From: Glen Gordon
Message: 7801
Date: 2001-07-04
Piotr:
>I wish I could fully understand it myself. It seems that occasionally we have *o/*e ablaut
>without a stress-shift, whatever the underlying mechanism (secondarily immobilised
>stress?). The *-s genitive is normally found with so-called "flexion fermée" in variably
>vocalised ("biform") stems (*dóru, Gen. dróus), but there seem to have been isolated
>CVC stems declined in the same way, i.e. Có:C(-s), Gen. CéC-s rather than C(e)C-ós
>-- or at least this is what the "auctores" suggest.
Thanks for the answer... but I goofed majorly. I forgot an important aspect that I discovered about IE inflexion. I should have said that the *-s genitive pops up on stems that were original consonant-final stems in Early IE (such as *xWawei-se > *xWueis "of the bird"), not just polysyllabic stems. Those with *-es/*-os are that way because they were once vowel-final (such as *kewane-se > *k^unos "of the dog"). So I guess, perhaps, we should analysize *k^unos more like *k^uno-s and not *k^un-os... or whatever.
So again,*doru is the way it is because it technically ends in a consonant and has always ended in a consonant (*drous < Mid IE *t:areu-se with accent on second syllable as opposed to the endingless nominative *t:aru with accent on first syllable).
So I guess what I'm saying is that *dems _is_ expected, well sorta. Afaik, the earlier form still would have been *t:am-se assuming it was originally a consonant-final stem (CVC)... but that should still yield Late IE **doms. Either it was once a CVCV-stem with shifting accent (and hence *o/*e ablaut + thematic genitive *-es/*-os) or it was once a CVC-stem (hence athematic genitive *-s and no ablaut because there is no shifting accent). It almost looks like *dems has been analogically changed via influence of *o/*e ablaut in polysyllabic stems like *po:t/*pedos (perhaps from earlier *pat:e/*pet:e-se).
BTW, it is *o/*e ablaut that should only occur in polysyllabic stems because the entire reason for its existence involves the "weakening" of unstressed Mid IE *a to schwa (written as *e here). I'm very sure this time :)
- gLeN