Wednesday, August 22, 2018

Arya-Tibetan case for the OIT

(Pragyata, 10 July 2018)

The latest linguistic progress for the OIT

In 2017, there has been a lot of jubilation in the Aryan Invasion Theory (AIT) camp over new genetic developments prematurely touted as proving an Indo-European (IE) or “Aryan” invasion into India. Mainly they proved themselves incompetent by claiming that the linguistic AIT had been proven with genetic evidence. The problematic aspects of the genetic picture they painted, I leave to specialists to deal with. Meanwhile, however, a different development is undermining the AIT, and it is one I am more competent on.

You can follow us
on Facebook too

NW-Caucasian transforms Indo-Uralic into Indo-European

Three years ago I stayed in nearby Leiden taking summer courses in Avestan, Hittite, and statistical methods in comparative linguistics. There, mainly in private discussions with Leiden professors of linguistics, I learned about the big new theory within the AIT camp: that Proto-Indo-European (PIE) is a continuation of Uralic ("Indo-Uralic") after the latter had been adopted by an originally Northwest-Caucasian-speaking (NWC) population.

Lexical items in common between NWC (now mainly Circassian) and PIE are few, and controversial. But two other things are emphatically in common and stand out: like the earliest PIE, NW Caucasian is an ergative language (a structure that developed into a different structure, viz. the well-known case system), and like PIE, it has numerous heavy consonant clusters, as opposed to the Indo-Uralic ancestress which, like her sister Dravidian, mostly has a consonant-vowel-consonant-vowel sequence. 

The Homeland of Indo-Uralic is deemed to be Bactria-Sogdia, at least according to Alan Bomhard, the main living theoretician of Nostratic, a mega-kinship stretching into or beyond the Glacial Age and connecting Indo-European, Uralic, as well as Afro-Asiatic (Semitic), Kartvelian, Altaic, Dravidian and others (but not Bantu, Sino-Tibetan, NW- and NE-Caucasian etc.). During a westward trek, it "converted" a NW-Caucasian population, which may then already have been restricted to its present speech area, but which may just as well have filled up the East-European or Central-Asian steppes. The Caucasus forms the southern rim of the Yamna (“pit-grave”) area in SW Russia and East Ukraine, where the standard AIT locates the Homeland of Indo-European. In the process, it took over this population’s clustering and ergative speech habits, yielding the typical features of PIE.

This theory, pioneered by Frederik Kortlandt and presently supported by Alan Bomhard, John Colarusso and the influential Leiden school of Linguistics, has its skeptics and enemies. Thus, Johanna Nichols has shown that there is no noteworthy lexical exchange between NWC and PIE, though there is a later influx to NWC from Iranian and Russian, which have dominated in that area. Indeed, many scholars, less equipped with Nichols’ expert knowledge on the Caucasian languages, are also sceptical of PIE’s indebtedness to NWC.

Nonetheless, it has become the most coherent and battle-ready version of the AIT. It is what the OIT is up against. So, I sat down to consider alternatives that respect all the data on which this theory is based.

The Tibetan alternative

Then already, I remarked that both ergativity and consonant clustering were as much characteristics of ancient Tibetan. These traits have disappeared in the related modern Chinese (though consonant clusters are still prolific in ancient Chinese) and receded in modern Tibetan, e.g. what is now pronounced as Gelugpa, the Dalai Lama's sect, is written after its older pronunciation Dgelugspa. But in ancient Tibetan, preserved for us by the conservative spelling in the Tibetan script, they were vigorously alive.

So, the population that turned Indo-Uralic into PIE, some 6000 years ago, may as well have been Tibetan. Indeed, their putative homeland in Bactria-Sogdia borders on Tibet. Moreover, even a slightly more westerly Homeland might have been in touch with Tibetan through a westward Tibetan expansion. It has been established that NWC (and its probable offshoot Basque, another result of a long-distance migration, from the Caucasus to Sardinia, then France and Spain) is related to the Sino-Tibetan family.

Indeed, NWC may have come about as just such a westerly colony of emigrant Tibetan settlers. Given that the latter family is only spoken in East Asia (Tibet, China, Burma) and is a demographic heavyweight, it is unlikely that it moved from the Caucasus, an inhospitable place where we find only linguistic left-overs (e.g. Kartvelian or South-Caucasian, featuring Georgian, was once spoken in much of West Asia), to China. Rather, it is North-Caucasian that moved westwards starting from Tibet. At any rate, even lexical loans to or from SW Caucasian in PIE will often also be found in Tibetan.

Regardless of that still-disputed link between NWC and Sino-Tibetan, it suddenly struck me that I had failed to realize the importance of Tibetan all these years. If the Homeland of PIE was located where Hindu tradition puts the Aryan ancestor Manu, viz. in Northwest India just south of the Himalaya, Tibetan must have been its immediate neighbour. It should then be eminently possible to find at least lexical and possibly deeper linguistic exchanges between the two. I do not mean between Tibetan and any Indo-European language: there are of course plenty of those with Sanskrit/Hindi (e.g. padma, “lotus”, borrowed in Tibetan as padma/pema). I mean the “proto” stage of the Indo-European family as a whole.

