming but may have continued a semi-nomadic existence in the highlands, which were unsuitable for the kind of intensive farming practiced in the BMAC homelands or in the regions of Khorezm. Fourth, if there was any Central Asian influence on
South Asian populations, that influence likely long predated any development of Iranian, let alone Indo-Aryan, languages, and most likely occurred during the late NMG IV to early NMG V period (ca. 2800–2300 BCE) and even earlier during the Eneolithic from Kelteminar culture groups (4000–3500 BCE)
.


Davidski said...

@Le skipper de Pytheas

It seems hard to argue for a steppe origin at these early dates.

Ukraine_Eneolithic I6561, the sample that I used as the proxy for steppe ancestry in Anatolia_MLBA MA2203, is dated to 4045-3974 calBCE.

And I'm pretty sure he wasn't the first of his kind; more likely the first to be sampled thus far.

Davidski said...

@epoch

The Z scores are weak because there's not enough markers (the lower the marker count, the lower the Z score). But EHG and MA1 give the best signals, probably because they're not muffled by too much shared ancestry.

Mbuti AG3 Anatolia_EBA MA2203 0.0033 0.282 33441
Mbuti EHG Anatolia_EBA MA2203 0.0101 1.961 139765
Mbuti I6561 Anatolia_EBA MA2203 0.0073 1.027 96600
Mbuti MA1 Anatolia_EBA MA2203 0.0123 1.765 103735
Mbuti Varna_o Anatolia_EBA MA2203 -0.0042 -0.476 54094
Mbuti Yamnaya_Samara Anatolia_EBA MA2203 -0.0003 -0.066 143983

Folker said...

@Open Genomes

Thanks.

So Maros and Bulgarian Yamna are fitting (even if BY is not the best fit).

So it seems that some LN/EBA Balkans admixture could be needed.

There were at least 3 different Yamna groups in Bulgaria with specificities. Plus others nearby. And their datation is older than 3000BC (the first are dated from around 3100 BC), so one sample is not enough.

Anyway, Kaman Kalehoyuk samples are very likely Hattians. Not the best choice to have an insight of Hittites genomes.


John Johnson said...

@ Open Genomes

Good point. Yes you'd expect highly diluted steppe ancestry with substantial Balkan if that migration route is correct plus early in time. Concerning this, I was wondering about Barcın Höyük since it dates to c. 4000 BC aka the Anatolian Calcolithic. So does that individual with Gedmatch kit number M091434 show any real steppe admixture?

I also wonder what Y-DNA would persist that is more in line with Anthony's 2007 scenario?

Seinundzeit said...

David,

Thanks! Very much appreciate everything you do,

This is a treasure trove of data; can't wait to give these samples a spin...

supernord said...

@ Le skipper de Pytheas

"But Kroonen points out to evidence for Hittite speaking individuals involved in trade with Ebla (Northern Syria) around 2500 BC."

Kroonen is wrong, he accepts Hattic names giving them out for Hittite. There's no talk of Hittite speaking. In the source code not written any definitions for Hittite names, as if Hattians had not existed. The Hattians were just on the verge of Ebla, they were the vast majority of the population of the Hittite Kingdom.

How is it possible that in Ebla was not present Hattians, but it was present mythical Hittites? So let Kroonen first show Hattic names and prove that these names are not the Hattians (or from other aboriginal tribes). We absolutely know that Hittites borrowed words and names from a before Hittitic population.


The fact that these burials have a Hattians are proved by the fact that they were buried in the rooms, the Hittites were never buried in the rooms.

In addition, it seems that this city was conquered by the Hittites only at Labarna or Hattusilis.

MaxT said...

They suggest Anatolian Indo-European is a sister language instead of daughter language and they established themselves in Anatolia by gradual infiltration and cultural assimilation.

Linguistic supplement to Damgaard et al. 2018:

"First, the lack of genetic indications for an intrusion into Anatolia refutes the classical notion of a Yamnaya-derived mass invasion or conquest.However, it does fit the recently developed consensus among linguists and historians that the speakers of the Anatolian languages established themselves in Anatolia by gradual infiltration and cultural assimilation.

"Our findings corroborate the Indo-Anatolian Hypothesis, which claims that Anatolian Indo-European split off from Proto-Indo-European first and that Anatolian Indo-European represents a sister rather than a daughter language. Our findings call for the identification of the speakers of Proto-Indo-Anatolian as a population earlier that the Yamnaya and late Maykop cultures."

They are supporting steppe theory, they also have EHG labeled Indo-European in the study.

mzp1 said...

Anatolian is not a sister-language lol. Anatolian is IE therefore by definition it is a descendant of PIE. They just want to push everything out until they have a nice PIE=Steppe identification.

Dont just take everything you read in an amateur linguistic supplement as fact.

Davidski said...

@mzp1

Dont just take everything you read in an amateur linguistic supplement as fact.

Guus Kroonen 1,3, Gojko Barjamovic 2, and Michaël Peyrot 3.

AFFILIATIONS
1 Department of Nordic Studies and Linguistics, University of Copenhagen, Denmark.
2 Department of Near Eastern Languages and Civilizations, Harvard University, USA.
3 Leiden University Centre for Linguistics, Leiden University, The Netherlands.

mzp1 said...

@Davidski,

"The earliest available archaeological evidence of two-wheeled chariots is dated to approximately 2000 BCE (Anthony 1995; Anthony and Ringe 2015; Kuznetsov 2006: 638–645; Teufer 2012: 282). This offers the earliest possible date so far for the end of Proto-Indo-Iranian as a linguistic unity"

So "earliest possible archeological evidence" = "earliest possible"???

This is exactly what the guy is saying. So many things wrong with that. Looks pretty amateur to me.

MaxT said...

@mzp1

Genetically, whatever steppe came to Anatolia is OLDER* than bronze age steppe, but it is still steppe and came from the steppes, which is obvious from qpAdm here.

Linguistically, PIE as language and how it relates to other languages in its pre-PIE stage will probably change in coming years. Anatolia Indo-European shows linguistic influences from Uralic for a reason and obviously came from steppes.

Here is some clue on where we're headed

Kortlandt (2001: 1) “we may think of Indo-European as a branch of Indo-Uralic which was transformed under the influence of a Caucasian substratum.”

We already know genetic landscape of Siberia, Eastern Europe, Central Asia and how hunter-gatherers* here were connected in way or another since Mesolithic. Proto-Indo-Uralic as Kortlandt suggests may have originated among such hunter-gatherers, later PIE emerging from contact of EHG groups with Caucasus groups in steppes and then Anatolia arriving to modern Turkey with such groups.

supernord said...

mzp1 said...

" "The earliest available archaeological evidence of two-wheeled chariots ""

chariots! exactly chariots! but not two-wheeled carts! therefore two-wheeled cart was founded in earliest Catacomb culture!

"So many things wrong with that. Looks pretty amateur to me."

So you write only wrongs. You are pretty amateur layman. It is fact which cannot be rejected.

Chetan said...

@MaxT Exactly what many linguists have been saying for long. Anatolian came from an Indo-Hittite root which is considerable older than all the rest of the IE proto-languages (including Tocharian).

Also noteworthy what Kortlandt says about the Uralic affinities of certain Anatolian forms rather than IE proper.

mzp1 said...

MaxT,

Ahh the joys of theoretical linguistics

"Kortlandt (2001: 1) “we may think of Indo-European as a branch of Indo-Uralic which was transformed under the influence of a Caucasian substratum.”
"

But lets not forget the semitic languages because similarities have been shown.

Indo-Uralic-Semitic

With enough THEORIZING we can get probably get to

INDO-URALIC-SEMITIC-DRAVIDIAN-CHINESE and then we will have found the first human language ever spoken.

In the words of the above 'Linguists' this would something like
Pre-Proto-Pre-Uralic-Proto-Aryan. This consists of

Proto-Proto-Pre-Uralic and Proto-Proto-Proto-IE-Pre-Anatolian

The direction is North West North North North West South.

And as linguists we can do away with those pesky vowels using Laryngeal Theory. Pre-Proto-Proto IE words include classic tongue-twisters like *H3H2kwH3s for Horse.


Some of us though, prefer to exist in reality.

Palacista said...

If Hattics and their non IE language did not exist the claim for non EHG IE speakers in Anatolia would be strong, however, the Hattic language and its speakers did exit at the time of the samples in question. Surely no conclusions can be drawn on such data.

Kurd Dgk said...

For those interested:

I have spent the past couple of days analyzing modern Europeans, W Asians, and SC Asians using the aDNA sequenced in the from Damgaard et al and the Narasimhan et al 2018 papers (Botai and S Asian papers) using ADMIXTOOLS.

I've used f3 (source1 , source 2 ; test subject ) of the form:

1 - f3 [ Iran-Chl / HajiFiruz / AnatoliaMLBA / ArmeniaEBA , Saka/Scythian/Hun/MLBA/EMBA ; Test Subject ] for W Asian and a couple of S European modern subjects, and

2- f3 [ Beakers/Corded Ware , Saka/Scythian/Hun/MLBA/EMBA ; Test Subject ] for European subjects.


F3s are generally stingy in giving a negative result which is indicative of admixture between source1 and source2 to result in the test subject. The larger the -ve result, the stronger the admixture signal as per f3 instructions.


The reason I went through the effort of using Iran-Chl, Armenia-EBA, HajiFiruz-C, Anatolia-MLBA as source 1 for W Asians, SC Asians, and S Europeans is because we are very certain that those populations valid sources because they form substantial substructure for those moderns. This was also for the critic that may ask why I used Iran-Chl and not Kura-Araxes and so on. So basically to cover all bases.

I tested all those source 1 ancient pops with a host of source 2 MLBA, EMBA, and 65 Saka/Scythian/Hun samples I had at my disposal. Those samples span Hungary all the way east to Altai.

For Europeans f3 [ Beakers/Corded Ware , Saka/Scythian/Hun/MLBA/EMBA ; Test Subject ] would mostly test for additional steppe above and beyond that contributed by Corded Ware and BB, as well as an additional layer of Hungarian Scythian or Turkic from later expansions.


