On Fri, 06 Feb 2004 10:01:14 +0100, Miguel Carrasquer <
mcv@...> wrote:
>Perhaps we can compare the situation in Berber (see
><www.zompist.com/numbers.shtml>). Proto-Berber is assumed to have had a
>system with numerals 1-5 and 10, with 6-8 expressed as 5+1, 5+2, 5+3, and 9
>expressed as 10-1. Later 6-9 were replaced (although there's nothing to
>indicate that this didn't happen all at once) with numerals borrowed from
>what looks like an early form of Semitic: sed.is, ssa(b), ttam, tz.a.
It's interesting to consider what the source may have been within Semitic.
For historical reasons, NW Semitic (Phoenician) seems the most likely
source. The initial sibilant in 6 and 7 is not distinguished in Berber, so
that doesn't provide a clue. But the fact that Berber has s- instead of t-
in the word for "6" is somewhat of a difficulty, given Ugaritic t_t_.
Hebrew/Phoenician s^e:s^ is ambiguous (it can represent assimilation
t_...t_ or s^...s^). Aramaic s^is^, however, must go back to an
assimilation s^...s^, so there can be no objections against a NW Semitic
form *s^ids^- (besides *t_idt_- as reflected in Ugaritic). The word for
'8', ttam, shows that when the word was borrowed in Berber, the (West)
Semitic form must already have been *t_ama:n- (Ugar. t_mn, Aram. t&ma:ne:,
Hebr. s^emo:ne:), not *c^ama:n-. Another indication of NW Semitic origin
is final -a in tz.a < *tis`a, with the typical NW Semitic loss of
nominative -u in favour of accusative -a. The non-assimilation of -ds^-
(*-dc^-) in Berber sd.is, as opposed to assimilation in recorded NW
Semitic, is the most surprising thing in the Berber forms. It suggests an
early (2nd millennium) NW Semitic dialect, but distinct from Ugaritic. I
don't know if any of the Old Canaanite numerals are known.
=======================
Miguel Carrasquer Vidal
mcv@...