From: fournet.arnaud
Message: 57834
Date: 2008-04-22
----- Original Message -----
From: "alexandru_mg3" <alexandru_mg3@...>
>
>I. The Romanian word is ujag& 'bottle, (glass)'
>II. The word is considered a loan from Hung. uveg 'id.'
>III. But I 'cannot stop' myself not to link Romanian
>uiag~a /ujag&/ 'bottle, (glass)' with Albanian /uj&/ 'water'.
>IV. Also, I cannot see also how an Ossetic /a/ (avg < *a(:)paka:) can
>be loaned in Hungarian as /ü/
>Also we see that Hung.ü correspond to Ossetic u,i (that is reasonable)
>----------------------------------------------------------------------
>Osset. bud "smell" - Hung. büz "stench";
>Osset. ävzist "silver" - Hung. ezüst "silver";
>So an Ossetic a for a Hung ü is completly irregular
----------------------------------------------------
>Osset. avg "glass" - Hung. üveg "glass";
>Also I couldn't find a Hungarian /ü/ loaned from an /a/ of any other
>language ...
============
I'll come back with my pet "hedgehog" word.
Hungarian is sün
and Cheremis is süri
Khant soule
(of course impossible to get out of **sijili)
The alternation e/o in IE data exists also in Uralic :
Hungarian, Cheremis, Khant and also Mordvin must have *u
Cf. Armenien ozni
Finnish can have *e
Indeed ü < a in Hungarian looks hard to get,
üveg < *uvink/*ovink looks possible.
Arnaud
==========
> Rom. ujag& > OldHung. uj-wa-ga [uj>ü, w>v, a>e] > Hung. ü-ve-g
>Marius
=======
The sequence uj "new" exists in Hungarian and seems stable.
Arnaud
==========