Suzanne McCarthy <suzmccarth at yahoo dot com> wrote:

> Instead of the evolutionary model of logographies, syllabaries, and
> alphabets, etc. I suggest that there are only two basic types of
> systems.
>
> These are alphabets or analytic systems, and syllabaries or
> wholistic systems. Each of these may encode to a lesser or greater
> degree the morphology of the language. Syllabaries may be non-
> analytic like Japanese and Cherokee, or have an analytic composition
> like Cree, Korean and Tamil.

I don't think you can realistically lump syllabaries and logographic
systems together like this. Characters in logographic systems take on a
life of their own, often independent of the meanings of their
components, while syllables can always be concatenated systematically to
form new words whose pronunciation is obvious.

Additionally, calling Korean Hangul a syllabary is at odds with the
perception of most Koreans, who see Hangul as an alphabet whose letters
just happen to be grouped into syllable blocks. Hangul has discrete
consonants and vowels, unlike true syllabaries such as hiragana,
katakana, and Cherokee.

> While the analytic nature of the syllabaries may be useful for
> technical encoding, these systems are still learned by some native
> speakers as syllabaries. Some members of these language communities
> will have reduced access to digital literacy if the syllabic nature
> of their system is not reflected at some level in the input method.

Even if Koreans read Hangul syllable blocks one block at a time, that
does not make the writing system a syllabary. Peter Daniels and others
have pointed out that fluent readers of English, and other languages
written with alphabets, read clusters of letters at a time.

Character encodings and input methods do not have to be designed
together. Keyboards can be built to bridge any gaps between the
character encoding, the way native speakers view their script, and the
practical limits on number of keys.

-Doug Ewell
Fullerton, California
http://users.adelphia.net/~dewell/