I'm still wondering whether it's possible that the
stems that _seem_ to show *a rather than expected *e
when neighbouring the "marked" series (*q, *G, *GH,
*h2) are originally long.
So far, I can't convincingly rewrite traditionally
written *legH- as *leGH- without explaining the
odd e-vocalism in place of expected colouration. But
I so enjoy regularity and I've been in shambles trying
to find a way to restore order once again. Then I
gots to thinkin'... :)
Here's an idea. You have some verbs that have
special inherent semantic qualities that would later
affect how they surface in a new durative-aorist
system. Some verbs would be naturally durative,
and some aorist, of course. However, there could be
other features that could further split up the
array of verbs such that a special subset of verbs
would evolve from lengthened roots (inherently
aorist ones?) to the shorter fullgrade.
Since this shift from an undifferentiated "mi-
conjugation" to a durative-aorist contrast would
happen just _after_ the start of IE's fragmentation,
this hypothetical class might have evolved from
*Ce:C- to *CeC- without being coloured by any of
the marked series... because the marked series no
longer existed at that point! Sound nifty? :)
So, an original *le:GH- would fail to show colouration
because the vowel is long. After IE starts to
fragment, it changes to *le:gH- when the marked
series merges with the plain in most of the dialects.
Then the lengthening is finally done away with in
order to confuse later IEists. Hence *legH-.
I just can't quite figure out what semantic class
these verbs might all share. Anybody have input?
= gLeN
__________________________________
Yahoo! Mail - PC Magazine Editors' Choice 2005
http://mail.yahoo.com