Re: Magyar uveg <-> Romanian uiag&

From: fournet.arnaud
Message: 57886
Date: 2008-04-23

>>So before to learn that rules, you are already sure about your final
>>conclusion: you will find 'a rule' that makes an /u/ from an /a/ (I
>>saw that you have hardly try alreday) when there are other words (see
>>my quotes) that show you something else?

>Of course. We don't even need to know how a in avg (if it stems
>from avg at all: we don't know!) gets umlaut-u in Hungarian. It
>suffices to see that uveg > uiaga fits some patterns, whereas
>the vice-versa thing doesn't work (as your Albanian and noian
>don't either).

>>The earlier Ossetic variant was *apaka: ==>
>>try now with this one... to obtain uveg :)
>>I can tell you that is preferable to remain with avg

>I know. But uiaga is no substrate word such as those listed in various
>lists by all linguists (Romanian and foreign) that dealt with them. I
>don't know
>how old uiaga is, but I expect it to be very recent. More recent than 1600
>or 1700. (In Hungarian it might be much older, but this is none of our
>business here.)


Another possibility that may reconcile all forms is to start with *uGag
whatever it might have been before.
Hungarian *uGag > uYag > ΓΌ_eg and _v_ is preferred to -j- as glide
Rumanian *uGag > uYag > ujag (direct)

Of course *avg (and worse apaka:) is dead as an hypothesis in that case.