The results for W Iranics such as Kurds and Persians, and a Turk were unmistakable. Topping the chart for signal strength, consistent C Asian Sakas, and a few Tien Shan Saka and Hun. In fact one of the Feyli Kurds had an admixture signal for 57 out of the 65 Saka samples!


By contrast other W Asians such as Assyrians and Jordanians did not produce even a FEEBLE signal for ANY of the STEPPE groups, Saka, MLBA, or EMBA. This indicates that those other W Asians DON'T have any additional Steppe admixture outside of what was contributed to them by Iran-Chl / HajiFiruz / AnatoliaMLBA / ArmeniaEBA ( there may be very little that the stingy f3 is not picking up, but it would be negligible ). Some may ask so why do we see some of those W Asians score steppe using other methods of calculation. My gut feeling is that most of it is likely ancient alleles shared between Assyrians and Arab ancestors such as Iran-Chl / HajiFiruz / AnatoliaMLBA / ArmeniaEBA and Steppe groups.

Just before someone things that a little SSA in some W Asians is dampening the f3 signal. That does not appear to be the case, as a SW Iranian project member who has the highest SSA from my members, still had good admixture signals for many of the Saka samples.


Although I have not test S Asians yet, I did look at a couple of S Indian tribals. As expected f3s did not produce any signal of admixture for MLBA, EMBA, Saka. NW S Asian will most likely have a very different outcome.

Detailed tables and graphs for the above to be incorporated into my website at a later date. Also, Europeans did not produce any significant signals indicative of admixture from Saka/Scythians except for a N Italian who showed some indication of a little admixture from Hungarian Scythian related.

A little more on this at:

https://anthrogenica.com/showthread.php?14237-1st-Look-at-Damgaard-et-al-2018-C-Asian-Samples

Unknown said...

Genetic studies simply can't debunk linguistic results, it is not possible.

Why?

supernord said...

@Unknown

"Genetic studies simply can't debunk linguistic results, it is not possible.

Why?"

And linguistics confirms that those names are not of the Hittite.

"Elements such as ali-, alali-, lawadu-, memi-, mula/i- are attested
in Anatolian personal names of the Old Assyrian period (Laroche 1966: 26–27, 106, 118, 120)." Archi, A. 2011. In Search of Armi.
-da/-du /= -nda|-ndu

As you can see they were spread in Anatolia is certainly not the Hittites, known in no Hettites forms, obviously the period when the Hittites were not.

Matt said...

Davidski: This doesn't work for two reasons:

- Trzciniec people (Baltic_BA), who were descendants of Corded Ware, and apparently ancestors of Balto-Slavs, show a close paternal relationship to Indo-Iranians (and especially Indo-Aryans) via R1a-Z645, well before Scythians ever appeared on the horizon

- Balto-Slavic languages don't show a special relationship with Iranian relative to Indo-Aryan languages


Responding here rather than in the other thread. Fair enough both points, but

- Re: y-dna, it would be interesting to see what the R1 in these Scythians is. You don't think there's even a possibility it could be phylogenetically closer to modal R1a in Eastern Europe today than that which is present in Baltic_BA?

- Re: Low level special relationship to Iranic over Indo-Aryan, is there really enough resolution and clarity and stability among models to discount this?

Anyway, leaving aside the language and R1a subtype ideas levels, it does really seem to me like the Hungarian Scythians with Baltic_BA are pretty ideal for a two-way model involving present day populations from East-Central Europe.

When I reprocess G25 through PCA again with just samples relevant to Europe (sort of simulating a modern+ancient Europe specific PCA), they do look like they are a good match to fit with Baltic_BA in both the general West Eurasian dimensions, and the Balto-Slavic vs West European (with MN European and Steppe EMBA&MLBA) PC3 dimension: https://imgur.com/a/Ma9lE0S.

Looks maybe more consistently than the late Bronze Age Hungarians (who are often rather more EEF+WHG?) and the Hungarian Bell Beaker individuals. As would be expected from a later population.

I suppose we could argue that there are likely to be a number of populations in the region at this time that autosomally fit the bill, though, on the other hand, and the Hungarian Scythians just happen to have been sampled.

So I do still think it would at least be interesting to see them dropped on the Northern European PCA, even if we settle on the language/culture argument for a connection between Baltic-Slavic and Indo-Iranian via Hungarian Scythians being pretty wrong and not at all likely.

Rob said...

@ Jaydeep

“Second, contacts between Bronze Age steppe populations and NMG V and BMAC populations appears to have been one in which the dynamic of cultural influence was stronger on the side of the well-established sedentary foodproducing populations, and this resulted in the partial assimilation of these initial newcomers to the region both culturally and, to a lesser degree, biologically as well. Third, not all of those who emigrated from the north turned to farming but may have continued a semi-nomadic existence in the highlands, which were unsuitable for the kind of intensive farming practiced in the BMAC homelands or in the regions of Khorezm. Fourth, if there was any Central Asian influence on
South Asian populations, that influence likely long predated any development of Iranian, let alone Indo-Aryan, languages, and most likely occurred during the late NMG IV to early NMG V period (ca. 2800–2300 BCE) and even earlier during the Eneolithic from Kelteminar culture groups (4000–3500 BCE).”

Thanks
Yes I think i agree with the Danes. Demographically and socially marginal impact of steppe migrants being late integrants to the “world system” would not have made linguistic change.
So MBA steppe admixture indeed, but deep down all credible academics are abandoning the steppe theory.

Arch Hades said...

If Mycenaean Greek commoners had 15-20% steppe ancestry why not Hittite commoners?

Rob said...

@ Matt
The R1 in these scythians is a G !

John Johnson said...

@ Arch Hades

Possibly because the separation of the IE-Anatolian group was deeper in time, pre-Yamnaya. Anthony 2007 talks about this:

'Migrations into the Danube Valley: The Suvorovo Novodanilikovka Complex'

Essentially, Anatolian is the first IE group to split but it happens long before Yamnaya during Sredny-Stog.

Matt said...

Comparing distance of Hungarian_Scythian vs previous sampled Hungarian ancient and present day "populations", in G25: https://imgur.com/a/9Eu5LHS

Compared to all previous averages, Hungarian Scythian is offset in being closer to recent Eastern European people, compared to Beaker Hungary, otherwise the most similar set.

Beaker Hungary is slight offset to more relatedness to Iberia and NW Europe; Beaker Iberia and Iberia BA are the ancient samples that peak the Beaker Hungary side of the residual and Slavic Bohemia is the ancient that peaks the Hungarian Scythian side. Baltic BA is the ancient which peaks relatedness to modern Hungarians relative to Hungarian Scythian.

@Rob, have these samples been re-analysed online then? Supplement matches only G to DA117 (Nomad_His, Tian Shan). Hungarian_Scythian provides R1 for DA197 (though I wrongly thought it provided R1 for DA198 also; this sample is actually not assigned). I used a lookup between their supplement spreadsheets to match the sample labels with contexts to labels with y haplos.

supernord said...

Arch Hades said...
" If Mycenaean Greek commoners had 15-20% steppe ancestry why not Hittite commoners?"

Therefore they are not Hittites, but Hattians.

@John Johnson

Souvorovo is part Sredniy Stog II culture, therefore, it have full EHG and probability steppe component. Pre-Yamnaya state doesn't explain anything.

old europe said...

supernord


"Therefore they are not Hittites, but Hattians".

then why in the paper they talk about anatolian IE languages.....

John Johnson said...

@ supernord

Pre-Yamna means before Yamna in this case. Sredny Stog Phase II has a wide date from 4000-3500 BC and earliest for Yamna is 3300-2600 BC. To further clear things up its the same EHG population that creates Sredny Stog but later Yamna although Yamna has all that CHG likely from Caucus women.

supernord said...

old europe said...

" then why in the paper they talk about anatolian IE languages....."

The fact of the matter is that they're in the paper don't say a word about the Hattians, as if they do not exist in nature. Specially. What is unacceptable is simply ignoring historical, linguistic, archaeological and cultural data.
In this case, the reference to IE languages in this case is a pure tendentious fiction.

John Johnson said...
" Pre-Yamna means before Yamna in this case. Sredny Stog Phase II has a wide date from 4000-3500 BC and earliest for Yamna is 3300-2600 BC. To further clear things up its the same EHG population that creates Sredny Stog but later Yamna although Yamna has all that CHG likely from Caucus women."

Sredny Stog II has a dates of 4600-4200 BC, but not 4000-3500 BC! I perfectly understand that means Pre-Yamnaya. Between these cultures is not a thousand years old. Steppe component was long before Yamnaya culture, he was in Dereivka culture 4000 BC, CHG component was in Khvalynsk culture 5 Millennium BC.

Sanuj said...

Of all the examples cited from Kuzmina, all the borrowings are from IIr to FU, not the other way round. It is pretty clear what is the direction of loans, but still she makes an exact opposite conclusion? Where are the FU loans in Rigvedic Sanskrit after such a presumed long contact?

And the "chariots", the less said the better.

But, I do think that weak arguments are used on purpose.

Le skipper de Pytheas said...

Re Anatolian IE ("Hittite") vs non-IE Hattic
Consider just this example from the Ebla archive
Tar5-hi-li
Closer to Hittite Tarhunt- (Name of Storm God) than to Hattic Taru- (same)
-ili/u is a common 'suffix' in Hittite PN
cf the king's name in the Old Kingdom
Hattusili
Mursili
Hantili


The evidence from Hattic is so limited that to claim that the PN quoted by Kroonen are Hattic is unsupported...

Chad Rohlfsen said...

Those aren't really pops from the same timeframe. We're talking about 2,000 years of difference. If you use something more close to the timeframe, you see that steppe is really no more than the standard error. Steppe ancestry isn't needed. It is more likely that 2203 is from more northern Anatolia or a region with less Levantine flow. This person probably just has less Levant N ancestry than the others. Nothing to do with steppe admixture. If the Levantine ancestry did not reach the northern part of Anatolia, then the pop there is likely to still be on the Anatolia_ChL and Armenia_ChL/EBA part of the plot. Comparing 2203 to other MLBA and EBA samples there isn't really anything going on there, but there is something else going on.

