At 3:19:19 AM on Tuesday, September 30, 2008, Arnaud Fournet
> From: "Richard Wordingham" <richard.wordingham@...>
>>> Sorry but I clicked on a link that does not provide any
>>> data, just the reconstructed form. If I missed
>>> something, I'd be glad to get the right reference.
>> Brian posted the link -
> Ge'ez forms here contradict the forms on Zompist.
The correct form, according to Patrick R. Bennett,
_Comparative Semitic Linguistics_, p. 35, is <s^ala:stu:>,
the words for '1', '2', '4', and '5' being respectively
<'ah.adu:>, <k&l'e:>, <'arba:`&ttu:>, and <xamm&stu:>. He
gives the PSem. (in numerical order) as *'ah.ad-, *þn-a:-,
*s'ala:þ(-at)-, *'arba`(-at)-, and *xams^(-at)-.
Mark uses <?> for <'>, barred-i for <&> (which I believe is
phonetically correct), and <sh> for <s^>, and he fails to
mark the length of final <-u>, but otherwise his forms are
in complete agreement except for '3'. Possibly his is a
more colloquial form, or possibly it's simply an error, but
I'll go with Bennett.
> Considering the fact that the root is *t_l,
'Fact' does not mean 'idiosyncratic opinion of a rank