NSemitic borrowings: in response to Greg Web

[Theo Vennemann]

Miguel Carrasquer Vidal mcv at wxs.nl wrote:

>Upon re-reading that passage, I find we are in agreement on
>Etruscan semph, as well as the Akkadian/IE/Basque/Etruscan
>correlation of 6 ~ 7 with s^- ~ s- (credit for the identification
>of Etruscan s'a with "6", not "4", by an alternative and
>plausible re-interpretation of the Tuscan dice, should be given
>to Beekes and van der Meer).

I am glad you straightened that out. It is so difficult in etymolo-
gicis to trace views to their origin. And who wants to be consid-
ered a pirate? Yet who wants to hush entirely?

>The Basque forms <sei> and <zazpi> are a bit problematical,
>however.  From (East-)Semitic *s^es^s^- and *seb- we would have
>expected to see <ses-> and <zeb-> c.q. <sesi> and <zepzi>, with
>-zi appended as in zortzi "8" and [analogical] bederatzi "9" <
>*bederatzu.  As to <sei>, Larry Trask's "The History of Basque"
>mentions a few cases of sibilant dissimilation through loss
>(*Sanso > Anso "Sancho"), but these affect the first, not the
>second, sibilant.  The <a> in <zazpi> is mysterious, and we need
>a metathesis *zapzi > zazpi, which in itself isn't too much of a
>problem, not in a borrowed item.  I'd like to know what the <a>
>is doing in Welsh saith (< *saxt < *sapt- ??).  Do we know the
>Gaulish or Celtiberian for "6" and "7"?  My guess would be,
>preliminarily, that the Basque forms are more likely to be
>borrowings from Celtic (We. chwech, saith) than directly from
>Semitic.

I am happy to see that you are willing to make those connections,
rather than, e.g., declare them to be "look-alikes". However, trying
to solve the Basque problem by referring to Gaulish or Celtiberian
is a bit like ignotum per ignotius.

Theo Vennemann
3 February 1999