Re: The palatal sham :) (Re: [tied] Re: Albanian (1))

From: Miguel Carrasquer
Message: 30454
Date: 2004-02-02

On Sun, 01 Feb 2004 14:31:35 -0800 (PST), enlil@... wrote:

>However, you mentioned East Semitic and I'm not understanding
>how that works. I would have thought that East Semitic wasn't so far east
>back in Neolithic. Unless you're saying that it was more north at the
>time, say, in Western Anatolia... :)

I mentioned East Semitic and North-West Semitic together (they both share
the development *s > *s^).

To answer the question, I would suspect East Semitic is an early offshoot
of Proto-(North)-Semitic (presumably from the Israel-Palestine area) in the
direction of Central Mesopotamia, i.e. Akkad, where Semitic speakers are
recorded as far back as records go. The expansion to the North (Assyria)
is probably a bit later.

I have no preconceived notions about historical scenarios allowing Semitic
(or East Semitic, or whatever) to be in contact, or not, with
Proto-Indo-European at this or at that period. One should try not to be
influenced by a priori notions of what *not* to expect (and then not look
for it) or what to expect (and not look anywhere else), but to focus on
what the linguistic data alone is saying[*]. My feeling about the
linguistic data in this case is that yes, "6" and "7" are borrowings from
Semitic, or at least, ultimately from Semitic. And that no, it doesn't
appear that the two words were borrowed at the same time and/or through the
same channel(s). The *k^ in PIE *swek^s suggests that the word was
borrowed from North Semitic early (not from the attested North Semitic
forms with assimilation of *s^idc^- to s^is^s^- [Akk.] or /c^ic^c^-/
[Ugaritic]), and that it was borrowed into a (pre-)PIE or through a
language that transformed the cluster /ts(W)/ into /ks(W)/. It was also
borrowed in its "feminine" form (without *-at[u]) and without mimation.
The numeral "7", on the other hand, suggests a later borrowing, more
specifically from East Semitic (because of *s-, not *sw-), and
incorporating the flexional paraphernalia (*-atu, mimation) of the
"masculine" form *sab`atum (Akk. sebettum).

[*] It's a thin line, of course, and some trains of thought can be readily
dismissed because of the impossibility of either the historical scenario
(even if some curious linguistic data exists) or because of the
impossibility of the linguistics (even if the historical scenario itself is
plausible). So yes, a few Austronesian numerals look suspiciously like PIE
numerals, but no, there never were waves of Austronesians handing out words
all over the globe. And yes, Dacian might have survived into modern times
(look at Albanian), but no, Romanian is a Romance language, as anyone with
any notion of linguistics can easily verify.

=======================
Miguel Carrasquer Vidal
mcv@...