DMB

I don't have a subscription to the genome site, so I haven't seen the full article, merely the abstract. Where does it say that the men were all from Andra Pradesh? Presumably, more research does need to be done and I hope it will be, in Tamil Nadu and other places.

Please explain what you mean. I don't see anyone here doing that. Surely this theory is of interest to all of us, Indian and non-Indian, if we are interested in the origins of our contemporary world.

phaedrus

DMB

I don't have a subscription to the genome site, so I haven't seen the full article, merely the abstract. Where does it say that the men were all from Andra Pradesh? Presumably, more research does need to be done and I hope it will be, in Tamil Nadu and other places.

Well there has been enough coverage of this research in India Upper caste Indians like Europeans, lower castes like Asians: study

Please explain what you mean. I don't see anyone here doing that. Surely this theory is of interest to all of us, Indian and non-Indian, if we are interested in the origins of our contemporary world.

Just look at the headline of the above article, it gives you a clear indication on how such studies can be used in a very contrived manner. And when did i suggest that anyone of this thread was doing such things? Hindu nationalists have a different agenda for trying to rewrite history (atleast something has come out of it and there is careful scrutiny of the AIT now) and on the other hand part of the academics who want AIT to stand for whatever reasons

lpetrich

What did Whitney say?

lpetrich

But the anti-AIT folx nevertheless insist on mixing the two issues up. To me, it's like seeing some great political scam in the hypothesis that English was brought to England by Angle, Saxon, and Jute invaders.

copernicus

Phaedrus, this whole issue seems to have been made up out of whole cloth by Hindu nationalists who perceive some kind of disrespect for their national heritage. You have presented no evidence whatever that there is any kind of general conspiracy among western scholars to impose the AIT, which seems a perfectly reasonable theory in light of linguistic and archaeological evidence. There is nothing of substance to refute it.

You point to articles by people who have no formal training in linguistics, which, I assure you, is necessary to evaluate the linguistic claims properly. Elst only cites a few works, and he appears to have little or no knowledge of the mainstream literature on the subject. I offered you Gamkrelidze and Ivanov's article as an example of how one might go about refuting the standard homeland hypothesis. I don't recall Elst citing Gamkrelidze and Ivanov's hypothesis, which is yet another view of the PIE homeland. Unfortunately for Elst, it doesn't support his own speculations.

You ask whether I believed Gamkrelidze and Ivanov's hypotheses. Well, yes, I find them very attractive. I had a chance to meet Gamkrelidze about 20 years ago, when he was even more controversial. Their theories are based largely on a novel method of reconstructing PIE. They relied on our 20th century knowledge of phonological typology to arrive at their conclusions. That argument impressed me. Not being a specialist in diachronic linguistics, I am not sure what the 'mainstream' view is nowadays, but I think that Gamkrelidze and Ivanov have had rougher going with specialists in PIE. So I wouldn't say that the homeland issue is settled to everyone's satisfaction. To the best of my knowledge, central or eastern Europe is still the favorite. I would also say that nobody is seriously looking at northern India anymore, and Elst's rather light survey of the literature does little to shake the foundations of modern research. Gamkrelidze and Ivanov's work is a bit ground-shaking.

I also find the genetic articles cited here interesting, although I don't really know enough about that field to evaluate them intelligently. It seems very plausible to me that the color-caste system in India could have arisen via the invasion scenario. While there are exceptions, the upper castes do appear to be lighter skinned in general. Does anyone deny this? How do those who oppose the AIT explain the origins of the caste system? I can understand why they might be really upset at the DNA study, which seems to support the AIT.

Lpetrich: I'll try to find something to quote you on Whitney. He is said to have dismissed the work of native grammarians as largely rubbish. On the other hand, his own grammar of Sanskrit bears an extremely remarkable resemblance to Panini's work. Coincidence?

[ July 30, 2001: Message edited by: copernicus ]

lpetrich

phaedrus:
Could you give specific reasons for which you find the article not convincing?

LP:
Read on.

[earlier:]
However, that article claims that it was India that was the Indo-European homeland, and I believe that there are significant reasons to disagree. Consider the Harappans, who were settled and who had writing; the writers of the Vedas appear to have been nomads without writing. Furthermore, the Harappan script has still not been decoded, though I once saw a half-convincing Dravidian interpretation of it.

phaedrus:
How does the harappan writing link to the significant reasons to not believe about the IE homeland?

LP:
It's what language that that writing was written in that counts. If northern India had been the IE homeland, then the Harappan writing would have been of some early Sanskrit-like dialect. But it has yet to be convincingly deciphered, and an early Dravidian dialect is at least as plausible. Also, the Vedas were originally not written down, but memorized by professional chanters, who transmitted them for some centuries after their composition. So why wasn't Harappan writing used to write the Vedas? Why did India have to re-acquire writing?

