The ante-PIE state of things (I mean the
"origin" of the PIE vowel system) is a a matter of speculation and debate. Not
so the Indo-European quality of *e. The vowel is normally reflected as mid-low
to mid-high front /e/ in Old Hittite, Greek, Italic, Celtic, Germanic (with a
partial shift to /i/), Slavic, Baltic and Armenian. It causes the palatalisation
of various consonants in several branches and individual languages. Tocharian
and Albanian show complicated developments, but these are strictly local
affairs, and the reflexes in question point to a mid front vowel as well
(consonants are palatalised before it in both branches) *e. In Indo-Iranian the
reflex is /a/, but again velars are palatalised before it. Being a former
schwa-like vowel is not enough to cause palatalisation in IIr at least, where
velars are _not_ palatalised before any /i/ that reflects syllabic laryngeals or
epenthetic vowels.
To sum up, the comparative evidence for *e
= /e/ is absolutely overwhelming. Only someone with an urgent
non-linguistic agenda could have a reason to ignore it. The lapse of time
between the disintegration of PIE and the Indo-Iranian merger can only be
guessed at, but must have been pretty long. My private estimate is ca. 2500
years or slightly more. There is no reason to suppose that during that period
the vowel was anything else but /e/-like (middish front).
Piotr
----- Original Message -----
Sent: Tuesday, January 15, 2002 3:55 PM
Subject: [tied] [pieml] Re: IE: likely home, India
*/e/? So I-I has */a/ > */e/ > */a/ and */a:/ > */o/
> */a/ ? BTW,
isn't <&> in itself front enough to cause
palatalization (the fact of
which I haven't
disputed)?