Christina E. Valaki , you might be interested in this:

https://www.bbc.com/news/science-environment-34175224

shalom! - Juju.

Linguists regard the Basque language as a linguistic isolate. Rumors of its origin in the Caucasus are inconsistent with mainstream opinion among linguists. The late Larry Trask specialized in the Basque language. You might want to consult his history of the Basque language. Besides linguistic opinion, a Caucasus origin is inconsistent with the genetic evidence (e.g. Nasidze et al. 2013). You may also want to read Valverde et al. 2013, a study that reports a Mesolithic origin of the Basques. As such, they predate the arrival of Indo-European languages.

Nasidze, Ivan et al. 2003. “Testing hypotheses of language replacement in the Caucasus: evidence from the Y-Chromosome.” Human Genetics 112: 255-261.

Trask, R.L. 2014. The History of the Basque Language. Routledge.

Valverde, Laura et al. 2016. "European paternal lineage M269: dissection of the Y-SNP S116 in Atlantic Europe and Iberia." European Journal of Human Genetics 24: 437-441.

While in graduate school at Harvard I specifically examined the claims for links between Basque and North East Caucasian (NEC) languages as one of my qualifying papers. Basque has a simple sound system, based on an aspiration, tenseness, contrast, so a two way contrast, while NEC languages have systems based on tenseness in the voiceless series and the glottalic series, plus a voiced series - so a five way contrast. NEC languages share with Basque a three-point contrast in the coronal fricatives: s, s-apical, and sh. But NEC languages also have numerous affricates, while Basque has only 3. Furthermore, NEC languages have rounded sounds, absent from Basque. NEC languages also have emphatic or pharyngeal sounds and pharyngealized consonants and even vowels in some. So, because of the lopsided number of sounds in Basque as opposed to the average NEC language it is not hard to find words in some NEC language that have some resemblance to a given word in Basque. What I could not do was find consistent sound correspondences that would have established a genetic relationship. René Lafon was a scholar who spent his career seeking to establish a linkage. I myself doubt, however, that a genetic link can be proven.

Extremely interesting contribution; thanks a lot, John, that was really useful.

Christina

You are welcome!

I agree entirely with Michael and John and would like to add a methodological consideration: If enough time has passed, it may be impossible to show that languages are genetically related. An instance might be the comparison of Modern Irish and, say, Modern Armenian. Knowing that the two are – by their ancient pre-stages – genetically related, there is no way of establishing regular sound-correspondences (as a rule taken as a diagnostic technique) amongst them, not to speak of morphological similarities. This is to say: our traditional method of identifying genetic relationship is only suitable for a specific configuration of data – admittedly still to be defined more closely.

Yes, Michael is right on with his remarks. I took up the issue of such a "specific configuration" in two articles. First, I once published a crude analysis in which I posited two arbitrary languages, each with a typical number of phonemes, about 30. I also assumed a lexical replacement rate of 10% a millennium, which I considered to be on the low side. I ignored core vocabulary, which probably lasts longer than peripheral words. Nonetheless, taking two words of "similar" meanings and choosing sounds for each at random, I found that after 15,000 years it was impossible to distinguish between surviving cognates and chance similarity. That 15,000 is an interesting number because the New World was settled around 15,000 years ago. Apart from Vajda's work comparing Na-Dene and Yeniseyan,(Na-Dene is generally considered a late wave settlement), we have not been able to find linguistic links between any New World language family and any Old World one. The argument was in "Dumézil and the Details, in an issue of Cosmos (1998, vol. 14: 103-117). I assume "mono-genesis," that all language emerged from some single, small proto-language, perhaps earlier than Homo or at least earlier than Homo sapiens. Though I have no justification for this idea, I did draw up (in 1983, with versions since) a theory of language evolution in which I treated the problem as a logical one, not a genetic or historical one, but could not get it published because the editor of Language said "Chomsky says language had no evolution." What is important, however, is that the comparative technique has inherent limits on how far back it can reach without some supplementation. Second, Tom Markey and I have just published a piece where we try to address this problem (Journal of I-E Studies, 2018, vol. 46: 312-360) and offer some ideas about long-range comparison, using the notion of exaptation in a new way. We hope our effort will make long range comparison more rigorous and therefore plausible. Still, the 15,000 year mark looms as a crucial horizon for comparative work. The Mother Tongue adherents may be able to argue for a more distant horizon by relying on core vocabulary, but even they will have a horizon where chance similarity blinds their efforts and I am confidant that it will be much less than the "mother tongue" date of the emergence of H. sapiens (150,000? a date itself subject to debate among paleontologists) that is sometimes asserted.

Dear Michael and John,

Thank you indeed for your answers. There are so many aspects I hadn't even considered when I first asked my question. Being involved in neurolinguistics and language organization in the brain for quite a long period has made me forget how interesting historical linguistics used to be when I was an undergraduate student. Thanks again for opening again all those windows for me...

Christina

Thanks for your kind words, Christina! And most of all, thanks to John who has made clear some crucial aspects of long range comparisons. I will have to read the articles you mentioned (now in retirement, I don't have regular access to our journals). The 15,000 yrs you posit certainly are an upper limit beyond which no principled statement about genetic relationship can be made. It all depends on the quality of the data you have at your disposal. Being agnostic with regard to mono- vs. polygenesis (too long ago), I even have difficulties to judge in a principled way, say, the Indo-Uralic hypothesis, not to speak of any of the Nostratic hypotheses. And even one of our common hobby-horses, the indigenous languages of the Caucasus, offers more questions than answers when it comes to the problem of a common origin vs. multiple origins. - Speaking of the quality of data, I mean the requirement of strict sound-correspondences on the one hand (often disputed, though), but most prominently the lack of hard criteria for the identification of semantic "similarity". Can't go into details here. Definitely, there is still a lot (and thrilling things) to do .

Yes, much has been done by linguist J. Cejador in this direction. He proved many ideas about the same roots in languages

Thanks, Aleksandr

Could you indicate specific papers by this author?

Christina

"El lenguaje" und "Embriogenia del linguaje" see on archiv.org , auch andere werke sind downloadbar hier. Ich lebte im kaukasus und die meisten Georgien seien der meinung ihre sprache mit Basque gemeinsam ist. Sie sind deshalb stolz, weiss nixt warum naehmlich. Ossetinische sprache weist auch besonderheiten auf mit IE sprachen und Ungarische steht alleine in europa mit Altaischen in verwandschaft

Thanks a lot for the answer and additional information, I'll read the files and get back to you for discussion.

Best,

Christina