The big thing, is that these aren't Harvard samples, so you're not going to get much in the way of accuracy. Just looking at this, there are mapping issues and we're going to have to wait for these to go through Cambridge. These stats showing how the Harvard Anatolia_BA is closer to everything shows that these samples are not aligned correctly and every statistic involving Harvard samples is really null and void.

Chad Rohlfsen said...

result: Mbuti_DG Levant_N Anatolia_BA Anatolia_EBA -0.003024 -4.642 10169 10613 146566
result: Mbuti_DG Levant_N Anatolia_BA Anatolia_MLBA -0.002897 -4.749 10279 10709 148273
result: Mbuti_DG Levant_N Anatolia_BA Anatolia_MLBA_2203 -0.003267 -3.780 6663 6983 98032
result: Mbuti_DG Levant_N Anatolia_EBA Anatolia_MLBA -0.000078 -0.137 11740 11753 168076
result: Mbuti_DG Levant_N Anatolia_EBA Anatolia_MLBA_2203 -0.000539 -0.665 7654 7713 111142
result: Mbuti_DG Levant_N Anatolia_MLBA Anatolia_MLBA_2203 -0.000827 -1.038 7684 7777 112563
result: Mbuti_DG Anatolia_ChL Anatolia_BA Anatolia_EBA -0.003434 -4.663 8442 8856 120381
result: Mbuti_DG Anatolia_ChL Anatolia_BA Anatolia_MLBA -0.003175 -4.022 8571 8957 121704
result: Mbuti_DG Anatolia_ChL Anatolia_BA Anatolia_MLBA_2203 -0.002961 -2.804 5550 5789 80704
result: Mbuti_DG Anatolia_ChL Anatolia_EBA Anatolia_MLBA 0.000504 0.703 9272 9207 130075
result: Mbuti_DG Anatolia_ChL Anatolia_EBA Anatolia_MLBA_2203 -0.000769 -0.764 5960 6027 86272
result: Mbuti_DG Anatolia_ChL Anatolia_MLBA Anatolia_MLBA_2203 -0.000717 -0.713 6033 6096 87315
result: Mbuti_DG Armenia_ChL Anatolia_BA Anatolia_EBA -0.002138 -3.915 11539 11889 163652
result: Mbuti_DG Armenia_ChL Anatolia_BA Anatolia_MLBA -0.001435 -2.834 11741 11978 165661
result: Mbuti_DG Armenia_ChL Anatolia_BA Anatolia_MLBA_2203 -0.002718 -3.555 7565 7863 109462
result: Mbuti_DG Armenia_ChL Anatolia_EBA Anatolia_MLBA 0.000755 1.671 14443 14290 203223
result: Mbuti_DG Armenia_ChL Anatolia_EBA Anatolia_MLBA_2203 -0.000530 -0.788 9345 9417 134104
result: Mbuti_DG Armenia_ChL Anatolia_MLBA Anatolia_MLBA_2203 -0.001372 -2.012 9398 9584 135977
result: Mbuti_DG Seh_Gabi_ChL Anatolia_BA Anatolia_EBA -0.002754 -5.138 11316 11764 162746
result: Mbuti_DG Seh_Gabi_ChL Anatolia_BA Anatolia_MLBA -0.002251 -4.290 11511 11882 164742
result: Mbuti_DG Seh_Gabi_ChL Anatolia_BA Anatolia_MLBA_2203 -0.002716 -3.693 7453 7748 108850
result: Mbuti_DG Seh_Gabi_ChL Anatolia_EBA Anatolia_MLBA 0.000558 1.180 14232 14120 202109
result: Mbuti_DG Seh_Gabi_ChL Anatolia_EBA Anatolia_MLBA_2203 -0.000880 -1.271 9196 9313 133404
result: Mbuti_DG Seh_Gabi_ChL Anatolia_MLBA Anatolia_MLBA_2203 -0.001488 -2.093 9281 9483 135282
result: Mbuti_DG Villabruna Anatolia_BA Anatolia_EBA -0.001613 -2.292 10500 10741 148932
result: Mbuti_DG Villabruna Anatolia_BA Anatolia_MLBA -0.001545 -2.094 10613 10846 150728
result: Mbuti_DG Villabruna Anatolia_BA Anatolia_MLBA_2203 -0.001128 -1.135 6950 7063 99733
result: Mbuti_DG Villabruna Anatolia_EBA Anatolia_MLBA -0.000124 -0.183 12101 12123 171673
result: Mbuti_DG Villabruna Anatolia_EBA Anatolia_MLBA_2203 0.000789 0.857 7964 7874 113644
result: Mbuti_DG Villabruna Anatolia_MLBA Anatolia_MLBA_2203 0.000819 0.844 8012 7918 115142
result: Mbuti_DG EHG Anatolia_BA Anatolia_EBA -0.001927 -3.606 11745 12068 167364
result: Mbuti_DG EHG Anatolia_BA Anatolia_MLBA -0.001261 -2.308 11935 12148 169458
result: Mbuti_DG EHG Anatolia_BA Anatolia_MLBA_2203 -0.000801 -1.067 7791 7881 111929
result: Mbuti_DG EHG Anatolia_EBA Anatolia_MLBA 0.000920 1.847 15775 15569 223370
result: Mbuti_DG EHG Anatolia_EBA Anatolia_MLBA_2203 0.000947 1.428 10306 10166 147060
result: Mbuti_DG EHG Anatolia_MLBA Anatolia_MLBA_2203 -0.000047 -0.069 10322 10329 149314

supernord said...

@ Le skipper de Pytheas said...

"Tar5-hi-li
Closer to Hittite Tarhunt- (Name of Storm God) than to Hattic Taru- (same)"

The root is the Hattian, so the original doesn't even suggest it's a Hittite name.

"-ili/u is a common 'suffix' in Hittite PN"

You think it is, because just taken out a full list of the names of Armi:
Šar-mi-lu, Ša-li-lu, Ar-gi-lu, Ar-i-lu, Mu-rí-lu, Zi-la-li, etc.
Even there is no suspicion on the attitudes of Hittites. The combination of sounds is the most common, occurring in all local languages. It is not especially Hittitic.

mzp1 said...

A basic phonological table..

https://ibb.co/ifUnfd


In the first table only Vedic and Greek keep PIE aspirates.

In the second table both Old Church Slavonic and Lithuanian agree with Avestan vs Vedic.

In the last table the Euro languages are closer to Iranian, with Gothic and and Lithuanian particular close to Old Persian vs Avestan, and far from Vedic.

Obviously there are no cases where Balto-Slavic-Germanic are closer to Indo Aryan than Iranian.

Unfortunately there isnt much information on Old Persian there. I would expect it make the pattern clearer.

supernord said...

mzp1 said...
" A basic phonological table..
In the first table only Vedic and Greek keep PIE aspirates. ..."

Designation of the last row as graduate students is a classic Convention of the 19th century (Sanskrit-basic reconstruction 19th century of PIE). This reconstruction has long been rejected, but it is so designated from tradition. See, you're trying to fiction something without understanding the basics, you just don't even have the basic knowledge.

John Johnson said...

@ supernord

just out of curiosity which publication are you going by for those dates of Sredny Stog?

andrew said...

The ancient DNA results are consistent with the fact that contemporaneous historical accounts from Akkadian traders suggest that this conquest started shortly before 1740 BCE.

mzp1 said...

@Supernord,

"Designation of the last row as graduate students is a classic Convention of the 19th century (Sanskrit-basic reconstruction 19th century of PIE). This reconstruction has long been rejected, but it is so designated from tradition. See, you're trying to fiction something without understanding the basics, you just don't even have the basic knowledge."

No, I got the whole thing from Beekes', it is a current text.

I am not even talking about reconstructions so not sure what you're looking at or seeing in your head.

RAGERAGE said...

@ Le skipper de Pytheas said...

"Tar5-hi-li
Closer to Hittite Tarhunt- (Name of Storm God) than to Hattic Taru- (same)"

The root is the Hattian, so the original doesn't even suggest it's a Hittite name.

"-ili/u is a common 'suffix' in Hittite PN"

You think it is, because just taken out a full list of the names of Armi:
Šar-mi-lu, Ša-li-lu, Ar-gi-lu, Ar-i-lu, Mu-rí-lu, Zi-la-li, etc.
Even there is no suspicion on the attitudes of Hittites. The combination of sounds is the most common, occurring in all local languages. It is not especially Hittitic."

Those names are not Hattian but Hurrian. Actually most of Hittite Labarnas bore a Hurrian name as royal name. Hittite kings took new royal names when became king. -li suffix means of or from, so Hattusili means of/from Hattus. Also Urartians, the descendants of Hurrians, called themselves and their core land as Bianili which means of/from Bian. Interestingly same suffix exists in Turkic languages.

Anthro Survey said...

@Seinundzeit

Good stuff(as usual)! A couple of things I want to remark on.

Re/Swat and Kalash:
That Botai-like signal is quite significant and consistent with both ADMIXTURE output from the paper(green is enriched for this ancestry) and previous models. Given the ratio and steppe_MLBA_east not being quite as enriched for it, I think it represents Central Asian Hunter Gatherer(CAHG) survivor elements in this context. Highland environments are conducive for the survival of older population strata. In the others, otoh, probably just steppeMLBAeast.

In fact, with Botai, GonurBA1_outlier and other samples available, I wonder how feasible it would be to derive a CAHG ghost? I'll have a stab when I get around to it, but maybe yourself, Matt or Arza can give it a shot in the meantime?

Interesting runs using Kangju to model Hepthalite influence. Have you tried any other western Hunnic groups? Predictably, these eastern Iranics show clear affinity to them---and to exclusion of Turkic input. Clause wouldn't hold for proper modern Tajiks, tho, I'm guessing.

"...literally". I agree. Don't know how neatly Kangju %s correspond to actual %s given shared ancestry b/ween them and Andronovo/Sintashta. Sometimes, monte can "shunt" ancestry to one of the two inputs if there's a lot of shared drift.

Re/BMAC in upper caste----HF is probably connected to this. When I use Alberto's ghost, they often do get Gonur about 15% Gonur.

