xW>v>w (was: -tlo- vs -tro-)

--- In cybalist@yahoogroups.com, Piotr Gasiorowski <gpiotr@...> wrote:

>
> On 2007-07-08 00:26, stlatos wrote:

> > I suppose you believe that uXe > ue > uwe in all IE;

> Yes, naturally, whenever the laryngeal is lost /w/ is the most natural
> hiatus-filler in this environment. Why not?

Again:

> > where does
> > *dov- come from?

This change can be undone by analogy in most cases; it's the
exceptions I'm looking at. There's also analogy like *doxW+ / dow+ >
*doxW+ / doxWw+ (and *gYnYoxWw+ > know, *gYhloxWw+ > glow).

> > Why Skt rayis not *ra-is > *re:s?

> Because it's an i-stem: *reh1-í- > PIIr. *raHí- > *ra.í- > rayí-;

What part of that would prevent a-i > e:?

cf.

> *raHi-wánt- > *raiwánt- > revánt- 'rich'.

That could be from *rayiwánt- just as easily; in rayíntama- 'very
rich' it can't have i>0 because it's followed by 2 C (sonor.+stop).