The implication is that some Tibetan words have crept into the pan-IE vocabulary, showing up in evolved form in Latin or Germanic. Such words form part of every comparative dictionary of IE (there are already a big handful), whose makers never realized such words’ status of loan-words. Conversely, some PIE-origin words should show up in Tibetan. In effect, to prove the Tibetan-PIE thesis, a few of such words are no less than necessary. If there is only borrowing from Indo-Aryan, that would fit excellently in the AIT scenario.

The only way to avoid this necessity is if there was a steep status difference between PIE and Tibetan. Thus, there are hundreds of Iranians loan-words in the Uralic languages, but none in known Iranian languages (perhaps some in the lost Iranian-Scythian languages spoken in what is now the Slavic world), because Iranian had a far higher status, as the language of a conquering and urbanizing population, compared with the Uralic language of “mere” reindeer herds. It is likewise not inconceivable that the Tibetan mountaineers looked up to the proto-Harappan civilization incarnated by PIE. Even then, Tibetan may still have conferred some words transmitted by any native population to any conquerors or immigrants: names of local materials, animals and landscape elements. (That would certainly be the case in the AIT, but then only affecting Indo-Aryan, not PIE.) Anyway, the search is on.  

If there had been known Tibetan words in Sanskrit/Hindi, my attention would at least have been drawn to the role of Tibetan in the linguistic landscape in which IE came about. It seems that somehow I still had not freed myself from the AIT prejudice that discards Tibetan from consideration on the same grounds as Algonquian or Maori: simply too distant from any possible Homeland. AIT champions congratulate themselves that Sanskrit but not PIE has Munda connections (e.g. kumara, “young man”, deemed to be “para-Munda”), which can however be explained by the Vedic tribes’ eastern contiguity to the Munda speech area, not shared with the northwesternmost IE-speaking tribes, which spawned the IE emigrations. Those not in the know of the linguistic state of the art, such as Tamil chauvinists, still think that Sanskrit does, but PIE does not, show a strong Dravidian influence; in fact, there was no Dravidian influence in the Vedic heartland, it came about during the Vedic expansion to Gujarat, where originally Dravidian was spoken.  

More Tibetan, less ostracized

To check this hypothesis, I decided to learn Tibetan. I enrolled for last summer’s speed-course of Tibetan, two weeks full-time, organized by the Orientalist department of the French-language section of my own Alma Mater, the Université Catholique de Louvain, Belgium (with parallel courses for other languages deemed to be “for freaks only”: Sumerian, Hittite etc.). However, not enough people volunteered for the course, so it wasn’t organized. Having many other things on my hands, I postponed the endeavour until some weeks ago, when good Tibetan handbooks fell into my possession by a lucky stroke: a local initiate in Lamaism got rid of her collection of Tibetan dictionaries and language textbooks.

At about the same time, however, I learned that someone beat me to it fair and square: Igor A. Tonoyan-Belyayev from St-Petersburg (which, incidentally, has a state-of-the-art Tibetan-Buddhist temple since 1911), who is fluent in Tibetan. He has in the last year written a few papers detailing lexical links between PIE and Tibetan. He is part of a multidisciplinary nexus of AIT skeptics. His friend Aleksander Semerenko’s collection of archaeological evidence for an emigration from India to Mesopotamia and Central Asia constitutes another major breakthrough, but falls outside the scope of this article.  

To be sure, I have never worried about the honour of discoveries, let alone about intellectual property rights. My name as an OIT defender has been made and I’ll share in the fun once the OIT gets accepted. I also know that in the field of Hindutva/secularism studies, ideas of mine have been borrowed without acknowledgment by even well-known professors, but I don't mind. With my name under it, they would only encounter lots of stonewalling and sabotage. So if someone else can make them accepted but under his name, that is gain. It has been said that "you can get nearly anything done, provided you are willing to leave the honour to others". So when I find that a project I am taking up has already been elaborated by someone else, I only say: so much the better.

Further, I was starting to feel lonely in the trenches of the pro-OIT struggle, on which only a literal handful of people are working, so I welcome any new voices. Contrary to the received wisdom, it is not true at all that the Hindutva movement or the Narendra Modi government has invested anything in making the case for the OIT. At most, some in there try to piggy-back on other people’s efforts, on the results of which they have taken a royal advance by ludicrously claiming that the Homeland debate is a thing of the past, long lost by the “foreign” AIT camp. But even that is less than official, for in state documents and on government websites, even the ruling “Hindu fundamentalist” Bharatiya Janata Party (“Indian People’s Party”, BJP) keeps on taking the Aryan invasion for granted.

