> Why is it that Pumpokol _only_ has a back vowel,
> when the other languages do not ?
> This is _dirty_.
>
> A.
>
> ========
>
> a) Pumpokol went extinct around 1750. Who knows what the precise vowel
> quality was.
> b) And yet, this is the observation. I will return to the problem later,
> anyway.
>
> =====
>
> You stated this example was clear,
> Now you say that we don't know what it was.
>
> A.
> ======
>
No. This is not the point I tried to make. You misunderstand. What I
have been saying all the time is that transcriptional variation in the
Pumpokol data is completely normal. It is normal to the extent that we
cannot really rely on them, we can hardly infer anything from a
difference in the non-native orthography caused by imperfect
perception and transcription by non-native speakers.
If you have a better explanation, please, put it forth right now and
here. Is /but/ a loanword? Then state clearly where it could be from.
Whether /a/ is back phonetically is not that important, but yes, it is
much further in the back than /e/ and /i/. At the same vowel height,
there is no opposition to be sure. Ok then. Let me rephrase the
descriptive rule: ALL vowels except /e/ and /i/.
> =======
>
> ok
> Then,
> why has Pumpokol /a/ in tat "river" when the other languages have *e ?
> What is the reconstruction of that word within Yeniseic if it is a Yeniseic
> word ?
>
> A.
> ========
No, Pumpokol doesn't have /a/ in /tat/. It has /a/ in /tataN/, which
is the plural form. The variation is parallel to Ket/Yugh/Kottish
singular /-e-/ versus plural /-a-/. Hence, we could reconstruct */ses/
in the singular and */sas/ in the plural (in the post-tonogenetic
Yeniseic). Is there a way to explain this by the Uralic origin?
>> t in Pumpokol where Ostyak and Samoyedic have t as well.
>> Conservation of *s elsewhere.
>> This "coincidence" is troublesome for a "spontaneous" Yeniseian
>> development.
>
> Well, whether spontaneous or contact-induced, it doesn't matter. These
> things can spread across language borders.
>
> ========
>
> funny,
> but you seem to be ready to accept shared phonetic innovations
> without any lexical borrowing,
> i wonder how this can happen in practice.
>
> A.
> ========
I have never stated there was NO borrowing at all. Of course, if there
was a /t/-type Uralic substratum in Pumpokol, that could have
"ignited" the Proto-Pumpokol fortition of /s/ and the
affricates/clusters. It's a pity we know so little about the
socio-linguistic situation in Central Siberia of those times.
Since some of the affricates (/ts/, /dz/ and /ts^/) have velar
reflexes in anlaut positions, I wonder if the same is true for the
relevant Uralic groups, as that might lend additional support to your
hypothesis.
>
> i haven't but I'm interested to do so !
>
> A.
> I'm not sure I have it. I'll try to find it though.
>
> ====
>
> It would be nice.
> A.
> ========
I will try hard.
> How come that none of the languages surrounding Basque share its
> typological profile? Why has Basque not influenced Romance
> considerably?
>
> ======
>
> Because Basque is a fossil language !
> A.
> ====
Please, define what "fossil language" is. I do not see how this can be
an explanation of anything in any way.
>> There is no doubt that Basque is not like Indo-European, nor Berber and
>> that's a major reason to think it's autochthonous.
>> Yeniseian is not like Mongolian, Turcic, Korean, Uralic, Tungusic,
>> Japanese,
>> etc and that's a major reason to think it's _not_ autochthonous.
>
> I don't get it.
> Basque is unlike the surrounding languages, hence it is autochthonous.
> Yeniseic is unlike the surrounding languages, hence it is not autochthonous.
> Interesting. :-)
>
> =======
>
> The issue of common lexical items and common typological features also plays
> a role in the location of PIE.
> I suppose nobody (a little serious) would locate PIE in an area full of
> languages with which PIE shares nothing.
> A.
>
> ========
Ok. So, what are the homelands of the Proto-Uralians,
Proto-Tungus-Manchurians, Proto-Koreans, Proto-Turkish, Proto-Japonic
in your opinion? I would like to know more about the picture you
envisage.
>> Which part of Siberia could Yeniseic be native to ?
