Re: [tied] -phóros, -phorós, -fer

> My idea is that the ablaut vowel was proto-proto-IE *a, which ->

*e

> when stressed, -> *& when unstressed. So far, so standard.

However,

> if stress after this stage was moved and landed on a *&, it turned
> into *o. *o's are thus the ablaut vowel (which was *a) in a re-
> stressed stage.
> For all three vowels
> PPIE *a -> PIE *e (stressed), *& (unstressed), *o (restressed)
> PPIE *i -> PIE *ey (stressed), *i (unstressed), *ey (restressed)
> PPIE *u -> PIE *ow (stressed), *u (unstressed), *ow (restressed)
>
> in some languages generalised (under Semitic 'Atlantic'

influence?)

> from the *a reflexes to
>
> PPIE *a -> PIE *e (stressed), *& (unstressed), *o (restressed)
> PPIE *i -> PIE *ey (stressed), *i (unstressed), *oy (restressed)
> PPIE *u -> PIE *ew (stressed), *u (unstressed), *ow (restressed)
>
> (not Italic, it has no *ew)
>

Note how it accounts for
*méntis, *mentéys
*pértus, *pertóws/*pertéws

and if I add
PPIE *eR -> PIE *eR (stressed), *R (unstressed), *eR (restressed)
generalised in some languages to
PPIE *eR -> PIE *eR (stressed), *R (unstressed), *oR (restressed)

that should take care of the other semivowels, without special
rules.


Torsten