First, I'd like to correct myself on something that
I said to Miguel. The past tense cannot have zero-grade
because it was formed at a late date when IE was
already fracturing. Zero-grading, on the other hand,
initially formed during the transition from the Mid IE
to the Late IE period (which I estimate to be around
5000 BCE). Thus, the past tense must have operated via
the processes that existed in the latest stage of Late
IE and guna should be the normal grade form (and in
fact it is). Thus *e-bher-t, not **e-bhr-t.
>The point is that if Ablaut is a *sound-law* (or rather a collection of
>sound-laws), it cannot care
>whether it's working on verbs or nouns.
But the resistance towards loss of vowel in the root
DOES occur in both nouns and verbs! Listen...
The process of Mid IE syncope operated on all stems
but DID NOT operate within a paradigm (at least, not
normally) whether that be conjugation or declension.
Thus, regardless of whether *pat:- (the Mid IE version
of *pod- "foot") was unaccented or accented in various
cases, its default form (the nominative case) was
accented.
The Mid IE genitive form *pet:-ase, operating under
the laws of Vocalic Constraint and Penultimate
Accentuation, contained schwa in the initial syllable
(transcribed here as "e"). Normally, the unstressed
schwas of the genitive case should have eventually
disappeared, causing zero-graded **pd-os in Late IE.
However, again, I stress that the process of syncope
DID NOT NORMALLY OPERATE WITHIN A PARIDIGM because
this would obscure the root. As a result, the
unstressed schwa in the root was strengthened
(ie "fronted") to *e as though it were stressed to
avoid its loss. These are the processes that caused
the strange *o/*e ablaut seen not only in noun stems
like *pod- but also in the paradigms of some *o-graded
_verb_ stems which also resist syncope
interparadigmatically.
A verb like *bher- also resisted syncope throughout
the paradigm.
However, if we were to take a noun stem like MIE
*kewane "dog" (> *kwon-), the first schwa was lost
because the first syllable was NEVER accented, even
in other declensional forms. Thus, syncope here was
not resisted because this did nothing to obscure the
root.
There are of course some verb stems like *?es- "to be"
that haven't even resisted Mid IE syncope _within_
their own paradigms (*?es-mi versus *?s-mes) but
these roots and their conjugated forms were used
often enough that the fears of root obscurity were not
a sufficient enough reason to resist the mundane
process. In fact, this is the very uneventful reason
why "to be" is such an irregular verb in so many other
languages.
- gLeN
_________________________________________________________________
Chat with friends online, try MSN Messenger:
http://messenger.msn.com