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

Hallo!

On Wed, 21 Jan 2009 19:35:58 +0000, R A Brown wrote:

> Jörg Rhiemeier wrote:
> > Hallo!
> >
> > On Wed, 21 Jan 2009 12:31:15 +0000, R A Brown wrote:
> [snip]
> >>       Whether
> >> all them would have been is a different matter; but I agree that such
> >> evidence as we have does point to a greater prevalence of the ergative
> >> model in the pre-IE period.
> >
> > Very likely.  Today, Europe is almost entirely accusative, but that
> > is only due to the spread of Indo-European and Uralic.  Basque and
> > most Caucasian languages are ergative; that of course doesn't mean
> > that *all* pre-IE European languages were ergative (Etruscan, at
> > least, wasn't), but some certainly were.
>
> True - Etruscan shows no signs of ergative structure. I suspect we would
> find a variety of structures in the neolithic languages of Europe.
Right.  I consider it likely that PIE made a shift from active to
accusative alignment earlier in its history; and the language of
the central European neolithic farmers (which I suspect to be a
sister language of IE) could have retained the earlier active
alignment, or shifted to something else, or whatever.  And the
languages spoken before that could have been *anything*.  Indeed,
I find it very regrettable that Indo-European bulldozed it all
away :(

> [snip]
> >> There are also vocabulary items. There is set of non-IE words common to
> >> Germanic & Celtic (e.g. *landa, *comba), but also a set of non-IE words
> >> found only in Germanic and another set found only in (insular) Celtic.
> >
> > Yes.  I have a list of non-IE Germanic words which I have
> > extracted from an etymological dictionary of German, and
> > another I once found on a web site which by now is gone;
> > I would be *hotly* interested in a list of substratum
> > words in Insular Celtic for my Albic project.  Do you have
> > one, or can tell me where I can find it?
>
> I do not & cannot. I feel sure someone (if not some ones) must have done
> work on these lists of substrate words. I must make a serious search.
I haven't found anything yet, either, though there must be
something like that *somewhere*.  I simply cannot believe
that either nobody ever investigated the etymology of Insular
Celtic languages, or nobody found blank spots at that.

> [snip]
> >>       There
> >> are theories that posit an IE origin for Etruscan! And I guess you have
> >> come across the 'IE Pelasgic' theory.
> >
> > Sure.  There are some vexing similarities between IE and Etruscan
> > in the morphology, but by far not enough to establish a relationship.
> > Indeed, Uralic is even more similar to IE than Etruscan.
>
> Agreed.
Indo-European and Uralic look so similar that one can hardly
dismiss the impression that they are related to each other,
though "Proto-Indo-Uralic" (or whatever to call it), if it
ever existed, must have been spoken about 10,000 years ago,
if not earlier.  Also, the homelands of PIE and PU are quite
close to each other.  An IE-Uralic relation was suspected
already by the venerable Jakob Grimm.

> [snip]
> >
> > An IE-Etruscan relationship hypothesis is also peddled by a
> > certain Glen Gordon of the Nostratic-L mailing list, but his
> > evidence is extremely shaky; he always cites the same "cognate
> > pair", namely IE *kWetWor- '4': Etr. _huth_ (which he claims
> > was '4', but actually probably was '6').
>
> Groan - I would be a very rich person, I think, if I could be given a
> pound/euro/dollar for every theory based on supposed cognates of one or
> two words where the meanings are dubious at best    :)
Yes.  One cognate set is no cognate set.  You can always find
chance similarities.  In order to prove a relationship between
languages, you need at least about 100 cognate sets displaying
regular sound correspondences.

> Indeed, Etruscan _huth_ is (almost) certainly '6'. On Roman dice the
> opposites sides add up to seven as, indeed, the do on modern dice. Are
> we to assume that on the Etruscan dice where the numbers are written as
> words, that while one pair, namely 2 ~ 5, did add up to 7, while the
> other pair of opposites were 1 ~ 4 (= 5) and 3 ~ 6 (=9)??
Fiddlesticks.  The dice argument is known to me, and as the
word for '1' is securely known (because it is used with singular
nouns while all others are used with plural nouns), _huth_ can
indeed only mean '6'.

> Normally such a suggestion would be laughed out of court. The only the
> reason that some claim that _huth_ = '4' is that according to Stephen of
> Byzantion the Attic place-name  _Hytte:nia_ means 'Tetrapolis'
> (Four-city), and _Hytte:nia_ is assumed to be of Tyrrhenian origin. But
> this argument commits a serious methodological error: it is assumed that
> Hytte:nia_ is derived from a language related to Etruscan to derive the
> meaning of Etruscan _huth_, and then the supposed Etruscan _huth_ = '4'
> is used to 'prove' that _Hytte:nia_ is derived from a source related to
> Etruscan! 'Tis a vicious circle.
Right.  There are so many speculative assumptions involved here
that the argument can be dismissed out of hand.

> IMHO any theory based Etruscan _huth_ being a cognate of IE *kWetWor- is
> worthless.
>
> >>       What Kretschmer would have
> >> made of the Nostratic theory/theories, i don't know.
> >
> > He would perhaps have considered "Protindogermanisch" to be
> > a branch of Nostratic.
>
> I suspect he would.
Possible.  However, the fact that he supported a particular deep
relationship hypothesis doesn't mean that he'd have supported a
particular different one.  I recently read a book (_Einführung
in die eskimo-aleutischen Sprachen_ by Jan Henrik Holst) in which
the author vitriolically inveighs against Nostratic, Greenberg's
Eurasiatic and Fortescue's Uralo-Siberian (of which the latter
looks somewhat better to me than the other two) and then goes on
to present a "proof" of an Eskimo-Wakashan relationship, which,
if you ask me, is far less convincing still.

> > Note also that Allan Bomhard considers
> > Etruscan to be Nostratic without giving evidence - he takes it
> > as given that it was an aberrant branch of IE, which is of
> > course bogus.
>
>      LOL!
Well, the whole Nostratic business is very problematic.  It tells
a lot that the Nostraticist community is divided between two camps
who use different sound correspondences, and yet both camps present
lists of 500-something alleged cognate sets.  Of course, at most
one side can be right, which probably means that both are wrong :)

If you ask me, it is likely that Indo-European and Uralic are
related, but that's about it.  I used to be quite convinced
(though not entirely convinced) of Fortescue's Uralo-Siberian
hypothesis (i.e., Uralic, Yukaghir, Chukotko-Kamchatkan and
Eskimo-Aleut), but recently, doubt has been growing in me,
especially in light of the huge distances and considerable
anthropological differences involved there.

> [snip]
> >> The first chapter of my "Evidence for Pre-Greek Speech on Crete from
> >> Greek Alphabetic Sources" (Amsterdam, 1985) reviews  several strange
> >> pre-IE eurolang theories, most of which give ample scope for the LLL  ;)
> >
> > Yes, there are many strange theories about the linguistic
> > landscape of pre-IE Europe, simply because so little is known
> > about it.
>
> Yes, and the certitude by which these theories are posited seems to be
> in inverse proportion to the evidence available.
Yep ;)

> Now, I must get on with my mesolithic grammar........
>        ;)
Have fun!

... brought to you by the Weeping Elf

Replies