Miguel:
>Since there are no Semitic forms *sab`t-u(m) or *sb`at-u(m) with vowel
>deletion, the word must have been borrowed into PIE with three vowels in
>it: *sVbVtVm.
More runaround. You only answered my objection with an "if" statement.
I said _validate_ your empty assumption, not reiterate imaginative
scenarios!
The phonetics of the Semitic word aren't questioned. The question is its
stress.
However, it's next to impossible for a stressed second syllable to be not
heard at all by Pre-IEs, regardless of whatever stress system was in place
at
the time for IE. We don't assume a priori that IE once had three syllables
simply because Semitic happened to have three syllables! Sometimes loans
don't work like that. It's faulty logic. We see _within_ IE that there are
_two_ syllables. Therefore we start with this hypothesis _first_ for Pre-IE.
Thus far, there is nothing _within_ IE to suggest that it derives from a
three-syllable form. Such a possibility has no relevance to its actual
likelihood
anymore than potentially being struck by lightning doesn't stop me from
going outdoors. Nor is it problematic to propose *s�b`atum over *sab`�tum
given the ambiguity of the Semitic data. We just don't ever need that extra
syllable in IE, regardless of whether Semitic had three syllables. And
"extra"
is wasteful hypothesis.
Why explain away a vowel we don't need in the first place?? I have no
objections whatsoever by you claiming that vowels disappear by lack of
stress, but you're missing the point that we don't NEED that vowel to
explain it away! Simply DON'T reconstruct it and your problem is solved.
>The development *s�ptm. > *septm.' is posterior. I have no explanation for
>it. I have nothing against analogy from *ok^t�:(w), I just don't think
>it's a very strong argument. I would prefer something that explains in one
>fell swoop all the other accent shifts (*wl.'kWos, etc.).
Well, until you do, this appears to be the most sufficient "default" theory.
I see *wl'kWos as being due to Acrostatic Regularisation, causing accentual
differences between noun and adjective. Later the accent gets blurred
anyways and some thematic nouns are given final accent like adjectives.
Two different analogical processes at work.
The pairing in numbers is self-explanatory and evidenced by other patterns.
I wouldn't call it a "weak" explanation because this occurs all the time.
We see the pairing of "3" and "4" by their special usage of the plural *-es
(*treies, *kWetwores) and we see *neun and *dekm agreeing in initial
accent. Pairing occurs elsewhere, notably in Japanese but in a different
way (hitotsu/futatsu, mittsu/muttsu, yottsu/yattsu). This also was caused
by analogy.
We both agree that *septm' is innovation on earlier *s�ptm. We need
only explain why the accent shift that we both agree on had occured.
Analogy with neighbouring numbers is the only reasonable and simplest
conclusion. IE *okt�:u provides exactly the answer we're looking for.
= gLeN
_________________________________________________________________
The new MSN 8: advanced junk mail protection and 2 months FREE*
http://join.msn.com/?page=dept/bcomm&pgmarket=en-ca&RU=http%3a%2f%2fjoin.msn.com%2f%3fpage%3dmisc%2fspecialoffers%26pgmarket%3den-ca