Re: [tied] About methodology...

----- Original Message -----

From: Glen Gordon

To: cybalist@egroups.com

Sent: Wednesday, August 30, 2000 10:48 AM

Subject: Re: [tied] About methodology...

Glen: However, these quantum-linguistic analogies may only go so far. You assert that greater and greater accuracy are possible up to a _certain limit_. Please prove that statement or gracefully withdraw the ad hominem.

What's ad hominem in my posting? "Parva"? I meant both of us and the informal level of our discussion as compared to the really deep debates between Bohr and Einstein. No problem, I withdraw "parvus" if you prefer to be "magnus". As for accuracy, see below.

Piotr: By stressing the tree/wave duality and its consequences I don't want to argue that "deep" research is hopeless. I only say that the comparative method has its limits  [...]

Glen: There you go again. How does one assert this rationally? What limit are you refering to? 2000 BCE? 4000 BCE? 10,000 BCE? 100,000 BCE? How can you be sure this affects Nostratic linguistics? How can you be sure this doesn't affect IE linguistics? I'm betting that you, along with many people before you, can't possibly support this viewpoint with any scrap of logic.

The limit probably varies depending on the local ethnolinguistic and cultural configuration. Without scholastic sophistry it's enough to look at the quality of concrete reconstructions at various time depths to support this viewpoint empirically. Compare three well-known and well-studied families -- IE, Uralic and Afroasiatic. What we have is:

IE: a more-or-less clearcut reconstruction (scores of securely evidenced lexical items, a very complete picture of word-structure, derivation, morphophonemics and patterns of inflection).

Uralic: a less clear-cut but still credible reconstruction (possibly signifying a slightly greater time depth).

Afroasiatic: a very much less clear-cut reconstruction (which suggests that AA isn't strictly monophyletic or is significantly older than IE). I consider AA to be a borderline example. Decades of research have not made AA reconstructions anywhere as accurate and convincing as IE ones.

What about Nostratic? And please don't tell me stories about the teething troubles of Nostratic linguistics. Illich-Svitych died in 1966. Dolgopolsky has been an active scholar for some 40 years now, Bomhard published "Toward Proto-Nostratic" in 1984. Where's the progress? Where's the "greater and greater accuracy"? Where are the long-promised etymological dictionaries of Nostratic? From decade to decade I see the same weary lists of lookalikes and hear the same mantras. If Nostraticists are treated as a kind of sect by some of their colleague, they have themselves to blame for it. For lack of real progress they spread enthusiastic propaganda and maintain the illusion that everything has already been demonstrated, so that all a novice really needs is faith. My favourite quotes are from Vitaly Shevoroshkin (University of MI, Ann Arbor):

"At each conference on Nostratic outside Russia, a few attending colleagues happen not to be Nostraticists but linguists who know something about Nostratic research (which in itself, is commendable). Such colleagues may be divided into two groups -- namely, those who are Cautious but Curious, and those who are Prejudiced. The latter category seems to be hopeless: they simply never learn (though they may be knowlegeable experts in their own fields). -- Curious but Cautious colleagues usually would like to learn more, and more than one would join the Nostratic club and contribute valuable papers to the field (as has happened in the United States over the last two decades)."

As a matter of fact, Illich-Svitych reconstructed about 750 roots and Dolgopolsky's still unpublished long list is said to contain more than 2300 cognate sets (not to mention grammatical paradigms, pronoun systems etc.). Wow, that's nearly twenty times the number of reliable root equations supporting Uralic! Why doesn't everyone join the club at once? You seem to believe that the sociology of science provides an answer. You may be partly right, but the other reason is, quite simply, that the vast majority of those reconstructions involve poorly substantiated sound correspondences, arbitrary variation and loose semantics.

Glen: To a degree I agree. Still, the relationship of Inuktitut to outside language groups, whether it is to be viewed as a group of closely knit languages or as a dialect continuum, can still be mapped out in a tree diagram, if not for complete accuracy but for the sake of clarity and brevity more than anything. Again, a tree model isn't something to be thrown away. It has its place.

I agree. Splits do take place and may become permanent divisions -- though, mind you, not so unbridgeable as the gaps between distantly related eukaryotic species in biology. I'm suspicious of dogmatic "arborealism" as a methodological principle, but accept the practical usefulness of "family trees" as such.

Piotr