> No. I did not meant the gothic "ahwa" (BTW why are you using the
laringial
> notation here for *akWa?; my Dictionary is not sure and means there
can be
> *akWa or *ak^wa:)
> I have to remember where I seen some toponims/hydronims in Germanic
space
> where "afa" is present and it was translated as "water". Maybe
Torsten find
> it quikier I do. OK, I will look for to see where I got that and as
soon I
> find I will revert to the subject.
>
I've read Kuhn again and had to correct my memory. The apa-/affa-
names (Low German example: Ennepe) are found in a broad swathe
_south_ of the Nordwestblock, in the area elsewhere called para-
Celtic (or the like). The ap- 'thing' comes (perhaps) in variations
ap-/up-/ip- which makes it tempting to assign it to Kuhn pre-IE ar-
/ur- language, which was dispersed wider than the later IE
Nordwestblock.
http://www.angelfire.com/rant/tgpedersen/Op.html
Perhaps it replaced it so thoroughly in the "classical" Nordwestblock
area that all apa- names were wiped out?
BTW on the flimsiest of evidence (the alternation akWa/apa, but see
also
http://www.angelfire.com/rant/tgpedersen/pd.html
on the assumption that this other 'waterfront' word is a derivative)
I proclaim that non-IE Nordwestblock / the ar-/ur- language was
wavering on kW/p, which, when doublet glosses were loaned into the
succeeding IE languages, caused p/kW to become a shibboleth, with a
purge of one variant as a consequence in the various successor
languages. One could choose to see Latin 'quinque' and 'coquo'
as 'cothrige' words, with the development running the wrong way (of
course the following labiovelar also helps;-) ).
On 'waterfront': The German designation for the lowlying Platt-
speaking areas is the 'Waterkant', cf Cantium -> Kent, of which
English 'waterfront' is a loan-translation? There's a 'Waterfront' in
Cape Town.
Torsten