--- In
cybalist@yahoogroups.com, "tgpedersen" <tgpedersen@...> wrote:
>
> --- In cybalist@yahoogroups.com, Rick McCallister <gabaroo6958@> wrote:
> >
> >
> > --- tgpedersen <tgpedersen@> wrote:
> >
> > >
> > > > I remember on another list, substrate list, I think,
> > > > someone referred to IE elements in a substrate of Pyrenees
> > > > Ibero-Romance.
> > > > There are IE looking elements in Basque that are
> > > > non-Celtic, non-Latin/Romance. Some of them also seem
> > > > to show up in Sardinian, which suggests they may be
> > > > related to Lusitanian.
> > >
> > > Many of those IE loans in Basque Trask meant were from Latin
> > > have -a-, as if from the 'mot populaire' layer of Latin.
> >
> > Yes, but there are others that can't be Latin, twisted
> > or not: urki "birch", abarka "bast sandals, originally
> > bast". Trask saw these as coincidence.
> >
> Sigh. Seems I have to translate another Kuhn article. This one's
> gonna be long.
>
Translation:
"15 years ago, I finally managed to track down a second large and old
river name complex which gives itself away by similar characteristics
as the Krahe hydronymics: the distribution over large spaces and the
limitation to certain word stems and derivational elements occurring
widely, and further an even stricter limitation in phoneme inventory.
Above all it is the phoneme sequence 'ur', beside that also 'ar' and
'ir', in the stem syllables which characterizes this set of names and
unite them. It is stems such as Ur-, Dur-, Kur- and Stur, further with
suffix consonants Durs- and Murs-, Urk- and Burk- and more in that
stile. With the Krahe system this new one is related not just in
general in the limitation to a few stems and the use of certain types
of derivation, but in particular in the lack or great scarcity of the
IE fundamental vowels /i/ and /o/. This draws them close together.
However, in Krahes sets /a/ the most frequent by far and /u/ the least
frequent of the three vowels, while in the newly discovered /u/
predominates; and /a/ follows somewhat distantly.
The striking limitation of my sets on not only the stem vowels /u/,
/a/ and /i/, but further also their position before /r/, confirms, it
seems to me, my conviction that also for Krahe's hydronymics the
limited phoneme inventory is an essential characteristic - Krahe
denies that vehemently - . The two systems also seem to have differed
in that in Krahe's system the suffixes always begin with a vowel (e.g
Sal-usia and Sal-ika), in my system on the other hand they equal a
consonant (Dur-s-, Ur-k- etc). I referred to the "ur-/ar- system", as
I call it, for the first time in my review of the already mentioned
book by Krahe (AfdA. 78, S.4 6), and in more detail in Namn och Bygd
59 (1971, S. 52-70). Here I now limit myself to what is important to
my present subject.
This is firstly the area of distribution. It is larger as that of
Krahe's name groups and seems to transcend by far those boundaries of
Europe which I included in my works. It stretches as well in the North
and West as in the South to furthest coasts of our continent, and in
the Southeast to at least the countries around the Black Sea. An
eastern boundary I can hardly make out. In most of the countries these
names, as far as I see, spread thinly. In a number of large landscapes
they are almost totally absent, in others their number far surpasses
the average.
This strong variation was one of the things which initially directed
my attention towards the ur-/ar- names. The area with the greatest
concentration, to my knowledge, is that around the Ardennes. It can be
approximately circumscribed with four Urk- names: in the West the
Ource (to the upper Seine) and the Ourcq (to Marne), in the North the
old island Urk (in the Zuidersee) and in the East the Orke (to Eder in
North Hesse, cf. NoB. 59, S. 55). Urk as island name shows us, as also
the following, at the same time, that the names in question were not
limited to rivers - there are many island names among them, from
Greece to Norway -.
