Jens:
> You are in a way saying that all forms have penultimate accent if
> postposed demonstratives and plural markers are disregarded. However,
> that looks precisely wrong. In the more elaborate endings the accent
> plainly moves to the right, not to the left. Why are you saying the
> opposite?
Accent doesn't move to the left. It's a right-based accent placement.
What are you talking about? Can you provide an example?
Perhaps you mistakingly believe that a 3pp *-éna-ta shows left shifting
accent but that's not correct. The ending was originally *-éna with
penultimate accent. The ending *-ta was simply a postposed demonstrative
and not part of the word. When it became considered part of the word,
the word accent automatically became antepenultimate without any "left
shifting".
> There are plural markers in the dat.pl., ins.pl., and loc.pl. also.
> There are also dual markers contained in the corresponding endings of
> the dual.
Of course they are, which I mentioned were created _later_. I don't
understand why you think this is a fair arguement to list a whole bunch
of examples that appear to contradict what I'm saying without any deeper
analysis of the _structure_ that you're talking about. You're merely
quoting them ignorantly in isolation from each other. May I remind you
that you're discussing the origins of an entire system, in this case
the declensional paradigm, not individual morphemes.
The example of loc.pl. *pe:dsu would hint that the original form was
*pe:d (identical with the endingless singular) because the length
indicates a monosyllabic word. We've gone over how even the locative
singular in *-i may not be original and merely tacked onto a bare
locative. The dative as a case is the latest case to be created of all,
it seems, derived from the locative. This is a conclusion other published
individuals have noted long before my uneventful birth.
The etymologies of endings like *-bHis are quite transparent and so they
too must be late in origin. We see unbound *bHi too (hence "by"). It is
nothing but this locative particle with a plural in *-s by analogy with
the accusative plural *-m-s... In fact, we even see *-mis as a dialectal
variant, don't we!
The dual is even later than everything else. It's clear from any
examination of IE's declensional system as well as understanding universal
markedness, that the plural came first to weak cases, then dual forms
were synthesized for them. While I've agreed that the dual marker itself
is ancient and relatable to *k-duals in Uralic and EA, my acknowledgement
doesn't undermine the pattern I've always seen which is that number of any
kind was only marked in the strong cases of the animate in some relatively
recent stage of pre-IE.
> Hey man, the perfect was your point of departure.
It was an untested idea but I've decided that it was a deadend. No biggy.
= gLeN