*ANTE*-penultimate accent in Mid IE?

I just had a major brainstorm today about my proposed
Mid IE accentuation. For the longest time, as you
all know, I've posited a strict penultimate accent for
this stage. I've found little reason to undo this theory.
It is too precious to reject given that it solves the
most basic yet unintuitive accent alternations like
*?es-t versus *?s-ent or *kwon-m versus *kun-os. Clearly
penultimate accent plays a big part here and can't be
totally off-track.

However, since my mind can't focus on every aspect
of IE and Mid IE grammar all at once, I naturally
pick apart different aspects of my theory bit by bit,
causing inevitable chain-reactions that destroy other
previously held notions. Aaah, 'tis the joyous,
masochistic process of learning that whips my tethered
soul into flights of intellectual ecstasy.

Well, long story short: Mid IE isn't quite "penultimate".
This is because I was thinking about a paradox that just
won't go away. If we work back the perfect conjugation
to Mid IE, and we use a root like *woid- "know", we
have a teensy bit of trouble coming up with a credible
and aesthetically pleasing system that accounts for
later IE. I mean, if we use strict penultimate accent,
we might reconstruct something rather absurd like:

*wait:á-xa
*wait:á-ta
*wáit:a

As you can see, the accent wanders and it predicts a
later thematic vowel that doesn't exist in Reconstructed
IE. The Penultimate Accentuation Rule if held too firmly
doesn't allow us to reconstruct something less ugly for
this example. Oh-oh. Major blech!

Then I comes to thinkin'... What if I'm partially right.
What if the accent on _words_ isn't strictly penultimate
but rather on _stems_ or _roots_. What we end up with is
something way cooler than what I initially devised and
something that can now account for later IE even better
than before without divorcing it totally from the initial
accent of Old IE I've theorized based on Tyrrhenian
connections either.

-------------------------------------------------------

My new theory will be as follows. All roots or stems are
still strictly penultimate in MIE. Words as a whole,
however had two choices: penultimate accent or
ante-penultimate accent. Confused? It's pretty simple,
really. Here's an illustration:

You have the same root *wait:a-. This root will have
had accent on the first syllable (making it penultimately
accented like we want). Now, the singular MIE paradigm of
*wait:a- then was:

*wáit:a-ha (antepenultimate)
*wáit:a-ta (antepenultimate)
*wáit:a (penultimate)

Notice that this now doesn't violate the new rules on
accent, since it doesn't show wonky accent alternations
or syllabics. The word as a whole is either
antepenultimately or penultimately accented and the stem
itself is ALWAYS accented on the penultimate as we state.
AND, most importantly, it now predicts properly, using
the previously deduced rules of Syncope and Suffix
Resistance, the later system. In early Late IE, the
paradigm can now be reconstructed nicely as:

*wáid-x&
*wáid-t&
*wáid-&

And this regularly becomes what we indeed find in
Reconstructed IE:

*wóid-xe
*wóid-txe
*wóid-e

(Btw, the change of eLIE *-t& > IE *-txe is a small
matter of analogical assimilation with the first
person.)

This new rule tells us a lot more, though. In the
aorist, the conjugation of MIE *k:en?a- "give birth"
(> *genh-) is instructive of the resulting nuance
of this new "quasi-penultimate" theory:

k:én?a-m k:en?a-ména
k:én?a-s k:en?a-téna
k:én?a k:en?-éna

Here we see that polysyllabic suffixes automatically
steal accent from the root or stem and that this
rule is good not only for verbal paradigms of all
sorts of roots, but also for MIE's declensional
paradigms for nominal stems and roots:

nom kawána-sa
acc kawána-m
gen kawan-ása

As you may notice, the nominative in *-sa can now be
_suffixed_ to the penultimately accented stem *kawána-
without problems (implying that the nominative suffix
was attached to the stem much earlier than I previously
thought, possibly in Old IE). Only the genitive here
steals accent because it is the only case in this
example with a polysyllabic suffix. This properly
yields what we later find, given my aforementioned
rules of the Late IE period:

nom kwó:n-s
acc kwón-m
gen kun-ós

So, I hope everybody got that. Basically, all I'm
saying is that MIE is now to be understood as
"quasi-penultimate", still being mostly penultimate
but allowing the accent of some words to be
ANTEpenultimate.

Thoughts on Old IE
------------------
As I mentioned, my theory so far gives Old IE a
completely regular accent on the initial. I presume
that the shift of the placement of accent was one
involving the transferal of primary stress from the
initial syllable to wherever the secondary stress
was, which may have often found itself on the
penultimate or antepenultimate.

This earlier initial stress is the last method with
which we can completely explain away the source of IE's
otherwise unpredictable accent. Further, it has
implications in connection with Tyrrhenian, a family
of languages that many linguists think is particularly
close to IE.

Based on syncope in Etruscan, it appears that the
accent was largely on the first syllable. I take this
as indicative of Tyrrhenian as a whole where I've so
far come up with a quasi-initial stress system where
the only times that stress isn't on the initial is
when the first syllable contains *a. I recall Piotr
mentioning a similar situation with accent avoidance
involving "yers" in BaltoSlavic.


= gLeN