[DGK:]
In my view, both 'six' and
'seven' are thus analyzable within Indo-European and not borrowed from
Semitic. If any borrowing has occurred, these numerals have gone from IE
to Semitic, along with the words for 'horn', 'bull', and 'tower' characterizing
tribes who bred large cattle, used them for draught and as currency, and built
defensive towers.
"
1) If Semitic borrowed these words from PIE, where did the contact happen?
DGK: Presumably any contact between early
IE-speakers and Semitic-speakers occurred somewhere in Southwest Asia. Looking at the Babylonian numerals, however,
I must admit that '6' and '7' do not look like particularly good candidates for
IE loanwords.
No. Masc.
Fem.
1 is^te:n
is^teat
2 s^ina
s^itta
3 s^ala:s^
s^ala:s^at
4 erba
erbet
5 hamis^
hams^at
6 s^edis^
s^es^s^et
7 sebe sebet
8 sama:ne sama:nat
9 tis^e tis^i:t
10 es^er
es^(e)ret
By a quirk of Semitic
syntax, feminine numerals from 3 on up are used with masculine objects, and
vice versa. I suspect the feminines were
originally collectives used in apposition with nouns, which were more commonly
(unmarked) masculine than feminine.
Using (demarked) masculine numerals with (mostly marked) feminine nouns
probably originated as reciprocal analogy.
Hebrew '7' and '8' begin
with /s^/ not /s/, apparently by analogy with '2', '3', and '6'. On the other hand Hebrew retains root-final
pharyngeals, so its masc. _s^éba¿_, fem. _s^ib¿a:h_ '7' show that an original
pharyngeal has been lost in Babylonian.
The triliteral {sb¿} is not at all what we should expect if PIE *septm.´
had been borrowed into Semitic; a quadriliteral *{sptm} (possibly with a
different sibilant) would be reasonable.
Nor does borrowing from Semitic to PIE (championed by Luján Martínez
among others) make any principled sense.
Pharyngeals might very well be phonetically close to one or more PIE
laryngeals, but substituting one for syllabic /m./ (or vice versa) in borrowing
is merely an ad-hoc flight of fancy.
The Bab. forms of '6'
show that both Heb. masc. _s^e:s^_ and fem. _s^is^s^a:h_ have undergone the
assimilation *ds^ > s^s^ (> -s^ in Heb. masc. word-final position). Had Semitic borrowed *swek^s from PIE, one
might expect Bab. masc. *suekis or at best *s^uekis^ (on the basis of
_S^uanna_, a poetic name for Babylon which is likely Pre-Semitic because it is
indeclinable). But again, interchanging
an unvoiced dorsal stop with a voiced dental one makes no principled sense.
On the basis of these
considerations, it appears that the resemblance between Semitic and PIE '6' and
'7' is no more than that, a mere resemblance.
2) Could *swek^s < older *usek^s?
DGK: I cannot think of any other examples which
would make such a metathesis plausible.