*ap-/*up-/*ip- , rivers in -r

From: tgpedersen
Message: 36363
Date: 2005-02-17

> > >
> > >
> >
> > CVC-VC means
> > a root (CVC)
> > with a suffix (-VC)
> >
> > cf PIE *up-er > Proto-Celtic *wer-.
> >
> > Torsten
> >
> > Did you not assert that *per- was derived from *(a/u)per-?
> >
>
> Make it VC-VC then.


Actually I think *h2ap- or *ap- is a loan, not an IE root.


Kuhn lists the following names of rivers (and assorted lakes and
mountains:
Durius (Duero) Spain
Duranius (Dordogne) S. Fr.
Duri (2x) N. Iral.

Upper Po valley:
Stura (several), Nure, Curone, Taro, Ira, Stirone, Lacus Larius (Lake
Como), Maria (Mera), Lura, Monte Duria

Lower Po valley: nothing

Rhone valley: nothing
Mountains to the west of it: many examples of -ur-, -ar-, -ir-

Mountains of Switzerland, around Aare
Dura (Thur) (-> Rhine)
Suhre (-> Aare)
Jura Mountains
with suffix:
Orbe, Surb (-> Aare), Urnäsch (-> Sitter -> Thur), Uerke, Murten
(town), Murtensee, Sorne (-> Birs -> Rhine), Sarine, Sarnen (town),
Sarnersee, ri (town), Urnersee, 5 x Murg.

with -k- suffix, especially many in the Nertherlands, in an area
delimited by
w: Ourcq (-> Marne)
e: Orke (-> Eder -> Weser)
s: Ource (-> Seine)
n: Urk (island)
many names of the type
Urk-, Burk-, Kurk-, Lurk-
Ark-, Bark-, Kark-, Sark-, Mark-, Wark-

in the mountains to the right of the Rhine
Or-pe/Urf, Dor-pe, Sor-pe, Lor-fe,
Ar-pe, Mar-pe, Sar-pe.

in Lithuania:
Durupis, Nurupis, Surupis, Urkupis.


Kuhn: the areas of the ur-/ar-/(ir-) names are relic ares.

He identifies the -pe/-fe suffix (and Lituanian -up-?) as PIE *ap-
"water" and thinks therefore these composites with the non-IE ar-/ur-
placename roots as late. But the suffix is ap-/up- and is therefore
itself an non-IE ar-/ur- word-


I think *ap-/*up- is the origin of the preverbs/pre/postpositions *ab
and *up. Since it means also "river", *ab/*up was loaned as a noun
and perhaps in a single or two cases (locative, allative?). So what
we see in Or-pe etc is perhaps "on" or "up the Or- river", depending
on whatever old case used to be discernible in the now eroded -pe.

In a language with cases, the number of "semantic cases" can be
extended by using a noun indicating position, in a positional or
directional case, as a post/preposition of a noun which is then in
the genitive or some other suitable case, eg. Finnish 'Porin ja Turun
välillä' "between Pori and Turku" (-n genitive, välilleä
adessive "on the middleroad", thus "on the middleroad of Pori and
Turku"; 'keskellä järveä' "in the middle of the lake", keski-
"middle" in the adessive, järveä inessive of järve "lake).
But one might also see keskellä as an adverb "centrally"
and järveä as "on the lake".


*ab/*up in suitable case also behaves like an adverb (one
might want to compare it to a similar "international"
system of directional adverbs, that of 'starboard'
(Danish styrbord) and 'port' (German Backbord, Danish
bagbord, French babord)).


This means that -r of all the non-IE ur-/ar- place names may
be a genitive or locative marker (if I can explain away the
-k- of Lithuanian Urkupis). IE does have a relic -r locative
in pronouns (here, there, where, Lat. cur).

Cf.
German dar-auf "thereon"
Dutch daar-op "thereon"
rivername Or-pe "on the U river(?)" etc

which means going back to the strange fact that the Germanic
languages use this old locative(?) with 'floating' (Dutch) or
postposed prepositions, for which there is no logical
explanation.

Especially Dutch is nuts about this preference for locatives with
prepositions.
Not only
er-voor
hier-voor
daar-voor
waar-voor
as in the other Germanic languages, but also
ergens voor lit. "somewhere for" ie. "for something"
nergens voor lit. "nowhere for" ie. "for nothing"
overal voor lit. "everywhere for" ie. "for everything"


Perhaps -r is the mark of a genitive/locative used as the subject
marker in stative sentences in an active/stative language?

"water" has at least four forms in IE:
*w-d- : Greek hudro-, water
*w-g- : Greek hugro-
*w-s-
*w-r- : Latin u:ri:na (Varro: u:rina:ri est merger in aquam), urium
(Pliny: vitium lavandi est, si fluens amnis lutum importet, id genus
terrae urium vocant, Iberian?, cf river Urium in Sp.), urceus "pot",
urna, Greek oureo: "I urinate", Skt. vá:r, vá:ri, Tokh A wa:r,
OPr wurs "pond", ON úr "drizzle". Cf. Basque ur "water", irura
"vega, valle". Cf the town Aurbach <- Ur-augia; the latter is then
PIE *akW- "water", which, being a variant of *ap-, also must be a
loan from the ur-/ar- language.

One is tempted to consider Latin 'urbs' which has had a whole
supposed substratum language in Latin named after it: Urbian, as
related to the ur-/ar- language.


Torsten