I said to Knut:
>>No, no. The correct PIE paradigm for *wlkWo- is:
>> *wl'kWom *wl'kWos

Whoops, I got mixed up, sorry. Yes, of course, the genitive is
*wlkWosyo (with *-o-, however). The nominative is *wlkWos.
Sometimes I get stupid ;) Piotr Gasiorowski has kindly presented
some IE paradigms, both for nouns and verbs, for quick reference


Anyways, *e is only used in the _vocative_ as a thematic vowel.
The rest of the cases show *-o-.

>My basic idea is simply that a group of nouns originally had the
>stress pattern.
>Acc: *welqw�m
>Gen: *welqw�se
>Dat: *welqwe`ie

First, I don't think the dative goes back very far in time. I don't
reconstruct it for Mid IE at all. My thought is that the nominative,
locative, vocative and dative cases were once all endingless and
identical. These cases only developed special endings in the
Late IE period. The dative and locative cases would have originally
used the nominative form plus postpositions (MIE *b�i "at, by" or
*d�i "in, within, amongst" > *-bhi/*-dhi) to express these ideas.

Second, it was the *loss of final vowels* which caused the change
of penultimate stress to mobile stress. This loss is the very
thing that eradicated the penultimate pattern in the first place!
You can't have the accent move to final syllable UNTIL the final
vowels are gone and the penultimate stress is obscured. For
instance, you place the accent on the final of the accusative,
but on the penultimate of the genitive... but there's no motivation
for this erratic pattern!

Finally, in re of the thematic vowel in *e, my view is that it
was always *o. What I notice, for instance, in Mid IE is the
addition of *-a- between consonant-final stems and consonant-initial
endings (ex: NOM *pat "foot" versus GEN *p�t:-�-s� "of the foot" >
*po:t/*ped�s). Without the euphonic vowel, the genitive would have
been **pats� (yielding Late IE **po:t). This inserted vowel
appears only in stems ending in consonants other than semivowels
(ex: *x�w�i/*x�w�i-s� > *xawis/*xweis) and so it seems like a
typical mechanism that one would see in a highly syllabic stage
of IE (and thus formed in Mid IE and not in Late IE with its
laid-back permittance for consonant clustering).

>Acc: *wlqw�m
>Gen: *wlqw�s
>Dat: *wlqw�i

Here, I can agree with you. The final vowels are gone and, now
that the penultimate stress pattern has been obscured, there
is room for analogical changes to the accentuation (and a move
of the accent to final syllable all in order to regularize the
paradigm). I'd rather reconstruct in Early Late IE:

Acc: *wlkW�m
Gen: *wlkW�syo
Dat: *wlkW�i

(BTW, please keep *kW, a labiovelar, and *kw, a consonant
cluster, distinct.)

>In a later period the o-grade appeared, that would leed to:

But why does this *o appear? From where? How? For what reason?

>For many thematic nouns the accent was then shifted to the initial
>syllable, but not for all nouns, leading to

Yes, but why "not for all nouns", Knut? You are proposing these
changes but not justifying why they occured.

>So: In short words. [...]

Okay, I thiiiiink I get it... maybe? Are you saying that IE
started out with a regular penultimate accent, then some nouns
were given a different accent pattern, then the vowels disappeared,
and then some *e's changed to *o (then, in the genitive case, the
*o changed back again!) for some hitherto unknown reason, then
the accent was regularized to the initial on _some_ nouns but you
don't know why only "some" and not "all"?

Is that what you're saying? Sounds like you've still got some
problems to work out. This only brings up even more questions
than it serves to answer so far. These changes have to be put
in a very precise order, otherwise it all simply unravels.

- love gLeN

Get your FREE download of MSN Explorer at http://explorer.msn.com/intl.asp