Re: [tied] Re: Albanian (1)

On Sat, 24 Jan 2004 01:49:23 +0100, Mate Kapovic <mkapovic@...> wrote:

>I agree. I think I already had a discussion about this with Miguel on
>another list. But here is a possible solution: Uralic is probably the most
>likely to be genetically related with IE. If we look at Uralic vocalic
>system we have there: a; o, u and i, e, ü, ä. And so if
>*ka > *k
>*ko, *ku > *kw
>*ke, *ki, *kü, *kä > *k'
>we see how we could get to Miguels
>*k^ 12+9+6+3 = 30
>*kW 9+3+3+1 = 16
>*k 4+3+2+1 = 10
>One needs not assume the IE is indeed related to Uralic, just that it is
>possible that PIE *k, *k', *kw developed like this (with *k being the rarest
>one becoming only from *ka, *kw second rarest because it is form *ko and
>*ku, and *k' most common because it is from four different sources, or three
>for that matter, it is not important).

The frequency numbers for Proto-Uralic[*] are:

u 28
o 22
a 17
e 15
I 14
i 10
ü 9
ä 8

I don't know where you propose to count the hard vowel /I/, but in any
case, the whole thing yields far too many *kW's. If we count hard vs.
soft, hard vocalism wins by a considerable margin.


[*] Proto-Uralic wordlist from Pekka Sammallahti, "Historical phonology of
the Uralic languages", counting first syllable vowels only (other vowels
are low or high harmonic).

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