>
> Until you prove that ALL the vocabulary relating to the Siberian
> lifestyle is borrowed from the surrounding languages, Yeniseic still
> can be considered as originating in Siberia.
>
> ==========
>
> I'm sorry,
> but the burden of proving Yeniseic autochthonous is on your shoulders as
> well.
> And your "stalinist" requirement is absurd.
> You already have no argument at all.
Well, I disagree. Why would the burden of proof that Yeniseic is NOT
from somewhere else than where it is located today or from somewhere
else than the oldest records seem to suggest be on MY shoulders?
> I have already provided about 20 clean cognates between Yeniseic and PIE
> and you have provided NOTHING that supports Siberian autochthony.
They are comparanda rather than cognates. Nobody has accepted your
theory yet. Yes, it might change in the future, I might even be the
first one to accept it, why not, but 20 examples are a very weak
proof, if proof at all. Just keep up the work and prepare a more
detailed, more convincing, more coherent theory and I am ready to
accept it once it makes a good sense. I am by no means against that
possibility a priori.
> By the way, Maloletko who is the collector of the hydronyms we are
> discussing makes Yeniseic come from the Caucasus Area.
>
> A.
> ========
Why not, but arguments does he have? I just insist that Yeniseic may
have been spoken in Siberia for a bit longer time than you propose.
Yes, it may have come from a totally different area some time in the
deeper past, but you have not given us evidence enough to reject their
"autochthonous origin". (Moreover, what does autochthonous mean?) The
formerly Yeniseic speaking areas may have been overridden by
non-Yeniseic speakers leaving little or no traces. Substrate influence
is not obligatory. Its extent cannot be predicted, especially when we
do not know what the socio-linguistic situation was like.
> ========
> These Siberian languages that have that same syntaxic and morphological
> system are most probably also related.
> And they have plenty of lexical items in common a well.
> Yenisean is therefore not related to this "macro-Siberian" phylum.
> And it's therefore not Siberian at all.
No. Yeniseic is not "Macro-Siberian". That it is also not Siberian has
not been shown yet. It might be, but it has not. The Tungusic and
Turkic speakers are known to have overriden the autochthonous
populations, for example. Russian, too, has been a major "deletor"
there.
Anyway, precisely which typological features link Yeniseic to PIE, for
example (apart from the 4-way gender system) and how stable
diachronically are these features?
> How long will you defend the absurd theory that a language can originate in
> a place surrounded for 50 000 years by languages with which it shares
> NOTHING but a handful of loanwords ?
>
> A.
======
I am not defending anything even close to that. I think the Yeniseians
may have moved to Siberia at the same or earlier date and none of what
you have stated so far has proven the opposite. My personal view is
that they might have come from the more southern and more eastern part
of Siberia, but I have not accumulated enough evidence yet.
And, I dare say we have very little idea about what languages (if
any!) where spoken there 50 000 years ago.
Moreover, you somewhat neglect the geological history of that region, Arnaud.
By the way, does archaelogy support your hypothesis at least indirectly?
Dear Arnaud, what is so "stalinist" (a term which is really insulting
in relation to the past of my country, but I am an easy-going person
;-), so...let us forget about it) in demanding more evidence? The
glimpses you have offered are simply not enough and I wonder who on
this forum thinks they are. You want to convince others, do you not? I
am not trying to discourage you, do not take it as a negative
approach. I just need something to be able to say: "Wow, this is it!"
All I can say at the moment is: "Hm, maybe..." and "Hm, why not?" at
best.
You know, what we need is a system of correspondences that is both
productive and predictive, which we do not have at present.
Interlocking jigsaw pieces, Yeniseic explaining PIE explaining
Yeniseic.
Cordially and supportingly,
Petr
P.S.: I have not forgotten about the conditioning factors, no worries!
;-) Yet still, you have not answered my question: If /t/ is not a
reflex of */s/ in Yeniseic, what is? And if all the /t/-cognates in
Pumpokol are Uralic, what is left? And if nothing is left, did
Proto-Yeniseic have */s/ at all? And if it did not have */s/, are all
those */s/-etyma loanwords from Uralic?
You know, claiming that the /t/-correspondence does not exist in the
native lexicon may have (de)vast(ing) implications.
P.P.S.: A link of interest:
http://www.esd.ornl.gov/projects/qen/lake.html