July 03, 2012

Proto-Indo-European and North Caucasian

Quite consistent with my idea that Proto-Indo-European is related to the West_Asian autosomal component.  This component occurs at a a level  greater than 50% level in modern North Caucasian speakers, is absent in Europe prior to 5,000 years ago, and occurs at levels greater or equal to 10% in most present-day Indo-European speakers from Europe.

Transactions of the Philological Society DOI: 10.1111/j.1467-968X.2012.01309.x

Areal Typology of Proto-Indo-European: The Case for Caucasian Connections

Ranko Matasović

This paper re-examines the evidence for early contacts between Proto-Indo-European (PIE) and the languages of the Caucasus. Although we were not able to find certain proofs of lexical borrowing between PIE and North Caucasian, there are a few undeniable areal-typological parallels in phonology and grammar. Some features generally attributed to PIE are not found in the majority of languages of North and Northeastern Eurasia, while they are common, or universally present, in the languages of the Caucasus (especially North Caucasus). Those features include the high consonant-to-vowel ratio, tonal accent, number suppletion in personal pronouns, the presence of gender and the morphological optative and, possibly, the presence of glottalized consonants and ergativity.

Link

21 comments:

German Dziebel said...

"Quite consistent with my idea that Proto-Indo-European is related to the West_Asian autosomal component."

You're spinning wheels here. North Caucasus is precisely where the Kurgan theory places the Urheimat of Indo-Europeans. And it's more Europe, than West Asia.

Dienekes said...

You're spinning wheels here. North Caucasus is precisely where the Kurgan theory places the Urheimat of Indo-Europeans. And it's more Europe, than West Asia.

Europe and West Asia are arbitrary divisions of the territory of Eurasia. Actually, they do carry some meaning in the sense originally used by the Greeks, since the Aegean is indeed a barrier between two peninsulas. According to the original definition, most of southern Russia, and certainly the Caucasus belonged to Asia. For whatever political reasons, the idea of Europe was extended, but personally I could care less where one draws the line.

The population of the North Caucasus is dominated by a component I've labelled West_Asian and which ties them primarily to South Caucasians, Armenians, Anatolians, Iranians and the like, rather than their European (primarily East Slavic) neighbors. You can call it whatever you want if you don't like the name.

German Dziebel said...

Okay, we're making progress: you switched from spinning wheels to creating a storm in a tea cup. Why don't you call your component "North Caucasian" and say that it's consistent with the Kurgan theory and with the areal similarities between North Caucasian and IE languages?

South Caucasians are a worse fit from the areal linguistic perspective, per the paper you referenced.

The real question would be whether this "North Caucasian component" is of admixture origin or of common descent origin in Europeans. IF linguists tie IE to North Caucasian languages (which is nowhere in the works but doesn't mean the link won't be found by someone like Casule or me), then the component will be due to common descent. Otherwise, it's admixture. Proto-Indo-Europeans resided in the Northern Caucasus area for a while and absorbed a bunch of North Caucasians.

Do you have any problems with this terminology and reasoning?

Dienekes said...

Why don't you call your component "North Caucasian" and say that it's consistent with the Kurgan theory and with the areal similarities between North Caucasian and IE languages?

You're wasting my webspace. I won't call it "North Caucasian" because it's modal in South Caucasian and Balochistan populations.

It is not consistent with the Kurgan theory, because the aforementioned component is very weak in present-day steppe populations. I do keep my mind open to the possibility, however, that the steppe may have acted as a conduit for some Indo-European groups emanating from the Caucasus, but we won't know for sure until we get some ancient DNA from the area.

By the way, you are also wrong that links between IE and North Caucasian necessitate presence of Indo-Europeans in the North Caucasus. After all, they speak North Caucasian there! It is much more likely that the Proto-Indo-Europeans lived in the Transcaucasus; this would explain their South Caucasian links discovered by Gamkrelidze and Ivanov as well as the abundant evidence for Semitic contacts.

And, indeed the links with North Caucasian languages could be explained via Hurro-Urartian which has been linked to Northeast Caucasian.

wagg said...

"the abundant evidence for Semitic contacts."

Their reality have been questionned by renowned linguists and even the considered most obvious cases (like for instance, septm / sabatum (7)) is thought by some to have arrived in the PIE vocabulary through intermediary language (and indeed some Caucasian languages also possess this stem).

And words such as for instance "goat" (latin haed-, dutch geit, etc... ) that have matches in semitic, can easily be neolithic wanderworts (or even picked late in this case and not part of PIE as it's actually a rare stem IIRC) and as such not very decisive as an evidence of a geographic origin.

Fanty said...

Thats rougly the +30% region of the component that is talked about:

http://www.freeimagehosting.net/dhokj

Sigh, I need a faster "free image hosting without need to register" ;-)

German Dziebel said...

"It is not consistent with the Kurgan theory, because the aforementioned component is very weak in present-day steppe populations."

What do you expect from the steppe - it's the place where populations churn like crazy. Mountains, on the contrary, are a refugium where older strata conserve. It's very natural for archaeology, rather than for genetics, to capture an ancient reality of a spread zone such as the steppe, while genetics to capture the same reality in the adjacent mountainous region.

