Digest Number 395 (concerning Japanese)

-----------------------------------------------------------------
(Anglophones may scroll down to the bottom, which is where you
filthy cromags belong, for the English translation... :P)
-----------------------------------------------------------------

Guillaume:

>En fait, le mot chinois medieval est tshit, pas *shit,

Oui, mais le reconstruction de Miguel m'a donne la chance de m'amuser un peu
:)

Guillaume:

> 1 fitotu Non c'est pyit�- en vieux japonais

Ay caramba! C'est possible que je me suis confuse, pardon. Je crois que t'as
raison.

J'ai dit:

>As for the other set of Japanese numerals, they are certainly native,
> >based on a binary opposition scheme using vowel harmony (eg: OJap
> >fitotu/futatu, mitu/mutu, yotu/yatu, etc). Here they are in Old
> >Japanese, afair:

La reponse de Guillaume:

>En fait, il serait faux de dire que la majorite des specialistes du
>japonais acceptent l'hypothese d'un systeme binaire, [...]

Mais voyons, donc! Selon tes opinions, ces paires seraient une coincidence
extremement enorme, une impossibilite statistique! Meme si les "classes
accentuelles" sont differentes, on doit accepte que l'harmonie vocalique
existe quand meme parce qu'il y a peu de langues qui ont des nombres avec
ses characteristiques specifiques.

Guillaume:

>4/ Les numeraux japonais ont de bonnes correspondances avec les >numeraux
>Tunguso-mantchous. Je doute fort qu'ils soient herite d'une >proto-langue,
>mais il est tres probable qu'ils aient ete emprunte en >japonais. Voir le
>site de Starostin. La comparaison avec les autres >langues 'altaique' est
>plus problematique.

Je suis d'accord que le japonais a des bonnes correspondances avec les
langues tunguso-mantchous. Mais, on peut voir clairement que seulement _un_
nombre de chaque paire peut etre compare avec le tunguso-mantchous:

muttu "six" (et non mitu "trois")
y�tu "quatre" (et non yatu "huit")

C'est un autre preuve que ses nombres sont binaires.

Je refuse de voir le site de Starostin concernant l'Altaique (J'en ai
discute dans le passe) car c'est trop apparent qu'il n'en sais rien du tout.
Ses reconstructions sont simplement affreuses (sans une phonologie stricte
et definee) et ses reconstructions de son "Caucasien du Nord" sont
improbables.

Guillaume:

>Le proto japonais est :
>1 *pit@-
>2 *puta
>3 *mi-
>4 *d@-
>5 *itu-
>6 *mu-
>7 *nana-
>8 *da-
>9 *k@...@n@
>10 *t@...@

C'est reconstructions semblent propre. Merci.

----------------------------------------------------------------
TRANSLATION FOR THE UNILINGUAL DUMMIES WHO SHOULD SHOOT
THEMSELVES IN THE FOOT FOR NEVER LEARNING FRENCH :)
----------------------------------------------------------------

Guillaume

>In fact, the medieval chinese word is tshit, not *shit,

Yes, but Miguel's reconstruction gave me a chance to amuse myself a little
:)

Guillaume:

> 1 fitotu No, it's pyit�- in old japanese

Ay caramba! It's possible that I got confused, mea culpa. I think you're
right.

I said:

>As for the other set of Japanese numerals, they are certainly native,
> >based on a binary opposition scheme using vowel harmony (eg: OJap
> >fitotu/futatu, mitu/mutu, yotu/yatu, etc). Here they are in Old
> >Japanese, afair:

The response from Guillaume:

>In fact, it would be false to say that the majority of japanese specialists
>accept the hypothesis of a binary system, [...]

Let's get real! According to your opinions, these pairs would be an
extremely large coincidence, a statistic impossibility! Even if the
"accentual classes" are different, one has to accept that the vocalic
harmony exists regardless because there are few languages which have numbers
with these specific characteristics.

Guillaume:

>4/ The japanese numbers have good correspondances with the >Tunguso-Manchus
>numbers. I doubt strongle that they would be >inherited from a
>proto-language, but it is very probable that they were loaned into
>japanese. Look at Starostin's site. The >comparison with the other 'altaic'
>languages is more problematic.

I agree that japanese has good correspondances with the Tunguso-Manchus
languages. But, we can plainly see that only _one_ number of each pair can
be compared with Tunguso-Manchus.

muttu "six" (et non mitu "three")
y�tu "four" (et non yatu "eight")

It's more proof that the numbers are binary.

I refuse to look at Starostin's site concerning Altaic (I talked about this
before) because it's too apparent that he knows nothing about it at all. His
reconstruction is simply hideous (without a strict and defined phonology)
and his "North Caucasian" reconstructions are improbable.

Guillaume:

>The proto-japanese version is:
>1 *pit@-
>2 *puta
>3 *mi-
>4 *d@-
>5 *itu-
>6 *mu-
>7 *nana-
>8 *da-
>9 *k@...@n@
>10 *t@...@

These reconstructions seem sound. Thanx.

- gLeN

_________________________________________________________________________
Get Your Private, Free E-mail from MSN Hotmail at http://www.hotmail.com