--- In cybalist@yahoogroups.com, Bhrihskwobhloukstroy wrote:
>
> 2013/1/30, stlatos :
> >> > tetró:konta Dor G; cethorcho OI; vs
> >> > quadrá:ginta: L; kHaRasun = 40 Ar;
> >>
> >> This isn't before */r/ because there's */w/ between */t/ and */r/
> >>
> >
> >
> > So, if:
> >
> > quadrupe:s L; cátus.pat- S;
> >
> > exist it means quadru- < kWetwru-? I doubt it; met. ur>ru seems likely.
>
> *Bhr.: You should have then mentioned quadrupe:s, because there's no
> metathesis in quadra:ginta (not quadrá:ginta, by the way; the Romance
> accent's shift follows the palatalization of /g/).
I'm following Sihler's analysis. Sihler, who also mentions direct statements of old that the accent was trí:ginta: , makes it seem likely to me that tri:ginta: \ trí:ginta: , quadra:ginta \ quadrá:ginta , etc., both existed, prob. due to changes when saying the counting chant.
Whalen, I think you're on to something because the Romance forms seem to bear you out