Southern Arc - Anatolian split review

Source: TW

Finally the much talked about “Southern Arc” paper of LAZARIDIS etal has been published. Below are just some preliminary comments based on the authors’ statements & results in the main text (nothing deep). First some background:

  1. We have good evidence that the likely phylogenetic model for the languages including Indo-European & Anatolian is Indo-Hittite (Indo-Anatolian), i.e., Anatolian branched first, then the rest of IE split up from PIE.
  2. Balto-Slavic was a likely sister group of I-Ir in a ``satem’’ clade lodged deep in IE.
  3. Greco-Armenian & Italo-Celtic were likely solid clades as well.
  4. All these non-Anatolian branches can be clearly traced to the steppe on philological/archaeogenetic grounds.
  5. So far the Anatolia archaeogenetics has not revealed a strong steppe signal.
  6. The authors have assembled hundreds of genomes from west Asia, including Anatolia & covering the period when Anatolian languages were spoken. Despite this expanded dataset, they found no clear evidence for steppe presence in the period when Anatolian languages were dominant.
  7. Hence, they posit a variant of the Caucasian Hunter Gatherer (CHG) hypothesis that had made its rounds in the early days of the archaeogenetic foray into IE origins.
southern-arc anatolian-split
southern-arc anatolian-split
  1. As per their current version, there were two pulses of CHG ancestry into the steppe. An old one which was pretty deep in time & a more recent one which they believe also took some Anatolian Farmer/Levantine ancestry along with it into the steppe. This second pulse is postulated as pre-Yamnaya and critical for its founding.
  2. They also see a rise in CHG ancestry starting from an Eastern Anatolian locus & moving west.
  3. They take this movement of CHG into Anatolia & the second pulse into the steppe as the signal of original Indo-Anatolian spread – Anatolian obviously the movement into Anatolia & the movement into the steppe as PIE. They present this as the below map – something that brings memories of the Soviet workers, Gamkrelidze and Ivanov.
  4. However, in the text, the authors are more circumspect & mention that it could Maykop, Eastern Anatolia in addition to Armenia as possible sources from which this CHG pulse that founded PIE might have originated.
  5. They also admit that a much wider sampling in Iran, North Mesopotamia, and Azerbaijan would be needed for more precise mapping. Note that the Mesolithic individual from Hotu Cave from Iran is a classic CHG.

To summarize the Indo-Anatolian cladogenesis as per the authors’ conception:

  • A CHG population prior to the invention of the wheel (as the word for wheel is unattested in the reconstructed Anatolian cognates with IE) expanded into Anatolia ~6000 YBP. They spoke the ancestor of the Anatolian languages. In Anatolia they encountered at least two other major ancestries termed Levant pre-pottery neolithic & a Northwest Anatolian ancestry. The authors suspect these might be associated with the local languages like Hurrian & Hattic.
  • Another branch of the CHG population moved north into the steppe and merged with a HG-neolithic population which was a mixture Eastern HG & CHG. They took the precursor of the PIE language to the steppe. By the time of PIE on the steppe (earliest Yamnaya) these people had the wheel & the horse.
    • The Yamnaya then expanded east (Afanasievo=Proto-Tocharian?);
    • into the Balkans, back to Armenia & Greece-> Greco-Armenian & Balkan tongues like Thracian etc (Albanian might be only survivor).
    • Another wave into Europe after absorbing some Globular Amphora farmer ancestry-> Italo-Celtic.
    • satem was another branch of the group that interacted with Globular Amphora. A branch of that expanded east as the Indo-Iranians (Sintashta). They had a clear contact with the Uralics. While the authors state that: “Contributions of Indo-European to Uralic languages (spoken in the forest zone of Eastern Europe and Siberia) appear to have involved only Indo-Iranian speakers ~4200 years ago,” we think the picture was more complex. There is a proto-Uralic component that seems to reflect earlier loans from a language close to PIE. This was later augmented by 3x more loans from I-Ir & IA. This means that the Uralics & IE were close by in the Yamnaya period itself but the I-Ir& IA reentered their vicinity from the west, intensifying the contacts.

A controversial aspect of the paper is the authors tentative suggestion that the R1b1a1 V1636 Y Haplogroup might reflect migration from the “Southern Arc” -> steppe. However, they admit that old versions of this were found in the steppe & it could have come down from there. This would then bring to question their favored hypothesis & invert it to a total steppe origins for Indo-Anatolian.

However, if we assume they are right it brings back a question we have been asking for a while. Does it imply that the CHG group spread carried a broader range of “para-Indo-Anatolian” languages? For e.g., Greece had a CHG entry prior to the Greeks. What about such an CHG population moving east to Iran & India (contributing to Harappan ancestry) did they bear para-Indo-Hittite languages.

Of course “Southern Arc” hypothesis needs a much closer look before reaching anywhere close to an acceptance threshold. ity alam vistAreNa.

Does this raise the possibility of Harappa having spoken a language of IE family pre-arrival of sintASta folks?

Yes: If a version of Lazaridis et al’s hypothesis survives then we can postulate that a broader swath CHGs spoke para-Indo-Anatolian languages (rather than IE). One of these could have reached Harappan India. Was that the sole or dominant Harappan language? Less clear for now

Source: TW

Some have homologized Greek kalia (cottage) with Skt shAlA descending from a PIE root. Nostraticists have noticed North Caucasian ~kala for hut. Wonder about its implications for the so-called southern arch hypothesis for PIE. With the southern arc hypothesis brought back into fashion by recent archaeogenetic studies one can revisit some of the deeper relationships of IE. One possible candidate in North Caucasian. This link could swing either way just like the archaeogenetic evidence. On one side it might suggest a homeland close to the steppe Yamnaya flanked by the Caspian + Black Seas. On the other hand, it might not be incompatible with an Armenian (or a nearby locale) as hinted by the recent publication,