Mike S:
> Thus, in most cases, I would have to say the phonemic system is probably
> optimum;
> except for languages with very simple syllable structures, I think the
> simplicity and
> efficiency of the phonemic system easily trumps all contenders.  If you are
> inclined to
> think this is mere bias, consider this: many conlangers have designed their
> own alphabets,
> but how many have designed syllabic sets?  If anyone *has* designed a
> complete
> syllabic system, I'll bet my hat that it implements markers or some similar
> regular device
> to correspond directly to final nasal, vowel length, or some other
> phoneme-level distinction.
> Possibly without knowing it, they are, in fact, conceding the superior
> efficiency
> of the phonemic system.

My previous message outlines some of the reasons why I favour syllabaries.
The Livagian syllabary is quasi-featureal by design; so rather as the
tengwar's shapes provide information about the phonological features of
the segment, so a character of the Livagian syllabary provides, in its
shape, some information about the phonological features of the segments
that compose it. At present the script is in a mess, with no form--
signification correspondences currently established, but in its earlier
(and probably also its future) state, it was not strictly possible to
analyse the script as based on a set of more primitive featural elements
involving a constant form--signification correspondence.

On another point, I would in a conlang want to reject a phonemic script
because I reject the very notion of the phoneme.

--And.