[tied] Re: Old Czech hpán vs. Modern Czech pan, pán

--- In cybalist@yahoogroups.com, Miguel Carrasquer <mcv@...> wrote:

> On Wed, 02 Jun 2004 09:25:40 +0000, tgpedersen
> <tgpedersen@...> wrote:
>
> >--- In cybalist@yahoogroups.com, Miguel Carrasquer <mcv@...>

wrote:

> >> On Tue, 01 Jun 2004 16:20:18 +0000, tgpedersen
> >> <tgpedersen@...> wrote:
> >>
> >> >Burrows: The Sanskrit Language p. 169
> >> >"
> >> >p-ati 'master'
> >> >
> >> >Footnote
> >> >From pa:- 'to protect, govern'. That the /t/ in this word is
> >suffixal
> >> >is evident from its absence inGk. <despoina>. Therefore <páti-
> is
> >to
> >> >pá- (<nr.pa-, etc) as <vr.káti-> to <vr.ka->.
> >> >"
> >>
> >> Burrows is wrong. There is no way *pot- can be derived from
> >> *poh3-.
> >>
> >>
> >
> >Not even zero grade *pH3ti- ?
>
> That would give Skt. *piti- (or *biti-).
>
> The only possibility might be a zero-grade *ph3-ot- (as in
> nr.pá- < *h2nr-ph3-ó-), if that hadn't given *bot-, but that
> would be a very peculiar root structure. On the other hand,
> it would explain the persistent o-grade in the word...
>
>

Yes, that's a problem, right? I was thinking of looking at it from
the semantic angle, starting all over from a meaning something
like "above/below" (whichever!) and get that to encompass *po-
"dominate", *p-d- "foot", *pa:- "let graze" (on the low lush banks
of a river). Of course normally there's no good path (there that
word was again) from /a/ to /o:/. The exception, curiosly enough, is
the Nordwestblock (ex. <hook>, German <Haken>, <cake>, Dutch
<koek>). That, incidentally, would also solve a problem I have
getting from *apa- to German Ufer, Dutch <oever>, which BTW involves
that pesky water word again. I'l have to find some way of
manipulating a long sequence of loans and developments.
And BTW, as Kuhn points out, many localised Celtic/Italic/Germanic
words (Gothic athn-, Latin annus comes to mind) have /a/ as stem
vowel.

Torsten