Re: [tied] IE numbers

On Wed, 10 Oct 2001 22:14:05, "Glen Gordon" <glengordon01@...>
wrote:

>Miguel:
>>So what about *kwete-sor- (*kweto-sr-)?
>
>Is this an acceptable reconstruction? Is it appropriate to
>reconstruct a Proto-IndoEuropean root based on a Celtic-Anatolian
>isogloss?

Celtic=Indo-Iranian, actually.

>The problem with reconstructing *-sor- also involves the fact that
>IndoEuropean (that is, IndoAnatolian) probably had two genders
>(animate/inanimate) rather than the later three(masculine/feminine/neuter)
>and so we have to wonder why *kWetwores
>is so special to have adopted a special feminine suffix within the
>sphere of an animate-inanimate language.

The same happened in the case of "3". In fact, the use of *<-sor>
"woman" as *lexical* feminine marker is precisely the kind of thing we
would expect in an animate-inanimate (or completely genderless)
language (cf. the use of -eme "woman" and -ar "male" in genderless
Basque in words such as <asteme> (<asto> + <eme>) "female donkey" or
<katar> "tomcat" (<katu> + <ar>).