Re: [tied] Sarasvati River

From: Piotr Gasiorowski
Message: 7450
Date: 2001-05-31

Dear Alexander,
Few things are "incredible a priori", but an etymological proposal should be laid out in some detail so that critics can evaluate it point by point. Linguists won't be impressed by the fact that two words sound similar. Historically unrelated words often do (e.g. English day and Latin dies), while genuine cognates may sound very different (Russian dva and Armenian erku).
The crucial part of any etymological demonstration depends on the knowledge of language-specific changes. We know quite a lot about the developments that have taken place in the various sub-branches, languages and dialects of Iranian. This means that an etymology ought to offer a plausible and pretty detailed scenario of the postulated process, constrained by what is already known about the history of Iranian. It is not enough to say that Haraxva(i)ti: may have been transformed into RaNha (which is an ancient form, by the way, not "later Iranian") via arbitrary manipulations like chopping off the first and last syllables and modifying the middle. Such steps are unacceptable -- not because they are universally banned (they aren't), but because they would require linguistic processes that certainly did not operate in early Iranian.
The traditional explanation of RaNh- as a derivative of *(h1)ros- 'moisture, dew' fares much better from the formal point of view, but is semantically suspect. Is the Volga moist? Well, just as moist as any other river. It is extremely long and wide, rather deep -- in a word, huge -- but is it characteristically "dewy"?
My tentative suggestion, made on the spur of the moment, that the name may be Indo-Aryan (*rangH-a- 'hastening, running') is not detailed enough at this stage to count as a serious proposal, to be sure, but might make sense in the light of the claims put forward by some Russian toponymists that several hydronyms in that area should be analysed as Indo-Aryan rather than Iranian. I know too little about Mordvin developments to decide if *-ngH- (or later Iranian -Nh-) > -v- is a possible process there, but such details can be checked without much difficulty.
If you need a parallel hydronymic formation, the River L/e,g (excuse the Polish diacritics) in SE Poland might well derive from *h1longWH-o-s.
As for your second question, the evaluation of Datia > Daik, etc. will only be possible if you are more specific about the source form, the historical background, and the languages that show the forms in question. As I said at the beginning, a lot depends on language-specific patterns.
----- Original Message -----
From: Alexander Stolbov
Sent: Thursday, May 31, 2001 12:06 PM
Subject: Re: [tied] Sarasvati River

Thank you.
Do you mean that first a linguistic scenario for "Harahvaiti > Ranha" is needed and only then every step of it can be considered as possible or impossible, probable or incredible?
BTW, can not Mordovian names for Volga - Rav, Ravo, Rava   help here? - If we accept in this case a borrowing from Iranian after *Sarasvaiti (common Aryan) > Harahvaiti (early Iranian) but before Harahvaiti (early Iranian) > Ranha (later Iranian).
I have another analogous question.
What about the theoretical possibility of
Datia > Daik, Daich (old names of the Ural river)?
I guess I'll receive an analogous answer.  But maybe such a development is incredible a priori?