Early Indo-European: The Caucasian Substrate Hypothesis and How It Shaped Proto-Indo-European

The Caucasian Substrate Hypothesis in Indo-European: Were the Caucasian languages responsible for the 'unusual' features of Indo-European as it developed out of Indo-Uralic?

In 1946, Uhlenbeck wrote about the "unmistakable kinship with Caucasian languages," marking the first serious consideration of a genetic relationship between Indo-European and Caucasian languages. This idea has evolved significantly since then. In the second part of our series on this fascinating topic, we explore how scholars like John Colarusso laid the foundation for Bomhard, Kordlandt and Matasović to expand on the hypothesis.

Starting with Bomhard’s 1994 proposition, we see how Proto-Indo-European (PIE) might have been influenced by long-term contact with Caucasian languages, despite not being genetically related. His extensive work, featured in the 2019 Journal of Indo-European Studies, highlighted these influences through grammar, pronominal systems, and lexicon.
We will look at Kordlandt's views, emphasizing the transformative impact of a North Caucasian substratum on Indo-European languages, altering their vowel systems, consonant inventories, and grammatical structures.
This video also covers the significant updates and debates surrounding the hypothesis.

We examine critiques by scholars like Johanna Nichols, who questions the extent of lexical matches, and David Anthony, who discusses archaeological and genetic evidence supporting potential bilingualism in ancient steppe cultures.

We further explore Matasovic’s reanalysis, highlighting specific loanwords and shared features between PIE and Caucasian languages, while considering the possibility of an intermediary language influencing both.
Join us as we dissect these scholarly contributions and the mixed reviews they’ve received. What do these linguistic connections mean for our understanding of early Indo-European development?

Share your thoughts in the comments, and don’t forget to watch part one (on Colarusso's Proto-Pontic) if you missed it!

Selected sources: (those not featured here will be found in-screen during the video)

⭐Multiple authors, (2019). Journal Indo European Studies. Volume 47.

⭐Bomhard, A. (1994). Comments on Colarusso's paper "Phyletic Links between Proto-Indo-European and Proto-Northwest Caucasian". Mother Tongue: Newsletter of the Association for the Study of Language in Prehistory.

Bomhard, A. (2023) Prehistoric Language Contact on the Steppes: The Case of Indo-European and Northwest Caucasian.

Ruhlen, M. (2006). Taxonomic controversies in the twentieth century. In New essays on the origin of language (pp. 197-214). Berlin-Brandenburgische Akademie der Wissenschaften.

Lehmann, W. P. (2002). Pre-Indo-European. Washington, DC: Institute for the Study of Man.

Chirikba, V. A. (2016). From North to North West: How North-West Caucasian evolved from North Caucasian. Mother Tongue: Journal of the Association for the Study of Language in Prehistory, (XXI), 1-10.

Nichols, J. (2010). Proof of Dene-Yeniseian relatedness. In J. Kari & B. A. Potter (Eds.), The Dene-Yeniseian Connection (pp. 266-278). (Anthropological Papers of the University of Alaska 5 (new series), special issue). Fairbanks: Alaska Native Language Center.

Lazaridis, I., et al. (2024). The genetic origin of the Indo-Europeans. bioRxiv.

Kortlandt, F. (01 Jan. 2010). Studies in Germanic, Indo-European and Indo-Uralic. Leiden, The Netherlands: Brill.

⭐Matasović, R. (2012). Areal typology of Proto-Indo-European: The case for Caucasian connections. Transactions of the Philological Society, 110(2), 283–310.

[christopherellis]

As a student of Kartvel and Adyghe, as well as Basque, Arabic and a handful of IE tongues, I know that loanwords do not require direct contact, even taking centuries to be transmitted. Eg, Persian Khane, room, is found ftom the Balkans through to the Malay Archipelago. ( balcony, han, &c)
Fascinating stuff, i have been so for two thirds of a century

[nascenticity]

honestly i think the idea that they could potentially detect an unattested language through similarities between two language groups is *way* cooler than those two groups just being directly related, but i’m pretty sure that the “cool factor” isn’t a huge component of academic discourse 🤔

[intelliGENeration]

What is your take on the introduction of articles in Ancient Greek and Latin (and their predecessors)? What did it look like before… and where did the influence come from?

[rezazazu]

This could be the greatest discovery in the filed. Knowing what made the PIE the way it sounded and was like.

[francisnopantses]

I think those cognates look very good and I'm intrigued by the barley word. There are so many unanswered questions about IE agricultural words and their distribution.

[glitchpoke]

really interesting question in regards to the different areas/times where the substrate influence occurred, especially with the idea of a very anicent ~NEast Caucasian influence and later influence between PIE-NW Caucasian coming possibly from an intermediate language; but seems odd that they're not considering the idea that the influence could come from Early European Farmers speaking related Caucasian-like languages. I've seen the shared word for sheep between Germanic and NWC used as evidence for an EEF/LBK substrate on specifically Germanic or maybe broadly on the NW 'core' of Beaker-like IE languages

[Rorgosh]

And interestingly, language models do not follow the events, processes what we know from archelogical record… If someone goes this supposed proto indoeropean language, we still see fragmanted languages. Maybe because the terrain is also fragmented….