[earlier]
This would imply that the Aryan presence in India is at most 3500 years old. Hittite and Greek speakers were already in their adopted homelands at that time, meaning that IE must have split a few millennia before.

phaedrus:
Aryan presence is 3500 years old??(Are you suggesting 1500bc, just like all those AIT chaps do?) How did you come to the conclusion?

LP:
The Vedas feature some things that are rare or absent from Harappan sites, such as horses and iron weapons; and they describe a society of small villages, again unlike the Harappans and their cities. So the Vedas must be post-Harappan, which yields an upper limit of their age of about 3500 years.

copernicus

Just a quibble here: whether or not Harappan was an IE language is irrelevant here. AIT could still be correct. It could even be correct that northern India were the original 'homeland' of PIE and AIT were still true. All that AIT says is that tribes associating themselves with 'arya' invaded northern India and subjugated the indigenous tribes. Those indigenous tribes might well have spoken an IE-related language. We won't know until Harappan writing is deciphered convincingly, if it ever is. However, the lack of horses and iron casts doubt on the hypothesis that Harappans were Indo-Europeans, as lpetrich says.

lpetrich

Interesting kind of plagiarism.

Also, I have unsuccessfully searched for examples of how Panini's grammatical rules work; from the descriptions I've seen, his rules seem a lot like present-day ways of describing grammar, understood in some generalized, mathematical sense. Here is a guess as to Panini might have described some English grammatical rules:

Indefinite article
If the next word starts with a vowel
then it is "an"
else it is "a"

Regular past tense / past participle pronunciation
If the previous sound is "t" or "d"
then the sound is /id/
else if the previous sound is voiced
then the sound is /d/
else the sound is /t/

I have had an abundance of programming experience, which is why I've used "else" here; the construction of these rules was inspired by common programming constructs.

lpetrich

copernicus:
[stuff on Gamkrelidze and Ivanov...]

LP:
My knowledge of comparative and historical linguistics is on an amateur sort of level, but I do think that G and I are on the right track with their attempts at typological reconstruction; that kind of reconstruction uses constraints derived from regularities among known languages. For example, if a language only has only one of sounds /p/ and /b/, it will always have /b/ and not /p/. This may be because /b/ is easier to recognize than /p/, and this is supported by how /p/ sometimes gets reduced to /h/ or to nothing.

However, G and I's proposed ancestral-IE migration seems rather roundabout to me, and the spread of domestic horses is from just north of the Black Sea instead of to the southeast of it (G and I's proposed homeland). The first domesticated horse ever found is from Dereivka, Ukraine; this horse had had tooth-wear marks typical of having had a bit in its mouth.

copernicus:
To the best of my knowledge, central or eastern Europe is still the favorite. I would also say that nobody is seriously looking at northern India anymore, and Elst's rather light survey of the literature does little to shake the foundations of modern research. Gamkrelidze and Ivanov's work is a bit ground-shaking.

LP:
That is quite correct -- north of the Black Sea or nearby is still the favorite IE homeland spot.

copernicus

First of all, I wasn't in total agreement with G&I's analysis. The voiced stops are less likely to be retained than the voiceless ones, and there are examples of languages with /p/ and not /b/. The problem with PIE is that reconstructed */b/ is rather rare, and it isn't clear why. What is really odd about PIE is that there are plenty of cognate sets to validate */bh/, and, if a language has /bh/, it usually has /b/. So the reconstructed PIE system seemed very odd. What G&I said (which I found attractive) was that the classical reconstruction of voiced stops really corresponded to glottalized voiceless stops, and that made all the difference. I can well imagine /p'/ being less stable than /t'/ and /k'/. It is easier to produce the glottal pulse when the vocal cavity is smaller. So one might argue that glottalized /p'/ is intrinsically harder to pronounce than /t'/ and /k'/, just as /g/ and /d/ are harder than /k/ and /t/. (The necessity of maintaining a pressure drop across the glottis for voiced sounds is easier with a large vocal cavity. So PIE */b/ really ought to have been retained longer than */d/ or */g/.) So I agreed with G&I's conclusions, but not necessarily their whole story.

That aside, I want to call your attention to an interesting URL: http://www.harappa.com/script/index.html , which addresses the views of 3 international scholars on the linguistic origin of Harappan. All agree with the AIT, which is accepted by most scholars. The prevailing view is that Harappan was Dravidian--for the reasons given by the scholars. However, one of the scholars held out for the idea that it was possibly not Dravidian, but not Indo-European either. Pakistan's Ahmed Hasan Dani felt that Harappan, an agglutinative language, might well have been related to another language family, e.g. Altaic. Once again, we see that serious scholars give no credit to the idea that Harappans were Indo-Europeans. Indo-European languages simply do not provide cognate sets to validate the idea of a city-based civilization, which the Harappan culture most surely was.

[ July 30, 2001: Message edited by: copernicus ]