Vadim:
> But at the same time I still don't see why Miguel's
> explanation must be necessarily wrong.
I see this all the time with theories. That's the trouble
with them... It's rare when a theory can be proven
conclusively beyond a shadow of a doubt to be false. Why,
look at the Out-Of-India hypothesis that still clings for
dear life and, yes, it's probably true here too in the
end that there's nothing to really show that it "must
be necessarily wrong" no matter how absurd the theory
often becomes. What you're asking for, absolute proof,
is not possible.
Note that word "absurd"? That's why, in the end, despite
all the reasons you think something should have existed,
it has to be tested against Occam's Razor as I point at
a nauseating amount of times already. Is it the simplest,
most efficient solution or is there something less
involved that can explain this and other things a
heckuvalot better?
If you keep waiting for something to be proven absolutely
false, you'll be clinging to weak theories till your heart
finally gives in from old age and it won't get you very
far. It's like being stuck on a crossword puzzle for
hours because you refused to think of the questions that
stump you in a simpler way.
>> The interrogative pronouns would suggest that IE *kW
>> corresponds to Altaic *k. Therefore if IE *wlkWo-
>> were to exist in Altaic, one would expect *k in it.
>
> Only if IE *-kWo- is not a suffix here.
If you wish, although I think a de-adjectival noun makes
more sense.
> Why necessarily adjective? And why necessarily descriptive?
This is what we see elsewhere with florofauna. For example,
we have *gWohWus "cow" (*gWo:us) (*gWehW- "to graze"),
*peku "herd" (*pek- "to comb (wool)"), *bHibHru- "beaver"
and Germanic *bero: (*bHer- "brown"), *su:xs "swine" (*seux-
"to suckle"), *bHerxgos "birch" (*bHerxg- "be white, bright",
"otter" from *udro-/*wedro- "wet", etc. I could swear there
was even an old post on this somewhere in the archives, listing
such words, possibly 6 months to a year old now.
Now how many eco-terms use *-kWo- in IE? Do you feel this is
a productive suffix here? Where do we find it? Certainly
we do find it but I can't think of other animal or plant words
with it.
> For *kwon the Altaic parallel exists as well: *kaNV.
Yes, I've seen mention of it supposedly existing in Altaic
or Uralic. I think the first to propose an outside
connection for *kwon- was Illich-Svitych.
At any rate, it's hard to tell. Looking at pre-IE myself,
I'd bet that *kwon- is a pretty old stem though.
> In Uralic *wete "water". So there's no reason to propose the
> primary meaning "to moisten" in IE. It is true that IE *w
> <=> Ural. *w.
Actually there is. The stem *wed- is used to form other words
like *wedro- "wet". It seems easier to call *wed- a verb,
rather than a noun. Afterall, we never see a noun **we:ds so
why would *wedro- be anything other than a verbal derivative
like *sedto- "seated" is from *sed- "to sit"? It's hard to
see a noun in Celtic *udskio- "water". Looks like a verbal
derivative to me. In fact, since the vowel changes from *e
to *o in these various formations based on *wed- like so
many other verbs, I can't conceive of it as a noun without
*-r or some other nominalizing suffix at the end of it.
Another thing I've noticed is that an old layer of IE (the
layer known as "Old IE" in my books) often attaches *-an
(> *-r/*-n-) to stems, sometimes for no real reason. Oddly, we
see this same prolific-but-pointless extensions in Mandarin
(eg: *dian "a bit", sometimes *dian-r) and there are the
l-diminutives attached onto Latin words where there weren't
any in Common IE (eg: oculus "eye" < IE *hWo:kWs, stella <
*xste:r "star").
This is where rabid Nostraic theorizing comes in...
So I'm presuming that there was a Proto-Steppe verb *wit?-i
meaning "moistens, dampens, is wet" and that this verb was
used to form a root noun *wit? "water". It was inherited
into IndoTyrrhenian, becoming *wet: by regular sound rules.
The word was automatically pronounced with a long vowel
(ie [we:t]) since IndoTyrrhenian shares an areal feature
with neighbouring Boreal, disallowing nouns with only one
mora. The two languages dealt with this common rule
differently as time went on (more below).
In Old IE, because of the pesky and productive *-an ending,
the word was expanded to *wat:an with stative *a-grade,
then Mid IE *wat:an > *wat:ar, then postSyncope eLIE *wadr.
After Vowel Shift in the last third of Late IE, the word
became *wodr.
In Uralic, the root noun survived but because of Boreal
phonotactics (Boreal is ancestor to Uralic/EA/ChukKam),
monosyllabic one-mora nouns (as opposed to grammatical
morsels like pronouns) were not tolerated and CVC nouns were
stretched to CV:C like IndoTyrrhenian and then CVC&. After
the merger of the three stop series to a single voiceless
one, this regularly produced Boreal *wit& > Uralic *wete.
(I think I've sometimes seen *wetä by others but I don't
know which form I should adopt.)
> Semantically it is the worst example I have mentioned. But
> phonetical correspondences are good: IE *w- <=> Ural. *w-
> <=> Alt. *b-.
I just can't agree with this yet.
> Where do you find Alt. *b <=> IE *m?
IE 1ps *me and Altaic *bi (I've seen *ben reconstructed as
it's found in Turkish, but whatever.) We can't relate a 1ps
stem with a 1pp stem in IE without having to explain why
there was such a switch. Since we don't need to when a
solution *m <=> *b suffices, this latter idea becomes the
optimal theory.
> but Alt. *b <=> IE *w exists as well. Note that Altaic has
> no *w at all. What is then the correspondence for IE/Uralic
> *w? I think that surely *b-.
I think it can just as easily be *w <=> Altaic ZERO. All you
seem to have is Alt *bol- <=> Uralic *wole- "be" although
I'm not sure what the basis for the *w- in Uralic is, other
than to make it look more Altaic-ish than it really is.
= gLeN