Re: [tied] -phóros, -phorós, -fer

On 2006-04-10 14:42, tgpedersen wrote:

> That would be an athematic *-bhr.t. I proposed that they would
> survive too.

There are good reasons to believe that we are dealing with a
compositionally reduced -nt-participle. Not a true root noun, at any rate.

> No o-grade, you mean, or the argument is circular. But reflexes of
> *bhorós and *bhorós exist. For semantic reasons they would mostly
> have been constructed with an object, so I should write them instead
> as *-bhorós and *-bhóros.

This is partly true of *bHoró- (although the type is very frequent also
outside of composition), but not even partly true of *bHóros, whose
meaning is passive ('burden'). It may occur as a second member in
bahuvrihis (like, say, *gWr.h2ú-bHoros 'having a heavy burden'), but
then just about any noun can occur freely in this position. *bHóros is
_typically_ non-compositional.

> Since they are thematic, they are derived
> by generalizing an athematic root noun N *´-bh&r.&s, G *-bh&rós,
> decompounded N *bhór-&s, G *bh&rós. First an adjective was formed
> from it by generalising its genitive: N *-bh&r-ós, G *-bh&rós-yo,
> then the action noun was formed by generalising the nominative: N
> *bhór&s, G *bhór&s-yo. (And then somehow the remaining schwa's were
> turned into *o's by mutual influence, or whatever)
>
>
>
>> Besides, O-infixation
>
> o-grade

O-infixation us not the same as other manifestations of the o-grade. Go
back to Jens's long posting on his theory -- he mentions a few
diagnostic differences between them.

Piotr