Re: [tied] PIE *akWa: 'water' (was: The role of analogy, alliterati

--- In cybalist@yahoogroups.com, "alex" <alxmoeller@...> wrote:

> Richard Wordingham wrote:
> > 1) Alex suggested that it did not meet the CVC root pattern, so I
> > supplied the initial laryngeal to make the point that it is not
> > inconsistent with the pattern.
>
> OK I understand the teoretic way but the "h2" cannot be seen as

consonatic

> or can it?
>
> >
> > 2) kW as opposed to k^w is supported by three facts:
>
> Can one agree the PIE *kW was as actually italian "qu"

in "aqua"?

No. In part see below. The choice seems to be between a labialised
velar ([k_w] in X-SAMPA) and a *co-articulated* labiovelar ([k_p] in
X-SAMPA). Neither is the _cluster_ [kw].

> Can
> one agree the PIE k^w was as in english "chew" ?

No. A better choice is [kw], say as <qu> in English _queen_.

As I understand it, the most popular opinion for the dorsals is:

/k^/ was [k]
/k/ was [k_q] (any takers for [q]?)
/kW/ was [k_w]

/k^w/ would then be [kw].

> > a) kW is commoner than k^w
> > b) Sanskrit _ka:m_ 'water' may be connected
>
> Walde Hoffman doesn't like to connect the sanskrit word here, but

the

> Letonian "aka"

What's the meaning? I'm surprised Pokorny doen't have it. If it
belongs, it clinches the argument in favour of *kW.

> > c) I think the root is cognate with *h2ap 'water'.
>
> That would mean that kW >p at a pre Ie stage?

Only *sporadically*, as I said below:

> > For 2(c) we might be seeing sporadic **pW > p outside Germanic

while

> > Germanic would have **pW > *kW; alternatively, *h2ap reflects
> > sporadic *kW > *p.

> Germanic "-apa" is to see under Latin amnis from *ap-nis and

Germanic

> "affa".

What's Germanic about Latin _amnis_?

"affa" looks like an Old High German derivative of Germanic *apon,
referable to PIE *h2abon- < PIE *h2ap-h3on-. Pokorny lists the
words that belong here under *ab.

Richard.