Anthro Survey said...

@Davidski

Can you substitute Varna_outlier in for Ukr_Eneo? Or this would be too similar to the other leftpops?

If that works, maybe you can also try this?:
-HF_chl
-Tepecik or Levant_N or Levant-Barcin hybrid synthetic
-Varna_outlier+Varna (50/50 synthetic)

Davidski said...

Tried it. Doesn't work.

Anthro Survey said...

@Old Europe

So far, it's looking like Hittites could have been enriched for EEF(with minor EHG from interactions w/what was to become repun area?), not just ANF. We should be careful to distinguish these two given the context of an OOEE scenario. Besides, and hopefuly, if we get data from early Sesklo, this can give us a better idea of things. I.E. it's reasonable to see Barcin ANFs as a *contemporaries* to proto-EEFs, not so much ancestral in and of Itself.
In fact, there are lines of evidence suggesting that Seskloans developed some important techniques not present in contemporary Anatolia. Independent developments....that kinda thing.

Anthro Survey said...

@Davidski

Hmm. Too similar, I guess. Is there another way to autosomally explore the EEF vs ANF angle in bronze age anatolia? What techniques would be sensitive and reliable enough?

Samuel Andrews said...

@Davidski,

Can you add Tagar to G25 spreadsheet?

Rob said...

@ Matt

According to OG, DA198 (“Hungarian Scythian”) is G2
I’ve not double checked , and I appreciate it’s probably an outlier by virtue of geography

postneo said...

A viable model for the spread of IE. is from either east Anatolia/maykop/south Caspian to the steppe and independently to the balkans.
The steppe population adopts it for ease of transacting with southern neighbours. and the spread has nothing much to do with genetics.

On a different note:
If you strip away the pseudo academic patina.. the so called Andronovo vedic association reminds one of the shroud of Turin. A supposed relic of a man born in Bethlehem is enshrined in far away Italy. It becomes an article of "faith"

Aram said...

Davidski

Concerning those possible Eneolithic infiltrations via Caucasus. What You think about Armenia Chl?
What causes it's northern shift? Ukraine Eneolithic, EHG or something else. Also do it have ancestry from Central Asia? FrankN was suggesting a migration from Namazga to S Caucasus.
You can use Hajji Firuz Chl as a baseline because it is archaelogically related to Armenia Chl.

epoch2013 said...

@Rob

A while ago we quibbled on the onset of pastoralism on the Pontic steppe on the basis of a paper. I lost the link to that paper. I think it was in German. Do you still have that link?

epoch2013 said...

@Chad

Mbuti_DG EHG Anatolia_BA Anatolia_EBA -0.001927 -3.606 11745 12068 167364

Now that is a surprise. Anatolia_BA are the Lazaridis samples, I presume?

Davidski said...

@epoch

Singleton D-stats aren't always as informative as it's made out to be, especially when mixing capture and shotgun generated data.

@Aram

I haven't looked at Armenia_ChL yet in the context of all of the new data, but I'd be very surprised if anything from Central Asia can explain its ancestry.

I'll have a look this week sometime.

Le skipper de Pytheas said...

@ Raregare

Those names are not Hattian but Hurrian
Not true for my example: the Hurrian name for the Storm God was Teshub


Actually most of Hittite Labarnas bore a Hurrian name as royal name
During the NEW kindgdpm, some had a Hittite and a Hurrian name. Not during the OLD kingdom...

Hittite kings took new royal names when became king. -li suffix means of or from, so Hattusili means of/from Hattus
Not true: the king's name Hantili is transparently Hittite
hant-: front (face, building) hence Hantili = the front, first one (the leader)
so -ili does turn a (common) name into a P(ersonnal) N(ame)

old europe said...


Anthro

Ok.... Can you clarify one thing . I do not remember where but I've heard that ANF were already mixed with western populations and that they score around 40% WHG ( villabruna maybe). Is that true?

supernord said...

@mzp1

If you see something then you still don't understand it.


@Le skipper de Pytheas
"so -ili does turn a (common) name into a P(ersonnal) N(ame)"

-ili- is a suffix accessories. Unknown origin.

"Hattusili means of/from Hattus"

In Hittitic, it is not Hattusili, but Hattusilis.

Davidski said...

@Anthro Survey

Is there another way to autosomally explore the EEF vs ANF angle in bronze age anatolia? What techniques would be sensitive and reliable enough?

Not sure? Maybe qpGraph trees like these?

Anatolia_MLBA_MA2203 (Barcin_N)

Anatolia_MLBA_MA2203 (Balkans_ChL)

Anatolia_MLBA (Barcin_N)

Anatolia_MLBA (Balkans_ChL)

Rob said...

@ epoch

Yep of course I remember.
-> , Der Übergang zur Rinderzucht im nördlichen Schwarzmeerraum by Elke Kaiser

Rob said...

^ A very important contribution. Along with other publications, it deconstructs the idea of 'elite horse-riding nomads' from Khvalynsk. She demonstrates that as late as 3500 BC and beyond, the economy of all but the western-most steppe (Usatavo culture) was mostly based on hunting wild game, before a possibly rapid transition to specialised cattle-herding. These Sub-Neolithic populations must have been replaced like elsewhere.
She further clarifies there is no evidence for mounted invasions from the steppe even during the Catacomb period, into the Balkans. Obviously not Anatolia, either.
Another one for the real scientists.

old europe said...


ROB

to read synoptically with this.....


Egalitäre Hirtengesellschaft versus Nomadenkrieger? Rekonstruktion einer Sozialstruktur der Jamnaja- und Katakombengrabkulturen (3. Jt. v. Chr.).

Rob said...

@ Old Europe
Yep that's the second one
Of course, that's not to deny that people from the steppe didn't move south. Of course they did. But as I keep repeating it's context that matters. I'm not simply mocking when I say they brought cattle and marriage partners

In that same article, she looks at the origin of Chariots. It is difficult to link charitos to the Wagons of the Yamnaya & Srubnaya. It's as if, they suddenly appear, then move westward.

Also to be read is Ivanova's work on the metallurgical links between Iran, the Caucasus, Aegean , BB, etc.
I'll leave Anatolia for now, it's much too complicated, although clear enough.

Davidski said...

@All

Judging by this, it looks like Y-hg N1c and Uralic-speakers started trickling into the southern Baltic region during the Iron Age.

https://drive.google.com/file/d/1WXr0G2Ksg6ASi-J3YqJ0zpVnVWbpffU5/view?usp=sharing

https://drive.google.com/file/d/1eMHm8ewrhXXhVq101Q1-lz_y-RFXNrzT/view?usp=sharing

@Matt

Here are the coords for the Hungarian Scythians that made it into this plot.

Scythian_Hungary:DA191 -0.0061 -0.0103 -0.0086 -0.0015 -0.0087 -0.0058 0.0108 0.0033 -0.0003 0.008
Scythian_Hungary:DA194 0.0173 -0.0154 -0.0057 -0.0068 -0.0047 0.0038 -0.0061 -0.023 0.0045 -0.0029
Scythian_Hungary:DA195 0.019 -0.0104 -0.0061 0.0077 0.0021 -0.0009 0.0003 -0.002 0.0033 -0.0086
Scythian_Hungary:DA197 0.0008 -0.0082 0.0027 0.0004 -0.0003 0.01 0.003 0.0159 -0.0048 -0.0099

epoch2013 said...

@David

"Singleton D-stats aren't always as informative as it's made out to be, especially when mixing capture and shotgun generated data."

OK. Nothing there.

On another note, did you ever try if a similar small signal such as with MA2203 can be eked out of the Lazaridis 2017 Anatolian Bronze Age samples?

Davidski said...

Not really. I don't worry too much about D-stats like that anymore. They're often not very informative.

Rafs said...

David, is there a place we can find the meaning of some of the population labels? I'm very interested in the Indian ones.

Davidski said...

@Rafs

You mean from the new Damagaard et al. paper? I don't know? Maybe in the supp info?

http://science.sciencemag.org/content/suppl/2018/05/08/science.aar7711.DC1

MaxT said...

Balkan route looks unlikely based on those qpAdm now.

Caucasus route makes more sense, and fits with Indo-Hittie model for Anatolia.

epoch2013 said...

@Rob

Yes, that's the one. Tanks a lot.

epoch2013 said...

@MaxT

Why do you consider a Balkan route unlikely?

Rob said...

Some interesting models for a couple of W & S Slavs.

Serbian
'Thracian' 35%
'Baltic_BA' 31%
'Mycenean' 16%
'Nordic_IA' 9%
'Alan-Cauc.' 5%
'Gepid_Serb' 2%

Polish
'Baltic_BA' 52%
'Thracian' 18%
'Nordic IA' 14%
'Suebic' 10%
'Alan' 2%

Bulgarian
'Thracian' 34%
'Baltic_BA' 27%
'Mycenean 17%'
'Alan 13%'
'Halstatt' 5%

Slovenian
'Baltic_BA' 34%
'Nordic IA" 25%
'Thracian' 24%
'Mycenean 13%'

All fits d < 0.12%
They will change with time.

'



MaxT said...

@epoch2013

I had earlier posted about this, lingustic papers that came with the study and that of Kortlandts.

Huck Finn said...

@D and re: "Judging by this, it looks like Y-hg N1c and Uralic-speakers started trickling into the southern Baltic region during the Iron Age."

Very interesting positioning, many thanks for the analysis. TMRCA of "Baltic" N-M2783 is 2700 years, but the root is so far Finnish i.e. the only N-M2783* so far has been found in Finland. However, it may very well be that the connection is linked to a third place, such as Estonia or even some more remote place?

The "Lithuanian" with paternal N anyway seems to be a mixture of Baltic Finn or something pretty similar and a Balt.

RAGERAGE said...

@Le skipper de Pytheas

"Those names are not Hattian but Hurrian
Not true for my example: the Hurrian name for the Storm God was Teshub"

Using a theonym in a personal name does not change that personal name is ending with a Hurrian suffix. For example Christian is an English name but Christ derived from Greek via Latin. Also Northern Syria was one of the lands that Hurrians lived.

