Re: [tied] Albanian outside the centum-satem division?

----- Original Message -----
From: Abdullah Konushevci
To: cybalist@yahoogroups.com
Sent: Wednesday, September 21, 2005 2:11 AM
Subject: Re: [tied] Albanian outside the centum-satem division?


On 9/20/05, Abdullah Konushevci <akonushevci@...> wrote:



On 9/20/05, Abdullah Konushevci < akonushevci@...> wrote:
> Palatals in Albanian
> *aig'- 'goat' > Alb. <edh> ‚goat kid' (*ai > e, *g' > dh);
[...]


When I started this thread, my idea was not to prove Albanian not to be a
Satem languages (because it is obvious for me). There is a view according to
which, of three velar rows, prevelars *K^ merged with clear velars *K in
Centum languages, and velars *K merged with labiovelars *KW in Satem
languages.

Albanian is said to be outside the division because it has preserved traces
of all three rows unmerged, namely *K^, *K and *KW. Especially, before a
front vowel, original *k^ should change into th, *k should stay unchanged,
while *kW should change into s.

The following examples are said to be typical:

1) *k^: thom < *k^ens-mi 'I speak'
2) *k: kohë < *ke:so- 'time'
3) *kW: se < *kWe- 'what'

I doubt it because I just know no more examples for preserving k unchanged
before a front vowel. And I assume that *ke:so- had *k^ originally, not *k.
Of course we have Slavic evidence for *k in this word but it contain -s-,
and according to Meillet's rule saying that if the stem contains -s-, then
*k^, *g^(h) > k, g in (Balto-)Slavic, not s, z (see also gQsI 'goose' <
*g^hans-, kosa 'scythe' < *k^os- etc.). So, the example is not very
convincible.

As one can see from the Abdullah's list, the development of *k^, *g^(h) in
Albanian was very complex and there are no one single continuants of the
previous IE prevelars ("palatals"). Together with "regular" th < *k^ and dh
< *g^ we can see:

1) d like in dyllë 'wax', cf. Greek khylos 'juice' < *g^hus-l-; dorë 'hand',
cf. Gr. kheir < *g^he:sr; dimën (Geg.) 'winter', cf. Greek kheimo:n; dje
'yesterday' < *dhg'hies- and many more

2) z like in zë 'voice', cf. Slavic zvonU 'bell', zve,kU 'sound'; njëzet
'20' < (d)k^mti- etc.

3) q like in quhem 'I name', cf. Slavic slys^e^ti 'hear' (with no Meillet's
change) and Lith. klausyti, etc.

4) k, g like in mjekër 'beard', cf. Lith. smakra (Meillet's rule!) and
Sansk. çmuçru- 'moustache' (possible assimilation < *k^mVkr-); gardh 'fence,
hedge' < *g'hr.H-dho; (btw., "centumic" development is known from all Satem
languages, even from Sanskrit)

5) NULL like in i dhjeti 'tenth' < *dek^m.-to

6) Possibly also b, palatalization of n > nj etc.

In other words, because Albanian development seems to be very various and
sometimes even enigmatic (to tell the true, no conditions for initial d/dh <
*d(h), *g^(h) are known), the only example of kohë is no evidence for
preserving three velar rows in proto-Albanian.

Really, if Albanian preserved all three rows, unlike Satem and unlike Centum
languages, please give me more examples (for distinction between *ke and
*kWe for example, not for *k^e). And I will be stating that Albanian is a
typical Satem language until I see more evidence.

Grzegorz J.



___________________________________________________________
To help you stay safe and secure online, we've developed the all new Yahoo! Security Centre. http://uk.security.yahoo.com