[tied] Re: Schleicher's Tale

--- In cybalist@yahoogroups.com, Miguel Carrasquer <mcv@...> wrote:

> On Sun, 27 Jul 2003 04:08:51 +0000, Rob <magwich78@...> wrote:
>
> It's what Adams reconstructs in EIEC. In fact, both *h2ówis and

*h3ówis (<

> *h3éwis) are possible. There are some forms pointing to *a- (TochB

pl.

> awi, Skt. avi- [not *a:vi-]), which are easier to explain as coming

from

> the oblique root of *h2ów-i-: *h2éw-i- (= *h2áw-i-), but *h3éw-i-,

obl.

> *h3w-éi- [or perhaps *h3ów-i-, obl. *h3éw-i- (= *h3ów-i-)] cannot

entirely

> be excluded.

I think I remember reading somewhere that PIE /o/ > Tocharian /a/.
But I could be mistaken.

> *h1e- > e-
> *h2e- > a-
> *h3e- > o-
>
> but:
>
> *h1o- > o- *h1e:- > e:-
> *h2o- > o- *h2e:- > e:-
> *h3o- > o- *h3e:- > e:-

I was under the impression that Pre-PIE /xWa/ > /o/ everywhere,
and /xa/ > /a/ everywhere; in other words, that they were immune to
e/o variations. Is this wrong?

Digression: Hittite preserves two laryngeals, which Piotr Gasiorowski
says were velar ('h') and uvular ('hh') fricatives, respectively.
Could it be that Hittite 'h' /x/ < 'h2' /x/, and 'hh' /Q/ < 'h3' /xW/?

> Well, I think *k^é:rd (Arm. sirt).

Ah, so you think the long /e/ was 'original' and not due to
compensatory lengthening by omission of the word-final stop?

- Rob