[tied] Re: Some new etymologies

--- In cybalist@yahoogroups.com, "Abdullah Konushevci"
<akonushevci@...> wrote:

>
> On 6/8/06, tgpedersen <tgpedersen@...> wrote:
> >
> >
> > > To conclude, Albanian suffix –tar(e) is directly derived from

PIE *-

> > > tor-.
> >
> > Basque has a suffix -tar in place-of-origin compounds, and the
> > unrelated Iberian language might have had it too, according to

Trask.

> > Are you sure the -tar suffix is not of substrate origin?
> >
> > Torsten
> >
>
> ************
> I am sure it is from PIE origin and has its cognate in Latin -tor,

as well

> as in Greek and Hittite, for PIE *o > Alb. /a/

A-hem. That proves it might be of PIE origin, not that it is.

>and it has same function,
> forms nomina agentis in all these languages.

As you said, the "first form" ie. noun (not infinitive) based
compounds are the most common in Albanian. Since there's no verb
involved, they can't be nomina agentis.

>Further, in Basque it has different function

Semantically, there's not a long way from "one connected with a
place" to "one connected with an activity". That transition is
possible. Further, the PIE *-ter appears with preverbs/-positions
too, and arguably, in the 3rd passive and the Germanic words
for "there". That sort of "scattered" activity is not typical of
a "native" morpheme, so I suspect it's a loan in Indo-European.

>and about Iberian language I have no clue.

Guess you'll have to take Trask's word for it, like I did, then.


Torsten