Re: Latinus geminus

--- In cybalist@yahoogroups.com, Alexandru Moeller <alxmoeller@...>
wrote:

>
> alexandru_mg3 schrieb:
>
> > > Latvian jumis (with short /u/) would be from *yeu-. There
> > > are other Baltic forms which *are* based on the root *yeuH-
> > > (as well as on the root *yeug-).
> > >
> >
> > I know that is with a short /u/ but as I know the root is yeuh-

'to

> > join , to mingle' with a laryngeal.
> >
> > I'm not aware about a 'similar' root *yeu- . If you know some

words

> > in Baltic or outside reflecting *yeu- please tell me.
> >
> > Marius
>
>
>
> seeing all this, it makes me to think that Rom. "juma( > jumatate,
> jumi-juma)", Alb. gysmë/gymë should belong to this root and not to

the

> *sem as Latin "semi" or Greek "emi" ( the Idea of Meyer Lübke that

Alb

> "gyumë" should be a loan from greek is to me anyway invalid).
> The semantic evlution can be explained as my pair= my another half

so

> the development from pair to half appers to be traceable. The

phonetic

> seems to work too but the "y" in Albanian speaks for a long "u"

here.

> Should have been there a laringial in the root here or another way

to

> get a long u: for having the "y" in Albanian?
>
>
> Alex

Romanian juma, jumate, jum&tate 'half' <-> Alb. gjym&s& / gjysm& it's
the reason for which I started the topic

PAlb/Dacian? *jumi-a:/*jumi-a-tya: [fem.] from a PIE *yu-mi-tyo
[masc.]

If we don't have a long /u/ in Alb gjym&s& the single reason for y-
is the influence of i of the next syllable

*jumi- > *gjumi- > gjymi

Lettish jumis seems indeed a direct cognate

Marius