"I won't call it "North Caucasian" because it's modal in South Caucasian and Balochistan populations."

Okay, you can call it "Caucasian." Depending on the genetic system and depth of sampling, this component or its analogues will likely shift a bit between Southern and Northern Caucasus, Northeastern and Northwestern Caucasus.

"the abundant evidence for Semitic contacts."

Wanderworter that don't prove a thing. You should take a class on it from Wagg.

libya said...

"Their reality have been questionned by renowned linguists and even the considered most obvious cases (like for instance, septm / sabatum (7)) is thought by some"

Who are those "renowned" linguists and who are those others "some" and what are their proofs for those words being arrived through an intermediary language and in wich Caucasian languages there does exist a word connected to "septm"
Septem could be a wanderwort but what about other words such as six,two,barley,wheel,sacrifice(daps)and even basic words such as I(ego)?

Justin said...

Placing IE homeland substantially above the Caucasus region allows it to border the Finnic lands to it's north and the Altaic lands to it's west, with Caucasians to the south. All the influences and relationships happen cleaner this way.

wagg said...

Libya:

"Who are those "renowned" linguists and who are those others "some" and what are their proofs for those words being arrived through an intermediary language and in wich Caucasian languages there does exist a word connected to "septm" "

Sorry, don't expect me to do a lecture about it. To know more about this, you'll have to read at least either Mallory or Anthony (and their sources). I know that at least Igor Diakonov questionned it though.

"Septem could be a wanderwort but what about other words such as six,two,barley,wheel,sacrifice(daps)and even basic words such as I(ego)? "

Most of them enter the category of wanderworts, and how many are really considered sure loanwords by the linguistic community anyway? I don't know about it (apparently many are discarded by several linguists (six and two, apparently, as septm is the only borrowed numeral widely accepted - BTW: I've written sabatum while it's more *šab'atam, sorry)), so I can't comment futher about it.

Anyway, I'd say the Caucasian and Semitic (possibly via intermediary for the latter?) loanwoards are easier to explain with a PIE urheimat in the pontic steppes than old IE (PIE? - not Indo-iranian anyway, which is also abundant in ancient uralic) loanwords within proto-Uralic with a proto-IE urheimat in west Asia, even if in eastern Anatolia (this proto-Uralic thing is mentionned in David Anthony's book about IE, in the part beginning p.93).

libya said...

@Justin
There is thousands of Km's between the Uralic homeland (northeast of the Urals) and the Altaic one (southeast of Siberia)

IE/Uralic&Altaic contacts were not at a level of PIE but at a level of IE daughter languages such as Iranic

The PIE/AA and Kartvelian contacts are at a proto level (and are also genetical not sprachbundic as the case of IE vis a vis East-Asian phenotyped Altaic and Uralic peoples) so the PIE homeland should be seen (as explained by Gamkrelidze&Ivanov) as the mountainous (thus explaining great diversity in rather small territory) area spanning between the eastern tip of the Elbruz and the western tip of the Taurus on an east-west axis and between the southern slops of Pontus&Georgian Caucasus and the northern slopes of the Taurus on a north-south axis (nearly 1 mln km square area,[France, Germany and Italy combined] that is very rich in metals and fertile valleys)

Also Altaic and Uralic peoples came much later to Kazakhstan (wich was an Iranic land) and north europe so that you could say that north of Caucasus there was Uralic speakers and east of Caucasus were Altaic speakers by as early as the Bronze age!!

terryt said...

"Placing IE homeland substantially above the Caucasus region allows it to border the Finnic lands to it's north and the Altaic lands to it's west, with Caucasians to the south. All the influences and relationships happen cleaner this way".

That is the most likely situation as I see it. In fact there was quite likely 'originally' (whatever that means) a language chain, or sort of cline, connecting the languages. But the cline has been distrubed by the expansion at various times of descendant languages as well as language borrowings.

"It is much more likely that the Proto-Indo-Europeans lived in the Transcaucasus; this would explain their South Caucasian links discovered by Gamkrelidze and Ivanov as well as the abundant evidence for Semitic contacts".

It seems obvious that wether IE originated north or south of the Caucasus it very early moved through them. We have Hittite/Luwian to the south and a spread of languages fanning out from the north. The Indian languages most likely moved east through Baluchistan but to the west the pattern is far from clear. Is Phrygian related to Greek, and did Phrygian enter Anatolia from the west? Or is Phrygian related to Hittite/Luwian and evolve locally? The answer might tell us something about the pattern of western movement of IE.

libya said...

Hi!
According to this recent pape, there are not PIE/Uralic borrowings but rather late Indo-European/Uralic borowwings
http://www.mv.helsinki.fi/home/jphakkin/UralicEvidence.pdf
On the other hand IE contacts with Semitic and Kartvelian are at a PIE level, please see the paper in the Dienekesarticle below (the 2 papers are both short around 10 pages)
http://dienekes.blogspot.com/2011/05/dolgopolsky-on-two-homelands-of-pie.html

This fits with a PIE homeland south of Caucasus (with area north of Caucasus being an expansion area of early Indo-Europeans)

wagg said...

