Whoops, almost missed this one:
Sergei:
>Ivanov and Gamkrelidze state that /b/ in PIE *ablu- 'apple' is >eventually
>from [m] (they mean borrowing from something like
>substratum(Hurrian?)/migrating *Samlu- 'the same'). If they are >right, and
>PIE speakers turned [m] into [b], why could not they turn >*m-l-? into
>*bVlH and then, rejecting unnatural *b, into pVlH?
Yes, "apple" is supposed to be from an *-m- but let's re-reconstruct that
PIE root, shall we? First, there is no evidence of IE *b, let's just accept
that little fact and stop trying to fight against it. So, the root should be
more accurately reconstructed as PIE *amlu-. The odd cluster *-ml- naturally
would yield *-bl- and then *-pl- in Germanic, without the need of
reconstructing **-b-. To substantiate this, *s-amlu- is attested in
Anatolian. The origin of *s- is unclear but it would seem to be somehow
related to *s-less forms found in Turkish refering to fruit. I believe it
was /alIm/? I'm too lazy to check in a dictionary. :( Exactly what language
it comes from is something I don't know. Is there actually a similar Hurrian
form?
As for, *m-l-?, even though IE has no *b from time immemorable, it's still
very conceivable for a Semitic infinitive *mala:?u, with accent on the long
vowel, to be misheard in an early form of IE as *pele?e (accent on second
*e, where *e equals "schwa", later becoming *ple?-) without there being an
intermediate phoneme *b in IE, due to the same clash of *m and *l as we see
in "apple".
- gLeN
______________________________________________________
Get Your Private, Free Email at
http://www.hotmail.com