Proper methodology (was: RE: [tied] Re: Mother of all IE languages)

> >>The use of a verbal or nominal root is unheard in the personal

pronouns.

> >
> >That's not a reasonable expectation. Not all pronouns may be

replaced by

> >"verbal/nominal roots" in any particular language. However, this

same

> >process _does_ occur in other languages. The immediate example in

> >mind is Japanese
>
> Japanese? You know that's not a valid comparison.

Does this mean that Japanese once had pronouns, which were replaced?

> >>If the root were *yeu-, a root noun derived from it should be

*yut-

>
> >>Actually, the root is in my opinion *yeuh1-,
> >
> >It's an "opinion", not fact per se. Since it's safe to say that

*yeu-g-

> >should be grouped with the above root, its existence negates your
> >theory.
>
> What counts is the facts. All apparent cases of *yeu- can be

explained as

> *yeuh1- (Skt. yá:u-ti (*yeuH-ti), yuváti (*yuHé-ti), yu:thá-, yú:na-

. Lith.

> jáutis).
>
> How to interpret the facts is another matter. I think the

alternation

> *yeuh1- ~ *yeug- follows naturally from the uvular hypothesis. If

*yeug-

> was originally *yeuG-, with voiced/glottalized uvular stop /G"/,

such a

> sound has a natural tendency to become fricative (cf. the Kartvelian
> cognate/borrowing *uG-el- "yoke", with velar fricative /G/ < uvular

stop

> /G"/).
>
> Speaking of Caucasian cognates/borrowings, the NE Caucasian word

for "yoke"

> is also interesting: *ruk.(k.)- (Lak ruk., Nakh duq., Dargva *duk.,
> Avar-Andi *ruL.L.V). Compare Armenian <luc> (*lug-). It's

possible that

> Armenian reflects a proto-form *liug-óm, and that we have a doublet

*liúG-

> > *yeug-/*yeuh1- vs. *líuG- > leig- "binden" (Lat. ligare, Hitt.

link-,

> etc.).
>