Re: Too Many Too Little Possible Roots!!!

Matt wrote:
>
>Malagasy, an Austronesian language with African step-parents, is
>an interesting example of the kinds of pendulum swings you're
>talking about. Proto-Malagasy, like Malay and the Philippine
>languages, allowed word-final consonants and word-medial clusters
>(CVC syllables, in other words).  But then, probably under the
>influence of the Eastern Bantu languages, it developed word-final
>and word-medial epenthetic vowels, leading to exclusively CV
>syllables.  Compare the Tagalog word for moon, /buwan/, with its
>Malagasy cognate /vulana/. The ancestor of these words was
>probably something like /bulan/, CV-CVC but in Malagasy an /a/
>was added after the final nasal to give CV-CV-CV.  So here we
>have simplification of syllable structure leading to longer
>roots.
>
>In contemporary Malagasy, however, certain unstressed vowels are
>normally devoiced in rapid speech, and in some environments
>almost completely disappear.  Thus for instance the word "olona"
>/uluna/, which is stressed on the first syllable, is normally
>pronounced more like /uln@/ or even /uln/.  If radical devoicing
>persists in the language, these unstressed vowels might
>eventually disappear altogether, leading to shorter roots but
>more complex syllables (CVC, CVCC, etc.).
An extremely interesting example, Matt, to convince me all-the-more
of my theory. I'm beginning to become more and more convinced that
there must be some *ideal* amount of possible roots that natlangs
attempt to obtain/maintain. You're a proffesional linguist, right?
So you wouldn't by chance know what this secret number is? Come on,
let us all in to your secret!! I know you know!! 8-D Unless I can
figure out or in any other way get to know what this number is, I
think I'll never be completely satisfied with the structure of
Lumanesian roots.

Regards,
-Kristian- 8-)