From: Glen Gordon
Message: 7810
Date: 2001-07-05
>I am tempted to suppose that:- The original vowel was "e".
>It (sometimes or always) changed very early to "o" after the stress,
Proof? Evidence? Anything? I disagree that *e and *o are terribly connected at all. This is a forced assumption.
>Unstressed vowels often dropped, causing zero-grading.
Actually, from what I see, there wasn't much strong stress or zero-grade happening for many thousands of years before IE.
>Where analogy tried to force zero-grading in a place where the result >would be unpronounceable, assorted sorts of helping-vowels got in. >The H2 laryngeal tended to change nearby "e" to "a". >The H3 laryngeal tended to change nearby "e" and "a" to "o".
Actually, Piotr might be correct in thinking that *x (*h2) and *xW (*h3) merged early (before Anatolian) such that the laryngeals had already coloured *e to *a (that is, AFTER Mid IE *a had already become *o). Once the two laryngeals started the colouring, *xW went further by rounding the new *a to *o and then at that point the distinction between *x and *xW was lost producing only *x. Anatolian then splits and the rest is history... literally!
>To use a metaphor, to those trying to reconstruct old forms of >languages, it is to be regretted when the bulldozer of Analogical >Levelling drags its dirty iron tracks and blade through a language's >internal linguistic "archaeology", destroying evidence, and fogs out >phonetic laws with the blastings of its dirty outsized exhaust pipe,
>but it happens.
Oh please! That's just a lazy (or uncurious) person's excuse to not try. Analogical changes (which there are many in pre-IE) have to be slowly peeled away like the skins of an onion. It's a very slow process but certainly not impossible like you paint it to be. Ideally, the various theories on theoretical analogical processes in Pre-IE must be, when taken as a whole, the most efficient package of solutions available.
- gLeN