Re: [tied] Re: Syncope

On Tue, 30 Mar 2004 enlil@... wrote:

>
> Jens:
>
> > Still, that is *very* far from constituting proof that
> > the selection of "parallels" is adequate for the history
> > to be uncovered here.
>
> It demonstrates that my rules operate under a strong sense
> of linguistic normalcy overall. This is the effect we
> want when reconstructing a language properly unless
> for whatever reason more bizarre rules are needed to
> explain things. However, more uncommon rules should be the
> exception, not the norm. O-fixing and double-long vowels
> seem to seep into all aspects of your pre-IE, making it
> more bizarre than it is normal because they are much
> less common processes. With O-fixing, I don't even
> know what real-world language operates like this and
> I can only think of Lapp and Mingo as having double-long
> vowels.

*Must* you apply name-calling to the infix theory for some of the IE o's?
You are apparently trying to evoke a sentiment of idée fixe or some
junky-like fix, just to discredit the idea. I would like you to restrain
your disrespectful attitude a bit. I am trying to be very serious.

The inixal o has no exact parallels that I know, but metathesis is not
unheard-of, and that is all it takes to get the element in position.
Vocalisation of some back spirant to a vowel of back articulation is not
strange either, nor is its chromatization to /o/. It is only the condensed
story of all these events in a row that escape your search. If the element
concerned was a uvular spirant (which is not known, but quite strongly
suggested by a number of facts), the result could well be coarticulation
extending the uvularization to the entire initial cluster, which could
easily lead to the development of a uvular off-glide after the cluster
which would later survive alone.

> > You can find other languages that do different things
> > with tri-consonantal clusters; [...]
> > How did you decide that none of all that constitutes
> > a parallel to your pre-PIE?

By observing that IE is different from what those "parallels" would have
given.

> It turns out that a CV(C) syllable structure works best
> for MIE because of the logical path I take, starting
> with the observation of Syncope and following it to QAR.
> We haven't begun to adequately talk about QAR because
> we can't seem to agree on Syncope yet.
>
> However, I think that there is a good case to propose a
> similar CV(C) syllable structure for Semitic. Observing
> loans between MIE and Semitic help us see how both languages
> worked in more detail at that time. Semitic accent is
> turning out to be cool too. You say that "six" in Semitic
> doesn't have a vowel at the end of it in the non-mimated
> feminine form but it would violate CV(C) and would rob
> us of the necessary case determining vowel important in
> Semitic morphology. Double no-no.

Which can be neutralized by a third no: You cannot insist on analyses that
are only derived by rule if you can just as well adjust the rules. Can't a
language be "almost-cool"? Can't the general impression of Semitic
morphology be that referring to a oyunger time than the borrowing? The
answer is plainly yes to both questions.

> So I think *s^idc^u as
> the form cited to MIE speakers is more sensible over
> *s^idc^.

That you think so is not reason enough.

> So we should accept that the MIE speakers were
> inputted the word _with_ a final vowel. What they did
> with this final vowel is still a debate with us, but
> it's simpler to conclude that they just accepted the
> final vowel and didn't do anything to it until Syncope,
> hence MIE *sWeksa.

It is at least just as simple to assume that *swek^s has no final vowel
because the donor language had no final vowel in the original form. But I
am in pretty deep water here.

Jens