"Actually most of Hittite Labarnas bore a Hurrian name as royal name
During the NEW kingdom, some had a Hittite and a Hurrian name. Not during the OLD kingdom..."

So, according to you, Hattusili I who moved Hittite capital to Hattusa, was not the second ruler of Old Hittite Kingdom.

"Hittite kings took new royal names when became king. -li suffix means of or from, so Hattusili means of/from Hattus
Not true: the king's name Hantili is transparently Hittite
hant-: front (face, building) hence Hantili = the front, first one (the leader)
so -ili does turn a (common) name into a P(ersonnal) N(ame)"

Hantili was an usurper and his name is Luwian, Hittitic form is Hantezzi. - li suffix also shows relevance. Also please explain for example why Urhi-Teshub became king with regal name Mursili.

RAGERAGE said...

@Supernord
""Hattusili means of/from Hattus"

In Hittitic, it is not Hattusili, but Hattusilis."

No, it is Hattusili. Hattusilis is nominative (singular animate) form.

zardos said...

@All: Is there a site which lists all ancient yDNA and mDNA predictions including the newest ones?
Whats the ydna of Thracian samples?

Shaikorth said...

@Huck Finn

Looks like the NLithuania sample is very close to modern Latvians in Global25, a bit more EHG.

"distance%=4.1946 / distance=0.041946"


NLithuania:DA171

Latvian 96.2
Latvia_MN 3.8
West_Siberia_N 0.0
Nganassan 0.0
Nordic_IA 0.0

supernord said...

@ RAGERAGE said...
"No, it is Hattusili. Hattusilis is nominative (singular animate) form."

No, it is Hattusilis. Because Hittitic is nominative language with nominative ending -s.

"Those names are not Hattian but Hurrian
Also Northern Syria was one of the lands that Hurrians lived."

Hittitic Kingdom was not in Northern Syria. Hurrians will influence on Hittitic only in New Kingdom time. In Assyrian time of Ebla Hurrians were not in Northern Syria. Hittitic names are Hattian but not Hurrian. Some Hurrian names of guests possible are in Armi with Hattic names.

TruthPrevails said...

1
The Indo-Aryan and Iranian languages share a common set of etymologically related terms related to equestrianism and chariotry (Malandra 1991). Since it can be shown that this terminology was inherited from their Proto-Indo-Iranian ancestor, rather than independently borrowed from a third language, the split of this ancestor into Indo-Aryan and Iranian languages must postdate these technological innovations. "
2
The traces of early Indo-Aryan speakers in Northern Syria positions the oldest Indo-Iranian speakers somewhere between Western Asia and the Greater Punjab, where the earliest Vedic text is thought to have been composed during the Late Bronze Age (cf. Witzel 1999: 3).

Based on what you posted from the paper, what they seem to be saying is the Indo-Aryan Iranian split happened between 2000BC to 1700BC.

This is such a weak argument, which I guess stems from the fact that they do not have any Sanskrit & Avestan experts on their interdisciplinary team.

Avesta and ancient Persian inscriptions know about Haptahandu (Saptasindhu of Rigveda), HaraKvaiti (Sarasvati), horayu (Sarayu), haoma (soma), airyans (Aryans). While Rig does not know of any geography outside the limits of Greater India.

So if the texts were composed in Late Bronze Age (1200BC-500BC) and the split is supposed to have happened before 1700BC (courtesy Mittanis) how does Avestan Gathas know of rivers in India's interior ? Does not add up.

Vara said...

@TruthPrevails

Because those were not in the Gathas. The Vendidad is dated to the Parthian-Sassanid era as most Younger Avestan texts are. Also, the Horayu and Harahvaiti are in Afghanistan.

Not that I agree with the split dates or anything.

supernord said...

@TruthPrevails

"Avesta and ancient Persian inscriptions know about Haptahandu (Saptasindhu of Rigveda), HaraKvaiti (Sarasvati), horayu (Sarayu), haoma (soma), airyans (Aryans). While Rig does not know of any geography outside the limits of Greater India."


All these areas/entities were originally located outside South Asia. Initially the RigVeda describes them, not Indian.

"Based on what you posted from the paper, what they seem to be saying is the Indo-Aryan Iranian split happened between 2000BC to 1700BC."

Glottohronology calculation also.

mzp1 said...

@Supernord

"All these areas/entities were originally located outside South Asia. Initially the RigVeda describes them, not Indian."

You've lost the plot.


Balto-Slavic-Germanic are most likely derived from some version(s) of Iranian.

Greek-Armenian-Hittite develop from Indo Aryan around West Asia/South Caucuses.

Latin and Celtic are closer to Greek-Armenian-Hittite but with later Northern influences and greater earlier divergence.

This explains most of the structure within IE linguistics, literature and archeology.

There is no other competitive hypothesis.

There is little left to discuss other than figuring out the details.

Dmytro said...

I'm beginning to think that unless inordinate luck prevails, no irrefutable aDNA will emerge conclusively proving the historical roadmap of the classical Hittites. So no genetics and just as likely no archaeology. At the moment all that's left is linguistics (which is very suggestive but not conclusive). The extant Hittite texts indicate a fairly long period of interplay with the locals (and just possibly some other contacts). The antiquity of their IE is acknowledged. Perhaps they were a group which wandered very far (from some northern point) and "nomadized" their way into Anatolia (either from the West or the East. Neither route seems totally certain). Eventually, after a very long time period, they start to become visible, and then their historical moment arrives... I continue to feel that the archaic-ness of their IE and the lack of involvement in the "horse word" by Hittites ("Ansekura") and Luwites (a borrowing from the later Indo-Iranic "aswa"?) points to a very early separation from the rest of the IE world, more evident that that of the Tochharians even). For now, but totally open to emerging alternatives...

mzp1 said...

It doesn't work because 'archaic' elements in Hitrire are only so because they are closer to Greco-Aryan. Otherwise they would be considered as innovations.

You cannot have Hittite moving out first very early but linguistically closer to Indo Aryan that branched out later.

The linguistic tree model aligned with the steppe origin theory is a massive failure hence there is no clear proposal of how Greek and Indo Iranian moved away from the Steppe to their current locations.

You cannot have IA sharing items with Greek and another group with Iranian and Balto-Slavic-Germanic.

Mr. Kulkarni said...

So now saptasindhu is in Ukraine or something? Lol

TruthPrevails said...

@vara


The Sapta-Sindhava region was bounded by Saraswati in the east, and by the Sindhu in the west and the five in between were Satudru, Vipasa, Asikni, Parusni and Vitasta. This totally locks up the geography of Rig composition.

There are many corrupt translations citing reflex and what not connecting them to Volga etc for Saraswati and Horayu with Sarayu.

So if someone wishes to connect the two names with rivers outside India, then they need to also make up the names for the remaining six rivers because of them being mentioned sequentially. And show how the river names outside also map with the remaining six.

In present times we are trying to find just one river and make sense of Saraswati with Ghagghar Hakra.
But for naysayers they have to find the remaining six outside India which is impossible.

TruthPrevails said...

If just based on Chariot and horse-keeping etymology they want to lock up 2000BC date then what about wine making? Based on it Indo-Iranian language groups should be the oldest.

An important word borrowed by IE languages from West Asian languages is the
word for “wine”: viticulture and wine‐making originated in West Asia (c.f.
ancient Egyptian wnš, “grape, wine”): “The wide distribution of phonetically
similar words meaning ‘wine’, ‘grape’ among various linguistic groups of the
Near East…(Indo‐European, Semitic, Kartvelian) indicatesthe extreme
antiquity of the migratory term, which must have passed from one language
to another at a protolanguage level.” (Gamkrelidze).

The distribution of variants of the word in the different IE branches gives a
break‐up exactly in line with the three groups of IE dialects, and their positions
vis‐à‐vis the wine lands of West Asia: this word is totally missing only in the IE
branches located to the east of these lands.

supernord said...

@mzp1

"There is no other competitive hypothesis."

Stop freaking out. you're ridiculous. your fantasies are delusional.

@Dmytro

"Perhaps they were a group which wandered very far (from some northern point) and "nomadized" their way into Anatolia (either from the West or the East."

Suvorovo of Sredniy Stog, Cernavoda, Usatovo, Ezero, Troy cultures is long time way out of Anatolia with many non-Indo-Europen languages.


" continue to feel that the archaic-ness of their IE and the lack of involvement in the "horse word" by Hittites ("Ansekura")"

ANSE.KUR.RA is Sumerian word, but not Hittitic.

"and Luwites (a borrowing from the later Indo-Iranic "aswa"?)"

You are wrong. In a completely rebuilt by non-Anatolian-Balkan traditions Troy VI (before Mitanni) bred horses, although before that horses are not seen at all.
In Mycenaean burials there is a chariot-burial in direct poses.


supernord said...

@TruthPrevails

Seven rivers has long been called the region of Central Asia South of lake Balkhash. Saraswati>Haraxvati>Araxs was known in different times many of the rivers Araks Caucasian, Syrdarya, Amudarya, and apparently the Volga (`Ρᾶ < Raŋhā < Rasa, Rava < hrava < Srava). Herodotus constantly confuses these Arakses, because he has a Persian King at war with different peoples passing Araks, and, fighting with North Middle Asian.

In Rigveda, the exact enumeration of rivers is local, and it has changed over time.

"In present times we are trying to find just one river and make sense of Saraswati with Ghagghar Hakra."

Refuted and irrevocably.

Sanuj said...

