"centum"/"satem" "exceptions" [was Re: Northwest IE attributes]

[Miguel Carrasquer Vidal]

Sean Crist <kurisuto at unagi.cis.upenn.edu> wrote:

>> Further, remember that there are other "satem"  branches, such as
>> Armenian and Albanian.

>That's not correct; Armenian underwent an independent set of consonant
>changes which roughly resemble Grimm's Law in Germanic.  Armenian did not
>undergo the satem consonant shift.

It did:

*k^  > s        tasn < *dek^m
*g^  > c [ts]   gorc < *werg^om
*g^h > j [dz]   jeRn < *g^hesr-

Additionally, *uk > us even if the velar was not palatal in PIE.
(Also, like the other satem lgs., Armenian underwent the
ruki-rule, although that's a bit harder to recognize).

>As for Albanian, I can't say much; it's so heavily mutated that it's not
>of much use in reconstructing PIE, and I know next to nothing about the
>sound changes it underwent.  I can say that I've never heard anyone claim
>that Albanian underwent the satem shift.

Albanian is satem, or rather "thatem" (*k^ > th).

=======================
Miguel Carrasquer Vidal
mcv at wxs.nl