[rashidegintkhoev]

It’s true, I’m Caucasion, Ingush, Galgha

[scottn]

I'm wondering - the words you give from. Matosevic's are all agricultural style words, goat, wool, grain, flour, horse are these like Potato and Train, loan words that came with the innovation ?? .Also - Kurdish in an ergative Indo-European language, but it exists in the same geographical area as ergative non-IndoEuropean Hurrian. I'm wondering if ergativity isn't base Proto-IndoEuropean, but a Caucasian element to the cocktail, and some languages got more dollops of Caucasian at more times than others? And if you opt for the Anatolian pre-proto-IndoEuropean theory, could there have been a first period of two way influence very early in Anatolia, then a second in the Yamnaya era in the steppe??

[LuisAldamiz]

Observations:

1. Bomhard premises are not incorrect but fail to account for Uralic substrate (and not just adstrate from the north) in the whole steppe region, which IMO is very clearly demonstrated by ancient DNA (notably the presence of East Asia derived mtDNA C and Siberian-admixed EHG autosomal DNA, all corresponding well to modern West Uralic peoples from Maris to Finns). It also fails to account for evidence of Caucasus or Zagros Neolithic populations moving northwards, which is again backed by ancient autosomal DNA: first the area especially at the Lower Volga basin was EHG (read: Uralic) and then it was a 50-50 mix of Caucasus (or Iran Neolithic, very similar) and EHG. My reading from this is that proto-PIE was one of many Neolithic West Asian languages and that it admixed in vocabulary with Uralic, as much as it did in terms of people/genetics, as these farmer/herders migrated northwards. They may have also been instrumental in spreading the concept of pottery southwards, as (counter-intutively) it was the Uralic peoples who first used potter in West Eurasia (a concept imported from China) and not the Neolithic farmers to their south.

This results in Indoeuropeans carrying Caucasian genetics around in their conquests and migrations but also Uralic/EHG genetics (handy when you have to discern with other peoples such as Pelasgo-Tyrsenians or Elamo-Dravidians also carrying Caucasus-like genetics, but not EHG, around totally unrelatedly to Indoeuropean expansions).

2. If the PIE grammar strongly diverges from the proposed Nostratic/Eurasiatic superfamilies, then, in good logic it should be out of them, just as Brahui is still Dravidian even if most of its vocabulary is now Indoeuropean or just as English is still Germanic even if half of its vocabulary is French-derived (and to much lesser extent from Latin or other Romances). This again seems to support that the Uralic influence is as substrate/adstrate and not as a proper phylogenetic root.

3. Unsure if "Pontic is real" but at least it seems more plausible than Bomhardt's theory. My take is that proto-PIE was one of many Caucasus-Zagros or Eastern West Asian languages to spread, plausibly from a northern location in the West Asian of "Fertile Crescent" Neolithic, maybe in historical Armenia but that it was surrounded by other languages, some of which would be ancestral to the Caucasian families (incl. Hurrian and Hattic) but also to Sumerian, Tyrsenian (via Halafian culture) and (probably from a more southernly source) Elamo-Dravidian.

[isimerias]

I thought I remembered a hypothetical intermediate population having been surprisingly detected in the new genetic papers. Seems like a lot of the pieces of the puzzle from different lines of evidence are fitting.

[Serendip]

In linguistics, and especially regarding previous states of languages that were never written and nearly nobody knows anything about them, anybody can claim anything, they spend their time insulting each other, and nobody comes any further.

[AxionXIII]

IE and Uralic just bordered each other, at a few different places, for thousands of years. Thats the connection. They did originate from the same sources.

[marjae]

Is there a side-by-side comparison of the proposals tying Indo-European and/or Indo-Uralic with Euskarian, pre-Greek, Northwest Caucasian, Northeast Caucasian, Burushaski, Turkic, Mongolic, etc.?
Also, since you are familiar with ittite and the other Anatolian languages, do you have an opinion of the various theories identifying Etruscan as an Anatolian or other Indo-European language?

[sebastienlopezmassoni]

Please no proto basque indo european theory, but I have read that Armenian had been influence by a Caucasian language and now we know that proto-basque and Paleo sardinian were the closest to the early anatolian that travel from caucasian moutains, balkans.Sargasyan Vahan found cognate/similar word between both languages.

[dorasmith]

It's hot air from the academic community. I ned mor information than this.

Howvr, since th peopl who gave rise to theYamnaya did live north of the Caucasus Mountains and the Caspian Sea, and there was a hybrid group that was part Caucasian and part Yamnaya immediately north of and in the Caucasus Mountains, it would be strange if their descendants didn't speak a language with part Caucasian background.

[dsyy]

lol the track at 16:30 had no business going that hard

[philipthornhill]

There's a lot I dont really understand here .... but the assumption seeems to be that 'Caucasian languages; were always in the Caucasus where they are now (?) Why so ? It seems to me that mountainous regions were often where once more widespread languages retreated to ....they are relics from previous eras in that sense. I'm thiking of Welsh (< Brittonic) in Wales, say, or Basque (< Aquitanian) in the Pyrenees. So the contact between PIE and the ancestor of, say NW Caucasian (assuming thats real) need not have been where, or adjacent to, where NW Caucasian is today that make any sense ?

[jahiemsterling]

😢 the native haplogroup i is paleo Caucasian splitting from ij this is the haplogroup that was present in the gravettian culture i might explain this Caucasian influence in pie

[merttuncer]

Indo Uralic is pure fantasy imo. Not only Uralic languages are grammatically extremely different they don't even originate in Europe. They originated in Siberia.