2009/1/28 Arnaud Fournet <
fournet.arnaud@...>:
>
>>> >> > The problem with this Yeniseic word *ses 'river" is that it
>>> >> > very much looks like a LW from Uralic.
>>> >
>> The problem with you is that you decide on faith alone that the
>> loaning was from Uralic to Ket, not the other way around.
>>
> ========
> I cannot figure out how this word could have been borrowed by many Uralic
> languages from Yeniseian which then spread this name all about the place.
> On the contrary, I can figure out how the five Yeniseian languages possibly
> independently borrowed from some Uralic dialects (but not all) the attested
> forms.
> The places with Yeniseian hydronyms with that morpheme seems to be rather
> clearly identifiable, when Uralic hydronyms show no limitation at all.
>
> The next argument is s > t is supposed to be regular in Yeniseian, I'm not
> convinced at all by this idea,
Well, that the correspondence exists in Yeniseic is an undeniable
fact. You can hardly ignore the following. Unless you provide
plausible Uralic parallels to the following items, the existence of
the correspondence cannot be rejected:
/tuj/ (back vowel) "moon" < */suj/
/but/ (back vowel) "hare" < */be?s/
/dat/ (back vowel) "eye" < */des/
/kut/ (back vowel) "horse" < */ku?s/
/lat/ (back vowel) "vulva" < */rJOs/
/tak/ (back vowel) "squirrel" < */sa?qa/
/to_/l (back vowel) "spend the night" < */saGar1/
/tú/ (back vowel) "half" < */su-/
/tuk/ (back vowel) "back" < */suga/
/túl-/ (back vowel) "red" < */sur/
/útu/ (back vowel) "to sleep" < */xus/
/uta/ (back vowel) "birch" < */xu:sa/
/hu-kút/ (back vowel) "house" < */Xu?s/
/xúta/ (backvowel) "one" < */Xusa/
a) Notice that I only listed the clear cases.
It should be noted that Pumpokol is poorly attested. Some of the
records are evident mistranscriptions, mistranslations and
misinterpretations (sometimes the form is clearly Yug, for example,
the transcription is incoherent and unreliable, etc.).
b) Let me ask you a question, Arnaud: If Pumpokol /t/ is not a reflex
of PY (> also Ket) */s/, what is???
> whereas there is no doubt that s > l y t s depending on the dialect in
> Uralic is certain.
Let us not talk about /l/ here now. I agree it is far from certain
that the /let/, /lat/ hydronyms are Yeniseic. We can also leave /y/
and focus on the /t/ ~ /s/ variation, which IS attested in Yeniseic.
Hence, if all the words that show this variation are somehow descended
from Uralic, this pressupposes that Pumpokol was a neighbour to the
/t/-reflex Uralic group, whereas Northern Yeniseic languages such as
Ket and Yugh were neighbours to the /s/-dialects of
Ugric/Samoyedic/Uralic/whatever subgroup may have been there.
So, the phonological change /s/ > /t/ must have started somewhere (a
cultural centre of some sort?) and began to spread. By the time it
began to influence Yeniseic, its speakers had to have settled there
for some time in order to become affected. This areal spread,
supported by lexical diffusion, may have equally affected the native
vocabulary resulting what is actually a regular correspondence today.
Would that be acceptable for you?
Still, I don't think you can just throw the correspondence away.
Apart from the 14 regular cases above, we have the following somewhat
less regular (or otherwise problematic) items supporting the /t/
correspondence:
/úttU-/ (back vowel) "warm" < */xus-/
/kutte/ (back vowel) "penis" < */gVns-/
/tag/ (back vowel) "larch" < */sas/
/tUt-/ (back vowel) "to sit" < */sVs/
Also, remember that deaffrication of alveolars is characteristic of
Pumpokol (PY */c/ > /t/ in non-initial positions, PY */dz/ > /d/ in
non-initiail positions) and it has been extreme in Ket, where PY */t/
has merged with */c/ and /*c^/ (Pre-Ket */t/), and where PY */d/ has
merged with */dz/ and */dz^/ (Pre-Ket */d/), and much less so in Yugh
(PY */d/ and */dz/ > Pre-Yugh */d/).
As for the other correspondences, the situation is less clear, of
course, but the most transparent cases can be exemplified as follows:
/bic^-/ < */bis/ "evening" (front vowel, auslaut)
/bic^/ < */b[i]s/ "brother/sister" (front vowel, auslaut)
/ec^/, /eg/ < */?es/ "sky", "God" (front vowel, auslaut)
/cía-N/ < */si-/ "four" (front vowel, anlaut)
/ciku/ < */s[U]Ga/ "year" (front vowel, anlaut)
/cel/ < */so?ol/ "sleigh" (front vowel, anlaut)
We could add:
/b[a~i]rc^oj/ < */pa(r)sa/ "high/height"
/kónc^adin/ < */?es-/ "to sow"???
But they are both dubious. The first might be a loanword from Russian
/bols^oj/ "big" or contaminated by it. The second is quite opaque and
may not be related to the other Yeniseic words at all.
The following exampmles seem to be irregular, too, since despite the
front vowels, we have /t/:
/ur-áit/ < */xur-?es/ "rain"
/tet/ < */ses/ "river"
/tec^/ < */siG/ "night"
/teksul-/ < */sVNgVL-/ "stump"
As for the first one, this might be an old recording BEFORE the final
/t/ eventually changed to /c^/, or simply a misrecording of the
palatal. Both "river" and "night" show irregular plurals, suggesting
the vowels may have been originally back. As far as "stump" is
concerned, again, this word is not transparent. Maybe you can find
some parallels in Uralic, I haven't checked that out yet.
Another small group of words preserves (or re-develops?) /s/:
/xatUs/ < */QoTVs-/ "bed" (dissimilation due to /t/ < */t/)
/útamsa/ < */?alVs-tamsV/ "hundred" (dissimilation due to /t/ < */t/)
/kasna/ < */Kas-/ "to take" < Yugh
/jasoru/ < */(x)asVL/ "ship" < Yugh
/lUcU/ < */K[u?u]s/ "ghost" (doubtful)
---
/sa_t/ < */sa?L/ "crucian" (following liquid)
/sálat/ < */se:Le/ "deer (rangifer)" (following liquid)
/selJ-/ < */seL/ "bad" (following liquid)
/sogo/ < */si:-/ "to eat" (???)
> I'm still waiting for a conditioning factor in Yeniseic.
See above for some examples.
> The next argument is this is not at all the only LW from Uralic.
> Many words dealing with Siberian realia are borrowed.
Please, be more specific as to what exactly "many" means and which
words are borrowed from which languages. That "river" is a borrowing
is only an opinion. "Snow" looks plausible, indeed, but these are two
words. Animals, ok, realia can be borrowed, the usual source being the
culturally more prestigious language. While rivers keep their names
when you arrive, why would you not force the few leftover aborigines
to use your terminology for the goods you want to trade? You are the
majority, the authority...please, be more specific and list the
suspicious items together with the assumed sources. The data need to
be considered before the view can be accepted.
Anyway, I do not rule out the possibility of Uralic loanwords in
Yeniseic. Of course not. But I reject the idea that the /s/ ~ /t/
variation is not a valid correspondence. /t/ is by far the most
regular reflex of PY */s/ next to back vowels. At least until you
prove the opposite, Arnaud. ;-)
Best wishes,
Petr