The landscape thus defined, which is particularly rich in ur-/ar-
names, stretches in the Northeast to a line, which I call the Borken
boundary, since it runs approximately from the island of Borkum to
Borken in North Hesse and four more towns with Burk- names are
situated close to it (s. NoB. 59, 56). Beyond it follows a space
stretching deeply into the East, which, with the exception of a narrow
strip at the water's edge, is almost free of ur-/ar- names. My next
question is now, to which language family did those communities
belong, which left the ur-/ar- names. Those characteristics mentioned
above, which this system has in common with Krahe's, make it very
likely, that they have come from closely related languages. From those
above-mentioned conditions which Krahe started from, it is almost
self-evident, that he saw his hydronymics as purely IE. He stresses
several times, that all their elements, stems as suffixes, are
provably IE (loc. cit., p. 32 and elsewhere), although he only needed
to have maintained, that they all could be explained from PIE - or IE
daughter languages - name stems on which that didn't succeed he
excluded! . He didn't recognize that almost all of it just as easily
could be explained from non-IE roots and his name sets in their
totality contain essential elements, which early IE cannot explain.
Most important there is the already mentioned striking preponderance
of the /a/ and beside that /i/ and /u/ in all syllables against a
great scarcity, if not complete lack, of the early IE fundamental
vowels /e/ and /o/, which positions his hydronymics, as mentioned,
close to the ur-/ar- names. Since in Krahe's name system I could only
recognize its distribution as properly IE, I came to the conclusion,
that the western IE people took over the foundations of this
hydronymics from earlier inhabitants, subsequently developed it and
distributed it over the lands in their possession, but that their
foundations must have been very close to the ur-/ar- system. In this
I didn't take account of the fact that the Indo-Europeans in Late
Stone Age had introduced into their language, particularly into its
morphology, the high-intensity use of /a/, as I had myself
demonstrated, this, granted, with great probability from a foreign
language family (KZ. 71, S. 143 ff., = Kl. Schr. I, S. 230 ff.), but
now could continue this use in the names they created, with no further
foreign influence. In spite of that, the relationship of their system
is so close to the second one which I have drawn attention to, that I
can't believe they are independent of each other. I'm thinking here
of the principle we would find difficult to understand, that name
stems should be chosen not just for their meaning, but that their
phonetic form should be at least an important, if not the sole criterion.
In my "second Old Europe" I see nothing which points to the
name-givers being Indo-Europeans. The phoneme sequence 'ur' was just
as unfamiliar to early IE as the /a/ in general and thus also /ar/,
and with /ir/ it would have been similar. Also ending-less
nominatives, as exemplified in Dur (in Ireland) and Nar (in Central
Italy), would be foreign to IE. Likewise the distribution area of the
ur-/ar- system that I sketched speaks against Indo-Europeans being its
creators and distributors. Where we with the highest probability seek
to place the earliest seats of the Western Indo-European groups, in
eastern and also central Europe, the ur-/ar- names within my subject
are the least frequent. The two old systems do not exclude each other,
where Krahe's system has expanded to, but in general it is so that the
lands with a strong share of -ur- describes a large arch around the
early IE core areas of Krahe's name sets. The explanation for this
geographical state of affairs as well as of what we otherwise learned,
I believe, is obvious: the ur-/ar- names are the oldest and were, it
seems, once in use in the furthest regions of Europe (and also beyond
them). Then from the east Indo-Europeans broke in, adopted the river
name system which they found, reshaped it, and introduced it into the
countries they conquered. Thus the distribution area of the older
name sets was hollowed out, and to a large extent pushed into fringe
and coastal areas, as well as mountain lands. In this process they
seem in the beginning to have adapted almost all names of the older
system to their new one or else replaced them, but later to have let
more and more of them be.- Thus wide seams of mixing and transition
were formed, in some stretches however - most likely where the
boundary was stationary for a long time - rather clear divisions.
That's the way it is, according to my material at the above-mentioned
Borken boundary and in western Denmark, further on the upper Rhine
between Lake Constance and the Vosges mountains, in the uppermost Po
area and probably also on the lower Rhone (but cf. p. 23 f. below on
the genesis of two first-mentioned boundary sections). All these
boundary divisions have in common that the ur-/ar- names are situated
on the far side from the Central European core area of those names.
