Jens on the absence of laryngeal in *septm:
>The only hard fact in this seems to be that it wasn't. I am not too
>shocked by that, for I have never assumed that any of the three IE
>laryngeals was a voiced pharyngeal spirant. I would rather bet on
>velar and labiovelar point of articulation for H2 (which was
>voiceless) and H3 (which was voiced), and for plain [h] for H1.
While I still can't agree with *h3 being voiced (more like [hW]), this
is hardly the point. The point is that there are evidently ample
phonemes to choose from in IE that could approximate Semitic
ayin, rather than merely ignoring the phoneme, even if the
approximation isn't perfect.
If "the only hard fact in this seems to be that it wasn't" heard by
IE speakers, it is more likely that this failure to hear the said phoneme
is because it was in a less audible, hence _unstressed_, syllable.
So either *s�b`atum or *sab`at�m would make sense here, but
*sab`�tum is so much more unlikely based on this absence of
laryngeal in IE alone.
Yet another obvious factor is that IE evidently shows no vowel
between *p and *t, let alone accent. To propose Pre-IE *-pVt- is
baseless (not based on any adequate and well proven Pre-IE rule)
and is just unnecessary assumption. Proposing that the accent
was initial however doesn't require any make-believe syllable
conveniently gone missing.
Other factors include *septm's original accent being on the first syllable
as indicated by the zero-graded second syllable. Zero-grading is
generally caused by accent shifts. While the rule isn't always true,
it happens enough in IE for many linguists to remark that zero-
grading was most probably caused by accent. So, *septm would
appear to be shaped in such a way that the accent must have
originally been on the first syllable, just as we see by the shape
of the word *wlkWos that the original accent was on the second
syllable. Not only is the second syllable zero-graded but the first
syllable in *septm is full. Analogy with accent in *okt�:u is logical.
So again this shows strongly that the accent was initial at one
time, regardless of a "lack of attestation". In this case, the word
shape itself is the attestation.
So we have three straightforward factors within IE that indicate
that the accent most likely wasn't on the second syllable in
*sab`atum but on the first at the time of the loan -- Absence of
expected laryngeal, the extra assumption needed to propose
an extra syllable *-pVt-, and the very word shape of *septm.
To add, not only are the above points very valid and hard to
object to, but since initial accent is unproblematic in both language
groups, I especially don't see why you are all trying to avoid the
path of least resistence. At least I've been basing my ideas on
general pre-IE rules which are evidenced elsewhere in the language,
like Penultimate Accent, for example. It's better than proposing ad
hoc rules for this one special case as you seem to do.
= gLeN
_________________________________________________________________
The new MSN 8: smart spam protection and 2 months FREE*
http://join.msn.com/?page=features/junkmail
http://join.msn.com/?page=dept/bcomm&pgmarket=en-ca&RU=http%3a%2f%2fjoin.msn.com%2f%3fpage%3dmisc%2fspecialoffers%26pgmarket%3den-ca