> What are the opinions of others on this group as to whether
> Tyrhennian once shared a common ancestor with the NE Caucasian macro-
> family, as Marco Marreti suggests, or with Indo-European, as Glen
> Gordon suggests, or as a language isolate? Or, is it smart to just
> let the jury remain out and not rush to any conclusions as of yet?
> -Michael
Well Michael, I'm biased towards the idea that IE and Tyrrhenian
share an affiliation and that they are completely _seperate_ from
NW or NE Caucasian.
Most accounts of the Nostratic hypothesis include both IE and Kartvelian
(_South_ Caucasian) together, although Kartvelian is surely very
different and seperate from IE with few non-areal commonalities.
Comparing ablaut systems between those two groups proves nothing other
than that perhaps they were in a general area together for long periods
of time, although ablaut can arise independently in both groups.
I will admit though that NWC loanwords could have infiltrated early
stages of IE (or IndoTyrrhenian) which may account for some of the
connections (most of these 'connections' are still dubious though).
Bomhard thinks so in "Indo-European and the Nostratic Hypothesis" and
lists a few roots that appear to coincide with IE roots in a certain
branch of NWC called Circassian. He concludes this because he feels,
and I agree, that IE took a root from Central Asia to its later
location in the NorthWest Pontic. The jury is still out on these NWC
similarities though.
Marco compares Tyrrhenian with NEC and he's not alone, sadly, even
though the opinion as far as I'm concerned is completely half-baked.
Personally, if I find that the Etruscan 1ps pronoun is /mi/ (IE *me)
with oblique case form /mini/, demonstratives /ica/ (IE *ko-) and
/ita/ (IE *to-), interrogative /ipa/ (IE *kWo-), genitives in /-s/
(IE *-os) and /-l/ (cf. Anatolian l-genitives and IE *-ol-), adjectives
in /-na/ (IE participle *-no-), a collective in /-(cH)va/ (IE *-h2)
and some straightforward parallels in vocabulary like /cara/ 'to make'
(*kWer-) and /nefS/ 'nephew' (Lemnian /nafotH/; IE *nepot-), I think
that there shouldn't be any confusion whatsoever. IE is related to
Tyrrhenian and retains archaisms that hark back well before even
Anatolian... hence IndoTyrrhenian, the fracturing of which would have
occured between 7000 and 6000 BCE. In contrast and just to nail the
coffin, the NEC 1ps is not anything like Etruscan /mi/. It appears to
be something like *zo (Ingush /so/) and has absolutely no similar
oblique form like /mini/ (only IE does and other Nostratic languages
like Uralic).
I've been constantly working and reworking my theories on sound
correspondances between IE and Tyrrhenian and have come up with some
latest conclusions. The IndoTyrrhenian, like IE, had a three-way
contrast of stops. Either something to the effect of *t (lax voiceless),
*t: (tense voiceless/half-voiced), *d (voiced) or perhaps *t (voiceless),
*t? (ejective) and *d (voiced). It's clear to me now that the Tyrrhenians
had deflated this system into a simpler 2-way contrast of aspiration,
doing away with any voicing contrast. Further, *p and *b (which become
*p and *bH in IE) evolved differently from the other stops, becoming
*f and *p respectively due to a pre-stop-shift change of *p>*f. It
evolved in such a way that the aspirated series of *[tH, kH] (**pH
didn't exist) was 'marked' and thus less frequent than the plain series
of *[p, *t, *k]. So, I now contradict some of my earlier ideas by
asserting that both the 'plain' and 'tense' (or ejective?) stops merged
to the more common inaspirate series while *b, *d and *g (which became
*bH, *dH and *gH in IE) became *p, *tH, and *kH respectively. This
explains some unignorable correspondances like /-pi/ 'beside' (IE *bHi
'by') and /-tHi/ 'in, at' (*dHi 'in, at').
The development of the velar series is further complicated but easy to
understand. Delabialization of *kW, *k:W and *gW to their respective
non-labialized counterparts is clear. So *kW merges with *k and both
become Tyrrhenian *k (hence Etruscan /cara/ <=> IE *kWer-). However,
in IndoTyrrhenian all velar stops were automatically pronounced as
uvular when they neighboured *a -- This is merely an assimilation by
vocalic quality of [+low] which *a has and which uvulars also are
known to have. One might say that uvulars are a-coloured phonemes just
as palatalized phonemes are i-coloured or labialized ones are u-coloured.
Well, these allophones were eventually phonemecized (no longer just
'automatic' but seperate phonemes) in both IE and Tyrrhenian. IE retained
traces of the previous allophony with *k (former 'palatal' *k^) and
uvular *q (former 'plain' *k). In Tyrrhenian, the uvular allophone
merged with aspirate *x (normally correlating with IE *x/h2 and *hW/h3)
which later oscillated in a predictable fashion between /cH(v)/ and /v/
as we see in the collective (cf. pulumcHva, susleva, etc). It becomes
/cH/ when preceded by consonant but /v/ when intervocalic. This now
also explains the otherwise unexplainable alternation between /macH/
'five' and /muvalcH/ '50', probably deriving from something like *mexa
and *moxalxa respectively. We also find uvularized velar stops reduced
to *x in initial position, as with *xota 'four' (IE *kWetwor-) both
from IndoTyrrhenian *kWatWan.
Well, I have to go, sorry for the long rant but I wanted to be clear
about my position on these language groups, how they specifically
related to each other and why a relationship between these groups and
NEC or NWC is completely untenable.
You decide.
= gLeN