Given the systematic and multi-pronged political use made of the AIT by the “breaking India” forces against India’s civilization and unity, this unconcern from a supposedly nationalist party must be seen to be believed. But most Western India-watchers have never seen the facts of India’s religio-political nexus with their own eyes and naïvely believe whatever the Indian anti-nationalist forces spoon-feed them. Even Western pioneers of the OIT tend to believe in a Hindutva hand behind the OIT. Well, contrary to what they believe, the BJP and its intellectual acolytes have no merit in this debate at all. 

In supporting a more easterly Homeland, our Russian Orientalist makes it a point to distance himself from "Hindutva". Though I am an avowedly original critic of Hindutva myself, starting with my book BJP vs. Hindu Resurgence (1997), I don't credibly have that option in Western academe, as anything at all said against Hindu-bashing there counts as “Hindutva”. So, anything I write in favour of the OIT has to overcome the nausea which most non-Indian academics have developed towards Hindutva. This distrust is partly because of the hate campaign by the likes of Steve Farmer, and partly because of the obnoxious show of incompetence and arrogance by lay Hindu polemicists on internet forums. At any rate, it greatly diminishes my audience. Probably Tonoyan-Belyayev doesn’t have that disadvantage.

The actual findings

This theory of a linguistic or at least lexical exchange between PIE and Tibetan is still in its infancy. A weak point certainly is that Tibetan is only attested in writing since 600 CE, some 4000 years after PIE unity. We may take a little help from the evolution in the related Chinese though, which is attested since the mid-2nd millennium BCE. But any comparison with NWC suffers far more seriously from the same flaw, being this language group being attested since only a few centuries and not having help from any related older language.

We will give a few promising examples, taken from the articles on Igor Tonoyan-Belyayev’s online page.

Bya rgod (“bird” + “wild”, “vulture”) is a Tibetan word yielding Vedic garut, “wing”, and garuda (“hawk”, Vedic celestial bird); and rgod ma (“wild one”, “mare”) yields ghoṭa, a widespread Prakrit word for “horse”, replacing the Vedic and virtually pan-IE word aśva. But these roots only affect Indo-Aryan languages and could be accommodated in an AIT scenario, where the Aryan immigrants would borrow some words from a native Tibetan population, which all the more westerly tribes speaking IE would miss out on.

But there are other cases, affecting the IE family as a whole. Words for “earth” often have the basic meaning “dry”, marking a distinction with swamps and rivers, and the related meaning “thirst”, e.g. Latin terra, related to Sanskrit tṛṣṇa, English thirst. The IE root *dhghom/dhghem, “earth”, whence Sanskrit kṣam, Slavic zemla, Persian zamīn, Greek chthōn, Latin humus, can likewise be derived from Tibetan skam, “dry”, skom, “thirsty”.

Germanic fehu, Latin pecus, both “cattle”, and Sanskrit paśu, “animal”, can be derived from Tibetan phyugs, “cow”. Clearly this word was not derived from Indo-Aryan, where the second consonant is satemized/palatalized as ś, but from PIE, with the original guttural k/g (Tibetan, like Chinese, doesn’t emphasize the opposition voiced/unvoiced but rather the opposition aspirated/non-aspirated), *peku-, as preserved in Latin pecus.

Latin tegere, “cover”, “thatch”, English thatch (< thak, cfr. Dutch dekken, “cover”, dak, “roof”) would be derived from Tibetan thog, “roof”, “cover”. Similar-sounding but distinct is Latin texere, “weave”, whence in English both texture and text, from Tibetan thag, “weave”, thags, “texture”. For a final example, Russian teplo, “warmth”, and Sanskrit tap-, “heat up”, may come from Tibetan thab, “hearth”, “oven”.

In Tonoyan-Belyayev’s approach, there are other exciting perspectives to this Tibetan connection, such as the grounds for the suspected shift in PIE from the ergative::absolutive contrast to the nominative::accusative contrast; the passive and the causative construction; and the genesis of grammatical gender. Further, indications are given for a small adstratal influence on Uralic, which must have been a western neighbour of Tibetan, probably in Sogdia. Enough material for some serious research. It is only beginning.


Gururaj BN said...

Probably, the article has gone over the head of most of the readers. It has certainly passed over my head. Hence, none of the trigger happy readers have left any comment on the article. Dr.Elst, may request you to tell me where can I start picking up some rudiments of Linguistics at least to the extent related to Indo Aryan languages, so that I can return to this article and try to understand it.

postneo said...

Dr Elst two random points:

1)One of your Tibetan examples reminded me about the word pair in English deck(bedeck, thatch implying a surface covering vs dock / tuck .. meaning insertion or subsurface related. Perhaps preservation of a weak contrastive vowel pair where fronted vowel implies surface and backed vowel implies volume related concepts.

In Bengali we have DhEk, DhAk(cover), Dhak, vs DhOk, DhUk (enter or gulp).

This does not seem an isolated case.

e.g spread vs sprout in English and phail(spread) vs phUl in Hindi (bloom, flower)

2) From recent ancient DNA publications, I recall that a botai genetic component is seen in the caucasus perhaps Maykop