On Saraswati:
http://www.omilosmeleton.gr/pdf/en/indology/RPSSC.pdf
"...consider also that the Vedic -s- is inherited from PIE, according to all IEnists,
whereas Avestan –h- is a devolved, not PIE, sound. Vedic √sṛ has many primary and
secondary cognates like sara, sarit etc. Now Avestan has no cognates for √sṛ and its
products, and the Avestan noun for lake is vairi- while vār- is ‘rain(-water)’=S vāri (?). The
stem hara- (cognate with Vedic saras) occurs only in the river name Haraxvaiti.
Consequently, it is the Iranians that moved away from the Indoaryans as, indeed, is shown
by their memory of having lived in a location they called Haptahәndu = Saptasindhu. The
root sṛ has cognates in other IE branches, Gk hial-, Latin sal-, Toch sal- etc as is shown by
Rix H., 1998. Now, it would be ludicrous to claim that the IAs left the common Indo-Iranian
habitat, as per the AIt, moved into Saptasindhu and turning the Haraxvaiti name into
Sarasvatī gave it to a river there to remember their past while they proceeded to generate the
root sṛ and its derivatives to accord with other IE languages. Occam’s razor, which here is
conveniently ignored by AIT adherents, commands the opposite: that the Iranians moved
away, lost the root sṛ and derivatives but retained the memory of the Sarasvatī river in its
devolved form Haraxvaiti and gave it to a river in their new habitat. This, together with the
fact that, like Greek, Avestan has no obvious system of roots and derivatives (as Sanskrit
has), should be enough to question if not refute various IEnists’ claims that Avestan retains
older forms of nouns and verbs and that therefore the Indoaryans were with the Iranians in
Iran in the common Indo-Iranian period – before moving to Saptasindhu."

Saraswati has always been the river between Yamuna and Sutlej, as attested by the Nadistuti Sukta, https://en.wikipedia.org/wiki/Nadistuti_sukta

Dmytro said...

supernord said...

Just so you don't misunderstand...

1. The Anatolians could have migrated southward quite early from some area of the northern steppes. So the first comment is irrelevant.

2. The horse was so unimportant in the ancient Hittite language that they needed to borrow a term for it. And the Luwians got theirs from a later IE group for the same reason.

So that's my basic hypothesis. But you can take it or leave it. Doesn't particularly matter.

TruthPrevails said...

@supernord

Go easy on hoama, over consumption has many side-effects.

That is what I said you have to find the other six which have to line up with whatever you want to call Saraswati,

mzp1 said...

Saraswati has internal etymology. Harakwaiti doesnt, it is derived from Saraswati.

It is locally known in South Asia that S sounds are more original where Iranian has H/K.

This is an early forerunner of Centumization, along with vowel shifts between IA and Iranian.

Linguistics has IA diverged from PIE because of mostly Vowels and Centum/Satem. IE Linguists CHOSE those more 'European' sound shifts to represent PIE, thus taking PIE away from Vedic, although they are otherwise very close.

This is problematic because IA is in all other ways the most conservative branch of IE, and those exhibiting greatest vowel divergence and centumization generally the most divergent.

Problems are caused by the model for all IE languages except Iranian-Balto-Slavic-Germanic. Hence Anantolian Theory, Greco Aryan and Indo-Hittite.

You've already made Anatolian a sister-language. Just do the same for everything else except Iranian and Balto-Slavic-Germanic and then your theory will be right.

TruthPrevails said...

@mzp1

thanks for that

"Saraswati has internal etymology. Harakwaiti doesnt, it is derived from Saraswati. "

Thats a clincher.

supernord said...

Dmytro said..
"Just so you don't misunderstand...

1. The Anatolians could have migrated southward quite early from some area of the northern steppes. So the first comment is irrelevant.

2. The horse was so unimportant in the ancient Hittite language that they needed to borrow a term for it. And the Luwians got theirs from a later IE group for the same reason."

No, I understood you perfectly well and I answered these two points - that you are mistaken, but you do not understand.


@Sanuj
"Saraswati has always been the river between Yamuna and Sutlej, as attested by the Nadistuti Sukta"
Propaganda cannot prove anything. You're just wrong, but a preconceived notion like it was anti-scientific can't win.

"Vedic √sṛ"

√sṛ roots are not only Vedic, < ser- and selos- are common PIE roots.

Sanuj said...

@supernord

"√sṛ roots are not only Vedic, < ser- and selos- are common PIE roots."

Not present in Haraxvaiti, but in Saraswati. Latter cannot derive from former.

TruthPrevails said...


Since after gong round and round people finally come to ask how do you explain the PIE-FU contacts if the homeland is South Caucausus or India.

Here is some more on IE-FU connections. From a published professor of FU studies at University of Rome.

I have argued here, in line with the
views of Janhunen, that many conventional FU/U vs IE correlations
are also most likely the result of chance resemblance, particularly the
correlations and reconstructions assumed to go back to the level of the
respective proto-languages.
This being the case, I personally would
not put too much effort into researching “pre-historical convergence”.
However, there is no reason to abandon this field of studies altogether
if one concentrates not on (more or less) distant, unrealistic protospeech
communities, but on real languages and real language contacts
or, at least, language contacts that are plausible geographically, temporally
and “contextually”, in the sense of Janhunen’s concept of “areal
context” mentioned above.


This is a good read to understand the reality about PIE/IE to FU links.

http://jeful.ut.ee/index.php/JEFUL/article/download/jeful.2014.5.2.02/56

Dmytro said...

supernord said...


0+0=0

mzp1 said...

ok BATMAN, thanks for the advice. I'll be sure to send the signal if I need more help...

supernord said...

@mzp1
"Not present in Haraxvaiti, but in Saraswati. Latter cannot derive from former."

you are completely wrong because you do not have any knowledge. To reason without knowing anything on the subject is free freaking.
PIE. *selos-wnti "water lake area" > IA. sarasvati, I. haraxvati(/si).
Avestan is not Proto-Iranian language and it does not cover all the Iranian languages.
There were a lot of Iranian languages that are now extinct.
In the Iranian name of the river Yaksart, -sart is the same word.

Rob said...

Yes Troy VI / VIi looks barbarian but that’s LBA. We’re looking at Proto-Phrygians etc now

Anthro Survey said...

@Davidski

What I'm asking is whether Tepecik, Barcin and EEFs are too similar to successfully fit into a scheme like this(only terminal portions of a potential setup shown):
https://justpaste.it/4lf4c

Or like this:
https://justpaste.it/1ug7q

Matt said...

@Davidski, cheers.

Pretty much like Welzin_BA average, as expected. Look closer to present day Hungarians in this specific drift than the Hungary Beakers do (may be an effect of cutoffs on what samples included in this PCA for each group though).

Plots: https://imgur.com/a/P6hoNxP
(Datasheet if anyone wants it: https://pastebin.com/JLwZrtx6)

mzp1 said...

@Supernord,

Iranians did not speak *Reconstructed PIE, they spoke Iranian. The etymology of Harakvaiti is not found in Iranian literature. Those roots of the name are not attested as being used.

Sarawati is an IA river name and exclusively a part of Vedic Mythology, it is not cognated with other IE rivers names or mythology. Thus it is not considered PIE, or even IIr.

Harakwati River can only be derived from Saraswati.

supernord said...

Troy VI is not Phrygian absolutely. Troy VI was started at 1900BC.
In Troy VI is only seals on Luwian language. The names of the rulers also have a clear Luwian etymology, not Phrygian.
The Phrygians only appear in Troy VII, after destruction the city.

@mzp1
You're wrong. I know the scientific truth, and you've only shown your ignorance of anything. Your amateurish reasoning can only affect the same dilettantes.

Anthro Survey said...

@Dmytro:

"....lack of involvement in the "horse word" by Hittites ("Ansekura") and Luwites (a borrowing from the later Indo-Iranic "aswa"?)...very early separation from the rest of the IE world, more evident that that of the Tochharians even)".

This is also how I see it, and, yeah, we may never get the irrefutable DNA evidence. Issues like teasing apart EEF from ANF or the heterogeneity of East Balkan & Ukrainian groups circa 4000BC aside, we have to be relatively lucky and stumble upon one of their migration routes. Perhaps searching sites along the Halys river can increase our odds, though?

https://en.wikipedia.org/wiki/K%C4%B1z%C4%B1l%C4%B1rmak_River

Samuel Andrews said...

I'm thinking Anatolia_IA MA2198 is part Scythian. She dates before Ottomans but has Asian ancestry as well as excess Steppe ancestry.

2.2907"

Anatolia_IA:MA2198

Levant_N,19.1
Scythian_AldyBel,17.6
Barcin_N,17.5
Scythian_ZevakinoChilikta,13
CHG,10.6
Scythian_Samara,9.5
Ganj_Dareh_N,8.5
Yamnaya_Samara,4.2
Kangju_European_Kazakh,0
Kangju_Asian_Kazakh:DA123,0
TianShanSaka_Outlier:DA56,0
TianShanSaka_Excluding Asian outlier,0
Wusun,0
Nomad_IA:DA129,0
Nomad_IA:DA18,0
WHG,0
Scythian_Pazyryk,0
Sarmatian_Rostov,0

epoch2013 said...

@Dmytro + Anthro

Turkish Thrace, the European part of Turkey, should be sampled. It saw a pretty large decline of settlements during the late Chalcolithic according to this paper:

http://etheses.dur.ac.uk/3994/

Chances are there is something undiluted enough there.

From a link Folker provided:

http://www.hurriyetdailynews.com/tomb-in-istanbuls-silivri-years-biggest-archaeological-discovery--99235

TruthPrevails said...

@supernord

On one side you say Saraswati is a reflex and corrupted form of Volga like below you mentioned.

Volga (`Ρᾶ < Raŋhā < Rasa, Rava < hrava < Srava)

And then you turn around and also show us the etymology of Saraswati in PIE.
PIE. *selos-wnti "water lake area" > IA. sarasvati, I. haraxvati(/si).

What is true? Either it is corrupted form or it has a proper etymology cant be both, but since you cite both to
suit the argument I am inclined to think you are arguing for the sake of it with nothing concrete.

Samuel Andrews said...

I have doubts that the Hungary Scythians actually spoke Scythian. I'm thinking they were locals who adopted Scythian culture. All Scythians show BMAC, North Asian HG, and Asian ancestry except the ones from Hungary.

The Hun-Sarmatians are almost entirely Asian. Their Asian ancestry is really similar to ShamankaEN. Hun Tian Shan is like earlier Iranians with additional ShamankaEN ancestry. Ottoman:MA2195 looks like 50% ShamankaEN and 50% Iranian-Steppe+Anatolia. What are we to make of this? Were the first Turks most of ShamankaEN-like decent? They definitely had a lot of it.

