Re: On the ordering of some PIE rules

> First a look at the havoc my rule 1 caused in paradigms which were
> later set right by generalization (by my rule 6) in all languages
> except Germanic (trad.: caused by the first rule 3 in Iranian)
> Gatha-Avestan
> Ns pta: "father"
> As ptaram
> Ds fTrai
>
> this hopeless and foreign-learner unfriendly paradigm is corrected in
> Late Avestan
> Ns pita
> As pitr&m
> Ds piTre
> (but still Ap f&Dro:)
>
> but, however, in Ossetian
> (Abaev: A Grammatical Sketch of Ossetic)
> Ns f&d "father" Np f&dæltæ
> (cf
> Ns mad "mother" Np madæltæ,
> the normal pl suff is -tæ, no doubt from -tai)

Actually, the Gatha-Avestan Ds is not fTrai but *fTrai > *f&Dro:i, due
to one rule that makes fricatives voiced after fricatives, and another
which inserts an epenthetic (svarabhakti) vowel /&/ between the first
two of three consonants. The interesting thing is that the result is
that the result is like that of Verner. Maybe a reformulation would be
in place here.


Torsten