Re: Gemination in Celtic

----- Original Message -----
From: Anders R. Joergensen

Well, I guess I'll have to disagree with you on that. I also find the
evidence for *kwh2C- > Latin caC- shaky, to say the least. Lat.
canis, which you gave earlier as an example of this, is usually (for
good reasons) reconstructed without h2.
======
I disagree too
Let's say this word does not count.
Arnaud
> ==========
> >
> > bukk- "male"
> > Tsigane, avestique buz < *bu-g-
>
> But we can't rule out a loan from Germanic for this word, can we?
>
> What would it be in Germanic ?

ON bukkr/bokkr, OHG boc, ModG Bock, OEng. bucca, etc.
< PGerm. *bukka- and poss. *bukkan-, taken to reflect an n-stem
*bHug^-on- ~ *bHug^-n- by Schumacher, as I mentioned earlier.

=======
We can't rule out that Kluge'Law applies to Celtic too.
So maybe this word counts for nil.
Arnaud
==========
> >
> > kass- "hate"
> > ka:dos
> > Av sadra
>
> I'm not sure I get it anymore. What does your geminate *-tt- (< h2t)
> become in the attested Celtic languages?
>
> I meant
> keH2- s- > kass in Celtic (gemination applies to s too)
> keH2- t- > kad in Eastern PIE

So if your *kWotta: was attested in later Celtic languages, I'd be as
*kWossa:, right?

Anyway, MBret. queuz /køD/ `regret', ModBret. keuz seems to be a
problem, as it is usually derived from PCelt. *ka:d- (< PIE *k^ah2d-
= Gr. ke:dos).

And PCelt. *katsi- (or the like) may just be from *kad-s- or *kad-t-
from PIE *k^&2d-.

>
> Ok
> how do we tell -kk- and -gg- in Celtic ?
> Arnaud

*kk > OIr. <cc, c>, ModIr. <c>, Brit. *x (W ch, Bret. c'h)
*gg > OIr. <c> (usually), ModIr. <g>, Brit. *g (WBret. g)

Examples are of course harder to come by than with simple *k and *g

=======
What are examples for that ?
Arnaud
=========
> >
> > lu?k "hiccup"
> > Gaelic aileag
> > Greek lug-mos
>
Yes, and as Brian pointed out, it appears to be a derivative in -eag.
So it should be left out of the discussion.

=======
What about slugim "swallow < *slukk ?
Arnaud
======
>
> >
> > makk "eat"
> > smag "taste"
>
> What Celtic words are you thinking of?
>
> French mâcher < *makk-
> Arnaud

Isn't this just from Lat. mastica:re? (and the <^> in French usually
implies the existence of a lost <s>, confirmed by the Old(?) French
maschier)

========
No
Northern French je mak- "I eat"
impossible from masticare.
Arnaud
==========

>
> >
> > mrekk- "speckled"
> > Lituanian margas
>
> What about this example ?
> Arnaud

What is the evidence for a laryngeal in these words?
Anders
=========
Precisely
the alternation -kk- ~ g.

Arnaud
=========

Do you have a list of all proto-Celtic forms
with -gg- and -kk- ?

Arnaud
=========