Re: languages of pre-I.E. Europe and onwards

Hallo!

On Tue, 20 Jan 2009 12:35:16 +0000, R A Brown wrote:

> Hi all,
>
> The site below was recently brought to my attention. I thought some of
> you on this list might be interested in this discussion on the languages
> of pre-I.E. Europe and onwards....
> http://languagelog.ldc.upenn.edu/nll/?p=980
Yes, that is a very interesting read, and I have already commented
on it here a few days ago after Roger Mills mentioned it in the
"Amibuity" thread:

http://listserv.brown.edu/archives/cgi-bin/wa?A2=ind0901b&L=conlang&T=0&F=&S=&P=4883

The linguistic diversity of Europe must have been quite high;
Ringe's idea that pre-colonial North America is a good model
for pre-IE Europe is quite a good one, I think.  The area now
occupied by the United States and Canada holds about 50
indigenous language families - a few moderately large ones
(20-30 languages each, such as Algonkian and Uto-Aztecan),
many smaller ones and a handful of isolates.  Europe, of
course, is only about half the size of USA+Canada, so we
should expect about 25 families in pre-IE Europe.

What were those languages like typologically?  We don't know.
I think the North American analogy ends here; I don't expect
pre-IE European languages to be as massively polysynthetic
as North American languages, but the survivors (Basque and
the Caucasian languages) are all strongly synthetic, and many
of them ergative, and pre-IE Europe may have been like that,
too.

I have a hunch, though, that there was an intermediate layer,
now completely extinct, between the Palaeolithic/Mesolithic
heritage and the spread of Indo-European.  There are two
reasons for this:

1. The spread of agriculture north of the Alps appears to
   have been demic, i.e. borne by people immigrating into
   the area, apparently up the Danube from the northwestern
   shore of the Black Sea (if the Black Sea flood disaster
   really happened, as some geologists assume, they could
   have been refugees from that event); this would also
   mean that those immigrants also brought in their language.

2. There is an apparently uniform network of geographical
   names, especially river names, covering a large area
   in western and central Europe, and appearing to stem
   from an unknown language or language family spoken in
   the area before the historically attested Indo-European
   languages moved in.

Of course, some people assume that this neolithic language
was Indo-European, but PIE lexical items such as *kWekWlos
'wheel' (and other words for wagons and parts thereof) and
*h2ayes 'copper' set a terminus post quem at 4000 BC.  Also,
the river names do not fit the phonologies of the attested
IE languages, so they must have been borrowed from an
unknown source.

My assumption is that the unknown language we are dealing
with here was a sister language of PIE which branched off
before the ablaut system emerged in the latter.  The homelands
of both families sat side by side on the Black Sea shore,
possibly both being refugees from the Black Sea flood event;
many of the "Old European" river names seem to make sense
when read as Indo-European, once the difference in vocalism
(the river names have /a/ all over the place where the
seemingly corresponding IE words have /e/ or /o/ - this is
precisely what one would expect from a sister group of IE
which did not undergo ablaut).

The German linguist Theo Vennemann has his own ideas about
pre-IE Europe, though.  He consideres the "Old European"
river names to be "Vasconic", i.e. from a language family
of which Basque is the last survivor.  He also entertains
the notion that Europe north of the Alps was uninhabited
during the last ice age.  This latter point, of course, is
patently false (and I am not even considering Neanderthals
here, which contributed nothing to the modern gene pool
and probably also nothing to the linguistic landscape of
Europe, either), and the assumption that pre-IE Europe
north of the Alps was occupied by a single language family
is unwarranted.

There are even more problems with Vennemann's theory.
One is that his etymologies make little sense, as one
gets names meaning "Wet River", "Water-River" and all
that.  Also, one would expect such names especially
in the only area of which we know that Basque or something
related to it was ever spoken, but the distribution of the
"Old European" river names shows a gap exactly there.
There are no OE river names south of the Garonne, and on
the Iberian peninsular, they are restricted to the northwest.
Third, Vennemann must take recourse to brute force to make
the OE hydronymy fit typologically with Basque.  In an area
stretching from the British Isles through Germany into Italy,
some of these names contain the spirants /f T x/ which do
not exist in Basque and which Vennemann denies to exist in
his "Vasconic".  How does he explain them away?  By assuming
an Italic intermediate substratum!  Italic in Italy, OK, but
in Germany?  In Britain???  Britain especially becomes a
crowded place for substratum languages in Vennemann's cosmos,
as he also assumes an Afro-Asiatic substratum there in order
to explain the typological aberrancy of Insular Celtic!
No, I don't buy all that.  Vennemann may be a knowledgable
expert on Old High German (or so I have been told), but
his model of pre-IE Europe is ridiculous.

All this is of course fodder for the League of Lost Languages!

... brought to you by the Weeping Elf

Reply