Re: [tied] IIr 2nd Palatalisation (was: PIE voiceless aspirates)

From: Patrick Ryan
Message: 41990
Date: 2005-11-10

----- Original Message -----
From: "Miguel Carrasquer" <mcv@...>
To: <cybalist@yahoogroups.com>
Sent: Wednesday, November 09, 2005 1:30 AM
Subject: Re: [tied] IIr 2nd Palatalisation (was: PIE voiceless aspirates)


> On Tue, 08 Nov 2005 20:02:57 -0600, Patrick Ryan
> <proto-language@...> wrote:

<snip>

> >PIE *kw regularly produces OI <c>;
>
> No. It regularly produces OI <k>. Before front vowels and
> /j/ this was palatalized to <c>.

***
Patrick:

First, to keep this discussion within manageable proportions, let us, for
the moment at least, look only at _initials_.

The only forms I find in Pokorny with OI <k> for PIE *kW are:

ka:çate:, *kWek^-, *kWok^[?];
kula-m, *kWel-;
karo:ti, *kWer-;
ka-H, kWo-s;
ki-H-, *kWei-s, *kWoi-s[?];
kri:Na:ti,*kWrei-;
kRmi-, *kWrmi-;
kSap, *kWsep-;
ka:s-, *kWa:s-.

Presumably, you would explain these as instances of:

zero-grade: kula-m; kri:Na:ti; kRmi-; kSap
presumably *o-grade: ka:çate, ka-H, ki-H
before *a: (*aH) - ka:s-

When I see imperative kuru, I suspect that karo:ti may be the result of
zero-grade as well: *kWréuti > *k(u)róti > *k-a-róti

Is that basically how you see it?


********************

> >PIE *k produces OI <k>;
>
> Which was palatalized to <c> before front vowels and /j/.

I have gone through the *k section in Pokorny for initials, and can find
only

candati, *(s)kendeti (metathesized from *(s)ked-n-)
camara, *kem-

showing, purportedly, OI <c> for PIE *k.

That seems to be a rather fragile basis for such a sweeping "law".

camara is suspect because it is an expressive, probably imitative word.

And we know the connection between PIE *sk(^) and OI <ch>.


***********************


> >PIE *k^ produces OI <S>.
>
> Yes. Assuming you mean <s'> (a.k.a <ç>).

***
Patrick:

Of course. I am used to writing <S> for <s> (from the time that esh could
not easily be inserted into email) so used <S> for Pokorny's <s'> and
Whitney's <ç>. If this sound is, in origin, truly a palatal sibilant, we
should expect PIE *se/*sy to be reflected in Old Indian as <S> but, no luck,
there. In view of that and its use in Old Indian to reflect PIE *k^, I
suspect that it was originally /ç/ rather than /s/, and am adopting the
practice of writing it <ç>.

***