TruthPrevails said...

@speculation

I do not know anything about Anatolian.

But seeing many of the words mentioned here and elsewhere the suffixes and the case endings sound very Dravidian especially Kannada/Telugu.

Maybe that is the effect of other languages in that area, as that area has been mentioned as being multilingual.

Grey said...

Anthro Survey said...
"Perhaps searching sites along the Halys river can increase our odds, though?"

given the lack of evidence of a land route a sea-then-river route seems the most likely to me - chalcolithic Rus

Dmytro said...

Anthro Survey said...
epoch2013 said...

I agree that both of these researches might well yield something. At any rate it's better to continue looking for evidence as long as this is possible. We're a long way from exhausting the available data "just waiting there".

Anthro Survey said...

@Old Europe

Don't recall specific numbers, but, yes, in terms of their deep ancestry, ANFs are less basal-rich than Levantine Neolithics and more enriched for Villabruna-related/UHG ancestry. In fact, this is easily visualized on simple West Eurasian PCAs.

My point is, though, that we don't know the exact demographic origins of the earliest Neolithic people in the southern Balkans. The Neolithic transition in places like France and Hungary clearly involved an influx/colonization of people from the Balkans, sure, but beyond that, we can't be too sure of the details until we at least get our hands on some early Sesklo. It may not be as clear cut as Anatolian agriculturalists jumping the Bosphorus and mixing with Greek hunter gatherers.

Consider the following scenario which would produce very similar autosomal signatures: Basal-admixed populations already exist in Greece and Macedonia around the late Mesolithic. During early Neolithic, or circa 7500BC, additional settlers hailing from either Anatolia or Cilica/N.Syria or both(introducing hg E?) arrive and introduce some new material elements.
Archaeologist Mehmet Ezdogan noted a collapse of north Levantine PPNB societies around the said date, which could have triggered a radiational efflux.

The Greek early Neolithic isn't a carbon-copy of what we see across the Aegean, exhibiting plenty of original motifs.

@Epoch

Thanks for the link.

old europe said...



@all

Hittite came from here, Hittite came from there, the route is this , the route is that, let for once apply the occam's razor maybe they were there since millennia ( above all if their genetic profile is confirmed) Otherwise we could validate the old chinese say:
" When the finger points at the moon, the idiot looks at the finger"

old europe said...

Anthro

thank you, I was not referring to you....maybe the Hittites are the product of WHG + Levantine ( or Natufians?) ( which could have also a western contribution of R1b V-88 if I remember correctly. In this sense they could be defined native to Anatolia

Chetan said...

@TruthPrevails Sarasvati is indisputably, from the PIE etymology supernord posted. And Harakhvaiti was perhaps the Iranian adaptation of the IA word.

But I don't think neither of these facts would allow you to draw the conclusions you seek.

Matt said...

Samuel Andrews: I have doubts that the Hungary Scythians actually spoke Scythian.

This seems like a reasonable suggestion, with the obvious caveat that I believe the Scythian language or languages are not attested and we don't know what they are. Which obviously makes it hard to tell.

At the same time I'd note, I would guess the languages of these people in Eastern Europe and the unattested Scythian language Iranian peoples in the steppe and Central Asia would have probably been in constant contact at a low level of frequency, certainly if they've been freely exchanging culture. Possibly via a currently unsampled intermediary population that on a genetic level would be a bit like other Scythians and a bit like Hungarian Scythians? Particularly if a population is trying to create solidarity, then they might emphasize aspects of speech they have in common.

So their way of speaking and language may have evolved some common features with Scythian language(s) and some intelligibility, even it is not exactly the *same* as Scythian.

Chetan said...

@Truth Prevails

http://jeful.ut.ee/index.php/JEFUL/article/download/jeful.2014.5.2.02/56

"many conventional FU/U vs IE correlations are also most likely the result of chance resemblance, particularly the correlations and reconstructions assumed to go back to the level of the respective proto-languages."

He's talking about the deeper connections in pronouns and verb ending between PIE and PU, not the very real Indo-Iranian loanwords in Finno-Ugric. I don't think anyone disputes that.

TruthPrevails said...

@chetan

"He's talking about the deeper connections in pronouns and verb ending between PIE and PU, not the very real Indo-Iranian loanwords in Finno-Ugric. I don't think anyone disputes that"

My point was about the location of homeland and the artificial constraint introduced by PIE-PU contacts, and since you agree there is no argument over that.

IIR-FU is understood as central Asia was Iranian speaking.

TruthPrevails said...

@TruthPrevails Sarasvati is indisputably, from the PIE etymology supernord posted. And Harakhvaiti was perhaps the Iranian adaptation of the IA word.

Yes that was the point, which it seems supernord was not able to understand, that Saraswati is IA.

postneo said...

Yes Kannada Telugu and Georgian have similar suffixes

aniasi said...

Sarasvati is IA, yes, but it is also Proto-Indo-Iranian.

Old Indo Aryan (Vedic, Chandamsi, Aryam) preserves more features of Indo-Iranian. The full series of aspirated and breathy stops, two post-alveolar series, and the accenting of vowels are all shared features of Proto-Indo-Iranian and Old Indo Aryan.

Iranian, on the other hand, reflects a greater number of innovations. This may be due to the fact that it has a later attestation, but it also comes down the fact that early Vedic schools tended to emphasise the idea that 'good language' was a sign of religious propriety.

By the time of Panini, it is clear that the "Samskritam" way of speaking was a decidedly artificial manner that embraced a clear continuation and imitation of historic religious forms, as opposed to "Prakrtam" speech that was more innovative, used by the majority.

In a sense, Iranian constitutes a number of older "Prakrtam" forms that evolved into the Gathic, Avestan, Median, Persian, and Eastern Iranian languages.

mzp1 said...

The best evidence yet...

"Ancient Greek and Vedic Sanskrit were also strictly mora-timed. Proto-Indo-European probably was as well."

https://en.wikipedia.org/wiki/Isochrony

You need long vowels for Mora-Timing.

Iranian drops long vowels and becomes syllable-timed, and this causes Balto-Salvic-Germanic to become stress-timed, as with Farsi.

Ancient Greek was Mora-timed, and possibly Latin and Celtic too, but these are now syllable timed languages.

Reconstructed PIE cannot be Mora-Timed because it doesnt have the vowels. PIE vowels need to be closer to Greco-Aryan. This shifts the whole IE tree into the correct position.

Mark Moore (Moderator) said...

I was hoping you could do a post on the paper that this article is based on... http://www.dailymail.co.uk/sciencetech/article-3320218/Missing-piece-Europe-s-hunter-gatherer-ancestry-discovered-Group-settled-Caucasus-45-000-years-ago-left-genetic-mark-modern-Europeans.html

They seem to leave Basal Eurasian out of the mix, but a lot of things they say about CHG, the "new" strain you have been talking about for a while, sounds like BE. But that makes no sense because I remember you measured Kotias and Sartublia for BE and they had rather low percentages. BE is not the same as CHG right? They just come from about the same area and were both supposed to be isolates.....?

Singh said...

@Jaydeep

"Here is the archaeological supplement A to the Daamgard et al paper,
https://zenodo.org/record/1243933#.Wvl19aSFPIU
The Conclusion of the section on South Central Asia is worth quoting in full

This survey of the archaeological and biological record of southern Central Asia yields four important findings. First, contacts between the sedentary food-producing populations of the Namazga culture populations residing in Kopet Dagh piedmont and Geokyur oasis of southern Turkmenistan who likely established the outpost at Sarazm had little to no contact with populations residing in the southern steppe zone. Second, contacts between Bronze Age steppe populations and NMG V and BMAC populations appears to have been one in which the dynamic of cultural influence was stronger on the side of the well-established sedentary foodproducing populations, and this resulted in the partial assimilation of these initial newcomers to the region both culturally and, to a lesser degree, biologically as well. Third, not all of those who emigrated from the north turned to farming but may have continued a semi-nomadic existence in the highlands, which were unsuitable for the kind of intensive farming practiced in the BMAC homelands or in the regions of Khorezm. Fourth, if there was any Central Asian influence on South Asian populations, that influence likely long predated any development of Iranian, let alone Indo-Aryan, languages, and most likely occurred during the late NMG IV to early NMG V period (ca. 2800–2300 BCE) and even earlier during the Eneolithic from Kelteminar culture groups (4000–3500 BCE)."

Interesting highlights Jaydeep, especially last paragraph.

Narasimhan et al. of Central & South Asian paper noted this on his twitter last month.

"I'm not beholden to the steppe origin of IE, it could also be the IE is spoken in Neolithic Iran, meaning that the the people of the Indus could be speaking IE, and then there is a meeting of 2 divergent IE strands. We are working on a synthesis-Please be patient. Data is so new."

He speculates there could be meeting of 2 divergent IE strands. IMO, maybe 1 of these divergent strands is not Indo-European but something Indo-Uralic from Central Asia maybe?

Singh said...

@Chetan & @Jaydeep

Under Indo-Uralic theory, what would WSHG's and Kelteminar culture be speaking? We have many dead languages in Central Asia before IE migrations

It's possible there was some Indo-Uralic spoken there and influenced IVC language in SA before I.E migration.

"Dravidian languages display typological similarities with the Uralic language group, suggesting to some a prolonged period of contact in the past. This idea is popular amongst Dravidian linguists and has been supported by a number of scholars, including Robert Caldwell, Thomas Burrow, Kamil Zvelebil, and Mikhail Andronov."

Narashiman tweeted this about Uralic connection..

"The supposed connections between Uralic and Dravidian have been highly fanciful and speculative. aDNA however, shows that Indus Periphery had a WestSiberianHG~ ancestry, as do ancient Fennoscandians, ridiculous or reasonable?"

Santosh said...

@ Singh,

"Under Indo-Uralic theory, what would WSHG's and Kelteminar culture be speaking? We have many dead languages in Central Asia before IE migrations"

I don't know at all and am completely speculating here, but probably pre-Burushaski, more widespread (Dene-?)Yeniseian, etc.?

Also, could you please mention more connections between mesolithic Gangetic traditions as inferred from archaeology and later day Indian traditions, if any, as you did in a post recently about a possible importance of fire in mesolithic Ganga valley? Thank you very much!

