isolates

[Ralf-Stefan Georg]

Please add to the list of isolates mentioned so far the (recently extinct)
Kusunda (Ban Raja) language of Western Nepal. It has been subsumed under
Dene-Caucasian, however ( a fate not at all unlikely for *any* language
isolate). Most general reference books put it into Tibeto-Burman, which is
certainly not correct. A few months ago, I posted on NOSTRATIC-L a brief
summary of what is known about the language together with an (almost
exhaustive) bibliography. I'm still willing to send this to every one
interested.
Somewhat hesitantly, I feel compelled to mention yet another potential
isolate: Itel'men. Yes, I *do* know that it is commonly counted as
Chukchi-Kamchatkan, and it is *more than likely* that it does. But the best
connoisseur of that language, A.P. Volodin, is now of the opinion that all
the similarities (systematic correspondences among them) between Itel'men
and, say, Koryak, are due to areal convergence rather than genetic
inheritence. I had some fights with Aleksandr Pavlovich on that issue,
since the languages share something which I regard as the most important
thing for a realationship: morphology. While it is true that only part of
the morphology is truely shared and the verbal morphology of Itelmen shows
whole subsets which find no matches whatsoever in the Chukchi-Koryak
languages (roughly it can be said that the *shared* morphology is that of
prefixes, whereas suffixal verbal morphology in Itelmen seems to be largely
independent - *roughly* I said, please don't count the affixes now !). The
other thing: the lexicon seems to basically non-Chukchi-Koryak at all.
Non-cognate (at least not *demonstrably* cognate, possibly an important
caveat) lexical items by far outnumber those with visible Chukchi-Koryak
connections (yes, I do know that the majority doesn't count as such, cf.
Armenian).
Now, I'm still inclined to see Itelmen as being related to Chukchi-Koryak
(and those of you subscribing ALTAINET will know that I'm a hell of a
splitter ! [just kidding, Sasha!]) *somehow*, but the simple dictum, as can
be found in all the major handbooks on language-classification, that
"Itel'men is Chukchi-Kamchatkan" (and period) is certainly an
oversimplification. There *are* degrees of language relationship (certainly
there are degrees of *discernability* of language relationship),
relationship is simply not always a binary +/- thing. The recent discussion
on mixed languages (CIA, Mbugu, perhaps Mogholi) is likely to reshape our
views on relationship during the time to come.
In the grammar of Itel'men I'm currently coauthoring with A.P. Volodin we
will have to say something about the genetic position of that language
(although there are *much more * interesting things about Itelmen than that
!). I think we will be working towards some compromise wording on the
matter, and after the book we will have some single-authored papers on the
matter, fighting again, but one thing I'd like to say now: if asked, what
language family belongs Itel'men in ?, I'd say at the moment: *I don't
know*. And I'm actually working on that language .... (maybe if I weren't,
things would be much easier ... ). Working on a language is perhaps the
best antedote against the morbus classificationis praecocis, which seems to
be plagueing some people in the field.
 
 
 
Stefan Georg
Heerstrasse 7
D-53111 Bonn
FRG
Tel/Fax +49-228-691332+