Re: Tonal Languages taken to extremes

In a message dated 10/1/01 12:21:24 PM, and_yo@HOTMAIL.COM writes:

<< That something is recognized orthographically is hardly proof it's
phonemic,
is it? >>

    Certainly not, but in the case of Arabic, it's indicative that those who
are knowledgable about the language (with or without linguistic knowledge)
are very much aware of the fact that no word can start with a vowel.  In
fact, there's a little symbol with the definite article which indicates that,
rather than start with a glottal stop, it can be swallowed by the end vowel
of the last word, if it's not utterance initial.  Of course, this is the only
evidence I have for this.  Other than that, to defend the view that a vowel
can't begin a word in Arabic, I would have to say simply that...it,
uh...can't.  Or, rather, it doesn't.  If an utterance begins with the
definite article (the most likely candidate for a counterexample, in my
opinion), there's usually a very noticeable glottal stop thrown in initially,
including with initial geminates, such as al-lajla (the night), which in the
middle of a sentence would be realized as [l:ejla] (in my Arabic), but which,
utterance initially, would be realized as [?(a)l:ejla] (where that (a) is
what my phonology professor calls a "vowelicle"--the short vowel-like thing
that proceeds liquids), and the glottal stop is actually quite noticeable.
One example that always comes to me is "the study(ing)", which comes out as
[?ad:ir&:sa].  For some reason the glottal stop is much more noticeable to me
when it comes before another stop with little or no vowel in between.

-David