Re: gast

> > On the subject of PGmc. *gasti-
> >
> > Jorma Koivulehto:
> > Frühe Kontakte zwischen Uralisch iund Indogermanisch
> > im
> > nordwestindogermanischen Raum
> > '42. Finn. kansa 'Volk' (karel. kansa 'Gesellschaft, Kamerad'
> > usw.) = lp. guos'si 'Gast, Fremder' < frühurfinn. *kansa
> > <- (frühur)germ. *xansa: > germ. *hanso: > got., ahd. hansa,
> > aengl. ho:s 'Schar, Menge'. Das ostseefinnisch-lappische
> > Lautverhältnis erweist eine frühe Entlehnung. Germ. a hier
> > wahrscheinlich < idg. o (s. Seebold 355 s.v. Hanse). Siehe LÄGLOS
> > s.v. kansa.'
> >
> >
> > (Jouppe has for this entry:
> > 'Saami gaddže 'party, company, crowd, people''
> >
> > Which reminds me that gasti- has IE cognates only in Slavic
> > (gostI) and Italic (Lat. hostis), in other words it might be
> > substrate. But kansa etc is also substrate in Germanic:
> >
> http://tech.groups.yahoo.com/group/cybalist/message/57233
> >
> >
> http://tech.groups.yahoo.com/group/cybalist/message/55551
> >
> > So, assignment:
> >
> > What kind of scenario would permit both
> > xx substrate -> Finn kansa, Gmc. *xanso: / Saami
> > guos'si -> Gmc. *gasti-
> > etc?
> >
> >
> > Torsten
>
> Perhaps it's from an unrecorded Gaulish form e.g. g,
> /o/ > /a/

No, the 'reconstructed PIE' form is *ghosti-. That's sufficient to
generate the form in the three branches it's attested in. So a loan to
Germanic before PPGmc *o -> PGmc *a would do the trick. We should also
remember it's a social institution; the accepted foreigner. It would
be interesting if that goes to back the meeting between the pre-Saami
and NW IE. Something like *kant-ja > *kansa or the like?
http://tech.groups.yahoo.com/group/cybalist/message/55557
(Saami already has PSaamFin *a > PSaami *uo, I think).

> Anything similar in Gaulish names?

I don't know. Anyone?


Torsten