Re: [tied] Re: A New language tree

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

From: Pat Ryan

To: cybalist@yahoogroups.com

Sent: Sunday, May 08, 2005 8:29 PM

Subject: RE: [tied] Re: A New language tree

Patrick writes:

 

As I was intrigued by Richard's comments about questions that might need addressing in Kazanas' essay, I waded through all 22 pages of it.

 

Frankly, I found absolutely nothing that, I thought, needed a response. Richard, if there is a question you feel is in that category after reading these comments, and if you will post it, I will try to specifically address it.

 

The first thing to note is that Mr. Kazanas is obviously a pious man who reveres the ancient writings and beliefs of India. There is certainly nothing intrinsically wrong with that but if one wishes to practice linguistics in this time, one needs tools that are a bit less dated.

 

Kazanas' main point seems to me to be that the old Indian regular treatment of root grades, in three categories, Ø-a-a:, is "original" and that the PIE model for root grades, in five categories, is chaotically irregular, and so must not underlie the original system, which, he presumes a priori, was "regular". Since "regularity" was original, "non-regularity" shows a later provenance. Old Indian preserves the earliest PIE root-grade system, and must be earlier than any PIE-derived language that operates within the five-grade system.

 

The point is made only at the sacrifice of reality. No PIEist of whom I know has set up a five-grade root system. This is a fantasy in Kazanas' mind. He believes PIEists grade roots into five categories, as follows: Ø-*e-*o-*e:-*o:; why he did not add *a to the mix? You would have to ask him.

 

Kazanas brushes aside without any sustained comment the subject of 'laryngeals'; as a consequence, it is impossible for him to explain vowel length as anything but a derivational feature, which, he considers, is essentially derivational rather than a response to stress-accent. Anyone who seriously doubts the existence of 'laryngeals', although legitimate difference of opinion can be held on their affects, simply has no place discussing PIE matters.

 

My understanding is that all PIE posit three root grades almost identical to what Mr. Kazanas postulates for Old Indian: Ø, zero-grade; *e/*o, normal-grade, and *e:/*o:, Dehnstufe, or lengthened-grade. Three, Mr. Kazanas, not five! The choice of *e(:) or *o(:) in any given root does not change the semantic meaning, and can shift back and forth with the stress-accent for exactly the same word. These *e/*o shift in the same word or formant show rather conclusively that no derivational intent underlies the variation. As for the lengthened-grade, I personally doubt it is derivational in any PIE except Old Indian; in my opinion, it is an Old Indian innovation. But I am willing to be better informed on this point.

 

In my present opinion, length is occasioned solely by the former presence of 'laryngeals', just as *a(:) is.

 

I have also found that Lehmann's approach to the *e/*o variation works well. In any period during which the variation was fully active, any early root-accent will produce *é. A shift of the stress-accent to a immediately subsequent syllable will produce *é in the newly accented syllable. The formerly accented syllable will now show *o. If the word has three syllables, and the last syllable is stress-accented, one of the two preceding syllables will show *Ø. After the full operation of these mechanism, a re-accented root-syllable that had become *o remained *o, i.e. became *ó. 

 

Kazanas makes an inordinate use of the root nR- (PIE *Haner-), 'mature man'. He claims that the initial *Ha seen in some PIE languages like Greek andros, proves the primacy of Old Indian! I have a message to impart to him: it is far easier to lose something than to acquire. That argument is counterintuitive and counter-commonsensical.

 

He also expostulates at some length on the various forms *sreu- takes in Greek derivatives. His remarks, in my opinion, point to the need for some linguistic training of a more recent provenance. He lists forms like Greek rhé-o:, the erh-rhu-e:-ka, and then rho-ós, and then pointa to the chaos. I think Mr. Kazanas needs to be reminded that the s-mobile stem has the actual root *rew-; and that between vowels, it would disappear in Greek; in zero-grade, followed by a 'laryngeal', we would expect u in Greek. In rho-é:, the -w- between vowels disappears, and the earlier stress-accented *rhéu- becomes *rho-'.

 

Some of the things Kazanas writes are just plain uninformed. As a proof of Sanskrit priority, he cites dah, 'burning', a root which he claims PIE does not have; therefore, its retention in Sanskrit shows Sankrit's originality. Quite aside from the circumstance that dah might have been borrowed into Sanskrit later than the PIE unity, dah corresponds to PIE *dhegWh-, 'burn', with Grassmann's law de-aspirated the first aspirate before the second aspirate was changed into h.

 

He also claims that padyate:, 'falls', related to pa:d-, 'foot' ('what falls'), and the lack of corresponding PIE derivations, as well as the lack of a verbal root *pad-, show the secondary nature of PIE. This argument is especially hollow. All other PIE languages could have had this verb and concidentally lost it. But the facts being what they are, they suggest that Old Indian innovated after the breakup.

 

Finally, Kazanas makes the highly suspect argument that since the Old Indian writings name many gods, and that, at most, only two of these are present in any given PIE-derived language branch, the primacy of the Old Indian gods suggests the primacy of Old Indian. For this circumstance, which I do not essentially doubt, here are other sufficient reasons. The Old Indians moved into India at an early date, and wrote down everything they could. India was able to defend itself successfully in the intervening years, and very old writings have been preserved. Hence every PIE god is probably somewhere to be found in Old Indian documents. I find that continuity admirable.

 

The other PIE branches, which I presume originally had roughly the same pantheon, had no writing; some gods were lost through lack of transmission (oral transmission only transmits what interest it). On their wide-ranging travels, they came into contact with many other peoples who, in many cases, gave them new deities that fulfilled the same functions as the older inherited deities, and, so far reason, were more attractive. Still more inherited deities were lost. Also, because of lack of writing, we have only the most superficial knowledge of the pantheons of, e.g. the Germanic and Celtic peoples. Probably, additional deities were there that corresponded to the Old Indian pantheon but we will never know their names. And so on.

 

In conclusion, Mr. Kazanas must broaden her perspectives beyond his admittedly admirable homeland.

 

 

Patrick

***