Re: [cybalist] Digest Number 32

[john]
Now this would mean that prior to 3,300 BCE we are talking of
Proto-Semitic languages. And we are suggesting that the Urheimat of
Semitic is somewhere netween the Northern Central and the Southern
Central Zone.

[moi]
I don't agree with this. 3300 BC seems to me much too early a date for proto-semitic. Semitic languages evolve very slowly (look at arabic which is in many regards more archaic than akkadian), because they have so few vowels and then there is the omnipresent semitic root -- the vowels cannot evolve properly, disappear and palatalise the consonnants, unless the root system would be obscured (this is maybe what happened in evolved semitic languages such as gurage).
Akkadian has a very archaic verbal and nominal morphology, but it underwent many mergers. Akkadian is first attested 2500 BC, and I would be quite suprised that in 800 year time it evolved more from proto-semitic than in 2000 years of subsequent evolution. It think a high date - 5000 BC at the very least is necessary for proto-semitic, and much more for the hypothetical proto-Afroasiatic.
The same goes for IE. I think proto-IE should also be dated back to 5000-6000 BC on the very least unless we supose the languages evolved xceptionnaly quickly.

I also think there are many loanword from Semitic to IE and conversely, although it is difficult to say how early it was. My only reference is the book by Gramqrelidze and Ivanov, who also cite loans from sumerian into 'PIE' - although I think the loans were made independently into the sister languages. Among them note the etymon *gwo- 'cow' from sumerian gu, that exist also in tokharian, which suggest it was loaned early into IE.

Guillaume

PS : please avoid using 'samoyed'. Each time I read this word I feel annoyed because it is utterly pejorative. It means roughly 'those who eat each other'. I suggest 'Nganassan-nenets' or something like that. It is like calling the Sami 'lapp' , and the Bristish 'rozbif (rosebeef in French)'