TruthPrevails said...


@postneo

1) Georgian and Dravidian languages are agglutinative.
(which also includes Elamite,Gutian,Hattic,Hurrian,Kassite,Lullubi,Sumerian,Urartian in the vicinity)

2) Scripts of Georgian and Kannada/Telugu look visually strikingly similar at first glance.

I could not find any more similarities than that. But if you say they share similar sounding suffixes as well then ok sure.

Santosh said...

@ Singh

"Narasimhan et al. of Central & South Asian paper noted this on his twitter last month.

"I'm not beholden to the steppe origin of IE, it could also be the IE is spoken in Neolithic Iran, meaning that the the people of the Indus could be speaking IE, and then there is a meeting of 2 divergent IE strands. We are working on a synthesis-Please be patient. Data is so new."

He speculates there could be meeting of 2 divergent IE strands. IMO, maybe 1 of these divergent strands is not Indo-European but something Indo-Uralic from Central Asia maybe?"

Very interesting! I wonder if anyone studied those 300-odd non-Indo-European words in Rigveda with the Indo-Uralic lens? Is any preliminary Indo-Uralic reconstruction done also or is it just a tentative proposal right now?

TruthPrevails said...

@singh

I had seen that tweet from vagheesh. Further in that thread he dismisses that there is any truth to it.
He just seems to be firing up the lemuria theory followers in a fun way.

Also there is no way to confirm that the second strand coming from Central Asia to South Asia is Indo Uralic, because Indo-Uralic borrowings are all one way, so we dont receive anything from the second strand to even attest it. There has to be only one real strand the first one.

As Vagheesh also speculates Iran Neo could have been native to South Asia as well, but which can only be confirmed through a Mesolithic DNA which can change the direction of the first strand too.

Ric Hern said...

@ Open Genomes

Thank you.

It looks like Hajji Firuz was a major trading settlement connecting East, West and Northern traderoutes.

Since Northern Bell Beaker People were so keen on adopting new arts and crafts I wonder if Maykop evolved in a similar fashion ?

postneo said...

dravidian and noth-central and western Indian, sanskrit are all mora timed languages.
Persian and eastern Indic languages are not. in fact both farsi and bengali have more backed and low vowels.
dravidian languages are extremely metrically conscious probably the most on earth.

Chetan said...

@TruthPrevails

"My point was about the location of homeland and the artificial constraint introduced by PIE-PU contacts, and since you agree there is no argument over that."

There are PIE loans into Uralic, possibly resulting from the northern contacts of Yamna/Poltavka. But the Indo-Uralic connection goes deeper than even that. Once you see how similar the pronouns, verb endings and root structures of PIE and PU are, it's very difficult to believe that these languages did not have a genetic relationship in the deep past. Pronouns and verb endings are almost never borrowed between languages. But of course there is no consensus yet.

@Singh

"Under Indo-Uralic theory, what would WSHG's and Kelteminar culture be speaking? We have many dead languages in Central Asia before IE migrations"

I don't think there is a properly sketched out model for Eurasiatic/Nostratic yet and not many linguists accept those proposals. You might want to check out Allan Bomhard's writings. According to him, Eurasiatic split (which includes Indo-Uralic + Altaic) occurred somewhere in Central Asia around 8000 BC. But the theory doesn't have many supporters.

epoch2013 said...

@ Samuel Andrews

Do you have any idea if mtDNA H7a1 is somehow related to Turkey?

mzp1 said...

@Chetan,

You didnt read my post.

PIE needs to be a Mora-Timed language. It isn't thus any proposed relationship between PIE and Uralic are highly speculative ie. simply wrong.

Either that or the relation is only to North Western IE Languages.

Timing is a very early and deep aspect of a language. It only goes More-Timed> Syllable-Timed>Stress-Timed in IE.

We cannot talk about about speculative relationships between theoretical Proto languages if the basic reconstructions are shown to have major problems.

mzp1 said...

@Postneo,

Do you have any links confirming North Indian languages are still mora-timed. I read somewhere they are syllable-timed but that didnt make sense to me.

Chetan said...

@mzp "PIE needs to be a Mora-Timed language."

Is this your own theory? And what are its implication for current IE linguistics?

But some of the other theories you have proposed here are absurd. Like how Indo-Aryan is more closely related to Greek than to Iranian. Even a layman can see why that's wrong.

mzp1 said...

@Chetan,

IE Linguistics are not concerned with Timing. I dont know why. But PIE does need to be Mora-Timed. Mora-Timing is probably the reason Greco-Aryan and Latin and Celtic have stronger Prosodic traditions than Iranian-Balto-Slavic-Germanic.

It's not my own theory. It's a basic fact...

Mora-Timing>Syllable Timing>Stress Timing. Structured > Unstructured.

Stress Timing is really bad for Prosody (Literature) compared to Syllable, and Mora-Timing, which is the only 'real' timing.

Implications are that Reconstructed PIE vowels need to be closer to Greco-Aryan than Balto-Slavic-Germanic.

This also brings PIE closer to Hittite and Armenian, so we dont need those Armenian and Anatolian, and OIT Hypothesis'.

I'm not saying IA is more close to Greek than Iranian. I am saying Greek is closer to IA than Iranian and Balto-Slavic-Germanic is closer to Iranian than IA.

Both Iranian and Greek are relatively close to IA, but Iranian loses long vowels early, which has a ripple effect further down the line, causing Balto-Slavic-Germanic to become very stress-timed compared to other IE languages. It is this vowel divergence that caused IE linguists lots of problems, and led to PIE being reconstructed further from Vedic.

Chetan said...

@mzp Indo-Iranian evolved out of the most "developed" PIE dialect ie. with full satemization, predominance of o-stems and a well developed verb system.

Tocharian, Italic-Celtic, Germanic and even Greek can't have evolved from Indo-Iranian because they all derive from linguistically "older" PIE forms.

There's no way Greek evolved from Indo-Aryan or Germanic Balto-Slavic evolved from Iranian. Simply false.

Chetan said...

Besides, Greek is attested from 1500 BCE, the same time or before Rigvedic Sanskrit. Can you tell me how Indo-Aryan can give rise to a Proto-Greek dialect in that case?

mzp1 said...

You are suggesting that IIr languages increased Form and Structure after IIr branched off.

But all real historical evidence points to IE languages losing Form and Structure.

It is obviously more likely PIE was more Structured than any of it's descendants, and that those IE languages simply lost Structure as they moved away.

You cannot go from Unstructured > Structured. And there is no way to go from Stress Timing > Syllable Timing, nor Syllable Timing > Mora Timing. Not theoretically and obviously never actually seen in IE languages.

You seem to be using cyclical reasoning in your arguments. It is only an assumption that IIr developed forms not seen in earlier IE.

mzp1 said...

@Chetan,

"There's no way Greek evolved from Indo-Aryan or Germanic Balto-Slavic evolved from Iranian. Simply false."

On the contrary, it is very clear.

postneo said...

Standard high school Hindi poetic meters(matras) have time allocation. Medieval Hindi poetry surdas etc never violate. Spoken Hindi also keeps things reasonably tight. Bengali ( I’m a native speaker) is extremely loose there’s very little length distinction except when forced to near geminates and compounds.

EastPole said...

@postneo
“Standard high school Hindi poetic meters…”

It is good that you look at poetic meters. There are very old links between Slavs, Greeks and Vedic Aryans in regard to poetic conventions.

In the Harvad publication “The Common Heritage of Greek and Indic Meter” we read about some similarities between ancient Greek poetic meters and Slavic meters:

https://s22.postimg.cc/a5wbsz7xt/screenshot_151.png

http://chs.harvard.edu/CHS/article/display/6442

The same archaic meters as in Slavic are also noticed in the most archaic phases of Rigvedic composition:

https://s22.postimg.cc/illwahasx/screenshot_149.png

https://s22.postimg.cc/jp60mfvg1/screenshot_150.png

http://chs.harvard.edu/CHS/article/display/6429

In my opinion these many links in poetry, religion and language between Slavs, Greeks and Indo-Iranians were much later than PIE. I would link them with the expansion of CWC –>Sintashta/Andronovo–>India confirmed by recent genetic discoveries.

It supports this model:

https://s17.postimg.cc/71fgzkv4v/screenshot_388.png

postneo said...

@mzp
Where did u here Hindi is syllable timed? Makes no sense. SunA sUnA pilA pIlA are totally unrelated in meaning.

Thx EastPole will review. At first pass Germanic and Slavic grammar don’t seem to converge without Sanskrit

mzp1 said...

@postneo,

Some website or wiki. Was skeptical myself.

Can you use Avestan or some Iranian form to converge Slavic and German instead of Vedic?

Lee Albee said...

I have been re-reading some 2016 papers " The Demographic Development of the First Farmers in Anatolia" and "Genomic Evidence Establishes Anatolia as the Source of the European Neolithic Gene Pool"

These papers demonstrate that CHG like ancestry had been seen spreading across Anatolia beginning in the Neolithic/Chalcolithic (Kum 6 6700 BP). They also find this relationship in the Remedello group in the Chalcolithic (northern Italy). And they propose that this relationship was likely mediated "These results point to gene flow from an eastern source into Chalcolithic Kumtepe and later into Europe, which could have crossed central Anatolia already before the Chalcolithic."

So finding CHG-only heritage in Bronze age Anatolia or in the Minoans for that matter should not be surprising. It has been circulating in Anatolia and Southern Europe for a good time now.

Association with a language is of course not possible but "the establishment of the Anatolian Indo-European branch in Anatolia is likely to have happened through a long-term process of infiltration and acculturalization rather than through mass immigration or elite dominance (Melchert 2003: 25)." So if by some off chance early PIE was spoken by some people with CHG heritage, pre-steppe, the slow integration of these people seen in aDNA may explain this pattern--or not. The putative age of PIE, when aligned to other languages does place it at ~8k years old--so timing is not necessarily too far off either.(http://www.pnas.org/content/pnas/early/2013/05/01/1218726110.full.pdf)

Just some thoughts