Libya:
"According to this recent pape, there are not PIE/Uralic borrowings but rather late Indo-European/Uralic borowwings
http://www.mv.helsinki.fi/home/jphakkin/UralicEvidence.pdf"


We'll see what other linguists will reply. It's unfortunate that the author you mentionned adress none of the major stem that D. Anthony mentionned (and the surmised contacts are not all about vocabulary though, it's also about declension case and pronouns).

He also claims that Tocharian (the most ancient IE branch, beside Anatolian) is out of question as source for the ancient IE loanwords in Uralic (*)). That's problematic because a proximity between both Tocharian and Uralic is a real possibility (e.g. with this rare stem, Tocharian kälk-, kalák- : to go / Finnish kulkea : 'to go'). Sometimes other features are also claimed to have been borrowed from Finno-Ugric (e.g. the fact that Tocharian had 9 noun cases and that some of them might have been directly borrowed from Finno-Ugric, the fact that tocharian endings are sometimes agglutinative (uncommon in IE))).

(*) He also bases it on an estimation of the date of proto-Uralic by 2 linguists. It might be flimsy ground, especially since his "archaic IE stage" he mentionned has to be clearly older than the proto-Aryan stage puts after in the chronology, but date-wise it seems to be problematic. There are good reason to think proto-aryan can't be older than 2500 BCE and younger than 2000 BCE (just look at rigvedic sanskrit and Avestan's resemblance). So with the fact that he links the south Siberian Afanasevo (starting between 3500-3200 BCE) with the Tocharian branch in his reasonning, how is that so out of question? And what would have spoken this archaic IE? and who? the ancestors of proto-aryans? AFAIk several stems of this archaic IE are not in indo-iranian. It looks like a different thing. Is there an archeological culture matching his requirement?

wagg said...

"a proximity between Tocharian and Uralic is a real possibility"

I was imprecise. I meant "a proximity between the PIE dialect that would be the source for proto-Tocharian, and Uralic"

eurologist said...

Baltic, Tocharian, and proto-Iranian are suspiciously ancient, compared to other IE languages. This points to the eastern Black Sea and beyond as receiving a few, early, and heavily isolated settlements versus a more ordered and more widely spread development on the western and southern seaboard.

The diaspora always stays behind - such as Island and Scandinavia did with respect to Germanic. Yet, Scandinavia eventually re-grouped due to Danish and German influences. In other words, even a large sea, with sufficient trade, eventually promotes language homogenization. If the latter is not apparent, peoples moved away into other, land-locked regions and retained ancient languages for a very long time.

Western IE languages heavily diversified more than 3,500 years ago - and as such, point to an origin that was widely spread at that time, west of the Black Sea.

Octavià Alexandre said...

IMHO, the flaw inherent to *all* the theories about PIE homeland is they wrongly assume the IE family is the result of a single linguistic event, whatever it was.

But actual data points to *multiple* language replacement processes happening over an exceeding long period of time, since the Mesolithic to the Iron Age. In fact, the Spanish IE-ist Francisco Villar, in Villar et al. (2011): Lenguas, genes y culturas en la prehistoria de Europa y Asia suroccidental, detects several dialects of a paleo-IE phylum by studying the ancient toponymy and hydronymy of Europe and SW Asia.

I think the "PIE" IE-ists is rather a cross-section of the last (post-Neolithic) stages of the IE family, to which the language spoken by the nomadic shepherds of the Pontic Steppes (which I call Kurganic) highly contributed. However, they're wrong in identifying Kurganic with PIE. The real PIE (i.e. the last common ancestor of all IE language) would have spoken much earlier, in the Upper Paleolithic, being part of a larger Eurasiatic phylum.

terryt said...

'the flaw inherent to *all* the theories about PIE homeland is they wrongly assume the IE family is the result of a single linguistic event"

I agree. I've just finished a comment at another post suggesting exactly that situation for the development of new species.

Octavià Alexandre said...

Which one?

terryt said...

"Which one?"

'Admixture Matters':

http://dienekes.blogspot.co.nz/2012/07/admixture-matters.html

Bear101 said...

Maybe a stupid question but how many words make up a language ? How many words of whole vocabularies today do "Indo Europeans" have in common ? If the word mother, mater, moeder etc. Is found thousands of miles apart but all other vocabulary is different from each other, does it really point to a common source of origin of a whole group of people ?

My view on Indo European is that it is a mixture of many languages by people previously mostly isolated from each other for thousands of years and with population growth and expansion eventually mixed and shared words becoming an economical force by means of the amount of territory and necessatating other languages to participate, add or adopt for economical gain and survival.

Seeing that there are Tocharian A and B, first Centum then Satem is quite interesting. Chinese drawings and descriptions do not show a difference in physical characteristics of the Tocharians wether they spoke Centum or Satem. This make me think of a more economical purpose to Satemisation, and not a different group of people with a Satem character replacing the previously Centum group.

So Centum must have been strongly represented in Central Asia to survive for quite a while economically before Satem became necessary. Lost links for trading to the west, due to the Mass Migration of most Indo Europeans might have been a key point for the forming of Satem. ?

Just some of my thoughts.