I said:
> The last two paradigms are not old enough to reconstruct back to the
> Thematicization period of mLIE. I've always said that AD is not
> ancient.
Wait! Hold the presses! While I pressed the send button, a sexy idea
had unexpectedly entered my head. I still stick behind my analysis of
PD and HD paradigms but I just thought of a way that an AD paradigm
could be ancient. This matters little to the problem of *-t&r- which
was assuredly proterodynamic and *-tér- which had to have been
logically hysterodynamic to begin with.
However, an AD paradigm could arise as predicted by QAR if we have the
following situation in MIE to postSyncope eLIE. Here, I'll use *a as
"any vowel" to demonstrate accent:
MIE eLIE
nom CáCaC-sa > CaCaC-sa
acc CaCáC-am > CaCáC-m
gen CaCaC-ása > CaCC-ás
Okay, what's happening here? Well, we see that if we have a disyllabic
athematic stem in MIE, we should expect initial accent in the nominative
because *-sa being just an agglutinated demonstrative doesn't affect
the accent of a stem as a general rule. We have accent on the second
syllable in the accusative (because *-am DOES steal accent because it's
a historically ancient suffix occuring in Tyrrhenian and Uralic inheirited
from Proto-Steppe in contrast). The accusative shows expected Paradigmatic
Resistance for the obvious reason of keeping the stem of the strong cases
the same form. The accent then is placed on the suffix in the genitive
because it's a disyllabic suffix and that's what disyllabic suffixes do
in MIE.
By way of QAR, all the above forms are possible since they display either
penultimate (acc, gen) or antepenultimate (nom) accentuation. Just
a little sidenote: the last "C" in the stem in this hypothetical stem
must be something other than a semivowel since *i- and *u-stems had
genitives in *-sa, not *-ása.
So this produces a wonky accent pattern in eLIE, one which I can now
see would be strongly alluring for any speaker to regularize. If the
accusative is put back in line with the nominative, we get:
nom CáCaC-s
acc CáCC-m (<- remember ablaut and zerograde?)
gen CaCC-ás
Hmmm.
So, while I don't think these thematicized animate suffixes were
originally anything other than PD when unaccented and HD when accented,
I can see some stems derived from them being accidentally sucked into
a coexisting AD pattern to produce the pattern that we see vis-a-vis
*-to:r in the nominative and *-tr-m in the accusative.
This I think might be a more optimal solution... but I just don't know
what ancient AD paradigms might have existed. Must think more on that.
An interesting idea for now anyway.
= gLeN