This is hardly a coincidence and only confirms, that what pushed it
back came from the east. If this is true in general, then my
Nordwestblock, which I believe I can count as one of the homelands
of the last Indo-European, was not long in their possession, and it
was further divided by the important Borken boundary. Already before
became aware of the ur-/ar- names and its boundary there, this line
had caught my attention, because other equally suspiciously non-IE
names and derivational elements coming from the West or Southwest,
stop at approximately the same boundary, and I assumed already
then, that an important boundary for pre-IE had been situated there
(Abh. d. Mainzer Ak. 1963, S. 562-68). Even a connection leading
along the North Sea to related names way up in the North, began to
take shape then. But already then I could present a couple of examples
which show that also in the northeastern part of the Nordwestblock
names of a pre-IE substrate language were not completely absent. Most
of it was then shortly thereafter expanded by the mass of ur-/ar-
names to an unexpected degree, and also, I believe, confirmed. But
here, at least with respect to the Borken boundary, we may only speak
of the names - and then perhaps of the languages (on this point cf. p.
23 f. below). Prehistory knows nothing of this boundary.
For the task I assumed here the above facts and considerations have
above all the significance that the last Indo-European in the only
area, in which we can grasp anything worth mentioning, is on top of a
foreign substrate. We might consider as certain that much of it has
been absorbed by the Indo-European, which prevailed there. This is
confirmed, I believe, by certain phonetic phenomena which makes its
legacy recognizable (on this point cf. Kuhn, Festg. f. L. L.
Hammerich, 1962, p. 122 f., = Kl. Schr. I, p. 398 f., Gedenkschrift f.
W. Foerste, 1970, p. 50 f. and Festschr. f. K Bischoff, 1975, pp. 7,
19, 21 and 26). The foreign material, which from the very beginning
must have entered the Indo-European in many ways, would here thus be
reinforced one more time yet. Above all, among those name equations
with which I will attempt a to add a portion of our Northwest to the
southwards migrations of western Indo-European groups and thus to
ascertain that it was part of the settlement area of the last
Indo-Europeans, there will be all manner of non-IE formations which
had been taken over by the Indo-European groups pushing westwards.
Their connection with southern European names perhaps rests in part on
relatedness already in pre-IE times. A further source of errors lies
in the fact that many names of the Nordwestblock which have close
relatives somewhere in southern Europe, perhaps also have existed in
the more easterly parts, no more accessible to us, of the Central
European home of the western groups of Indo-Europeans and have arrived
in the South from there. These are sources of uncertainty which are
difficult to get around. On top of that there are the general ones,
namely the coincidental and mostly secondary similarities - the oldest
name forms are famously seldom certain -. I therefore assume, as
always, that I have run into all manner of false combinations, and I
know that as many parallels as possible must come together in order
that the false ones among them do no damage worth mentioning. On the
other hand I know that only it is small relics of what once existed
that we can still grasp and use for comparison. And further, I have
only accessed rather small parts of the still available and accessible
old names, and this on both sides. It is hardly more than samples, and
yet most of them brought more than I had expected.'"
This resolves the puzzle of the *-k- suffix of Latin etc *ped-k- (in
pecco: etc) and *man-k- (in mancus): they are pre-IE words with pre-IE
suffixes loaned by Venetic and then loaned from Venetic by Latin,
Germanic (before and after Grimm), Celtic etc.
Note on Kuhn's use of the term 'The last Indo-European': it seems he
identifies Krahes Alteuropäiasch ewith the earliest IE, perhaps with
PIE itself, which is of course wrong. If the Veneti (new etymology:
from *wes-n- "merchant", later "beggar" (Pokorny), cf the Dutch 'sea
beggars".
Note that the -k- suffix may occur also in the non-Basque examples
from Basque Rick provided:
urki "birch",
abarka "bast sandals, originally bast".
Torsten