On Sat, 27 Sep 2003 14:11:52 +0000, Glen Gordon <
glengordon01@...>
wrote:
>
>Jens to Miguel:
>>The alternation /e:/ : /e/ does not point to /i/ by any rules for which I
>>know any evidence.
>
>Miguel:
>>I do.
>
>Two against one. I don't see it either. Your rules are convoluted and
>hard to follow/swallow here. They're unmotivated after a proper analysis
>of IE itself.
No, I believe they do follow from a proper analysis of the PIE phonemic
system. The near-absence of vowel phonemes /i/ and /u/ requires an
explanation. The solution is that they largely merged with the other short
vowel, /a/ at the time, as a short schwa. If we then analyze the nominal
Ablaut-patterns, we see that roots with **/a:/ give /o/ when stressed, but
/e/ in the weak forms (*wódr, *wédnos), where originally unstressed. Roots
with original **/u:/ and **/i:/ show a different pattern, with */o/ and
*/e:/, respectively, in the strong forms, but zero grade in the weak forms
(e.g. with **u: póntoHs, *pn.thós, with **i: k^é:rd, *k^r.dés).
>There's nothing wrong
>or terribly abnormal about *ye:kWr. It is a vrddhied root
It isn't vrddhied.
=======================
Miguel Carrasquer Vidal
mcv@...