Re: [tied] Evidences for a Mediterranean pre-IE Substratum?

On Sat, 23 Feb 2002 02:37:46 +0100, David Sánchez
<davius_sanctex@...> wrote:

>C. Babaev in his article:
>http://indoeuro.bizland.com/archive/article8.html - Materials about
>Iberians and Iberian Languages
>
>makes this statement: "The Spanish /s/ sound is somehow special and
>dissimilar to correspondent phonemes of other Romance tongues. The
>palatal Iberian consonants which influenced Celtiberian and turned its
>/s/ into /s^/ (compare Gaulish noun endings /-os, -es, -abos/ with
>Celtiberian /-os^, -es^, -ubos^/),

We do not know what the Celtiberian (or Iberian) pronunciation was of
the signs transcribed as <s> and <s'>. In the Botorrita inscription,
<s'> is significantly more frequent than <s>, which speaks against a
equation <s> = /s/, <s'> = /S/.

>(a) In addition I have found a lot of non-Latin words in catalan with
>/s^/ = <x> like <xerinola>, <disbauxa>, <xiruca> ... of unkown origin to
>me.

According to Coromines, <xerinola> (Cast. chirinola) is from the name
of the battle of Cerignola (Italy), <xiruca> is of unknown origin,
possibly from Basque, <disbauxa> is through Occitan from French
<débauche>.

>Also in catalan we have evolutions of s > s^ in some latin words:
>Latin *bassu 'short' > catalan <baix> = spanish <bajo>,

BASSU should have given Cat. *bas, but the form has been altered under
influence of the verb <baixar> < *BASSIARE (where <x> is regular).

>Latin <susu(rrare)> > catalan <xiuxiu(ejar)>, and <pis> 'pee' > catalan
><pixar>.

Onomatopoeic.

>I think some of these facts maybe originated in iberian.
>Iberian has two voiceless sibilants <s, s'> and also vasque seem to have
>complicated sibilant contrast <s, z, x> (all voiceless sibilants!)
>
>(b) But in fact other european languages have a similar /s'/ to that of
>Spain, namely Greek, Finish, some dialects of Northern Italy, and
>perhaps Sardinian and Corsican. In these zones we now historically have
>existed non-indoeuropean mediterranean languages:
>-Northern Italy: Ligurian (not just Celtoligurian or Lepontic) and
>Rhaetic (Tyrrhenian Family). We now about Tyrrenian family that Etruscan
>have also two voiceless sibilants <s, s'> (and perhaps also Rhaetian?)
>-Greece: Pelasgian and/or Minoan
>-Corsica and Sardinia: Nuragian (Culture of Nuraghe)
>-Finland: perhaps its /s'/ is inherited from proto-uralic, hungarian
>also have <sz> contrasting with <s>?.

Hungarian <sz> = /s/ and <s> = /S/, which is another example of the
unpredictability of the way in which s(h)ibilants are spelt (cf.
Polish, Old Czech <sz> = /S/, <s> = /s/). We see the same thing in
many other instances (North Etruscan uses <s>, <s'> for what was
written <s'>, <s> in Southern Etruscan; Hittite uses syllabograms with
traditional Akkadian transcription <s^> to the exclusion of
syllabograms with <s> (so Hittite <s^> was in fact /s/). One
additional complicating factor here is the confusion regarding the
Semitic s(h)ibilants, combined with the fact that most "Indo-European"
scripts are derived from Semitic ones (Etruscan < Greek < Phoenician;
Hittite < [Hurrian <] Assyrian < Babylonian [< Sumerian];
Brahmi/Kharoshthi < Aramaeic < Phoenician; Celtiberian < Iberian <
Phoenician+Greek etc.) In Attic Greek, the shape of the letter
<sin>/<shin> was given the name of the letter <samek> (> sigma), to
represent Greek <s>, while other Semitic s(h)ibilants/affricates were
assigned sound values <zd> (<dz>) and <ks>, or were used as variants
of <sigma> (e.g. <san> -> Etruscan <s'>).

>On the scarce evidence presented here my question is if it is reasonable
>to think in a Pre-IE Mediterranean Sprachbund with these
>characteristics:
>(i) Existence of al least two voiceless sibilants <s, s'>

If this stands for /s/ ~ /S/, it's hardly significant.

>(ii) Tendence to not consider voicedless/voiced contrast as phonemic (as
>in Basque, in Etruscan, and Iberian)

Don't know about Iberian, but voiced/voiceless is certainly phonemic
in Basque. In Etruscan, the distinction is one of
unaspirated/aspirated.

>(iii) Tendence to aglutination (also in Basque, in Etruscan, and
>Iberian)

Most languages can be described as "agglutinative".


=======================
Miguel Carrasquer Vidal
mcv@...