Peter_Constable@... wrote:
> On 11/11/2001 06:38:11 AM "Peter T. Daniels" wrote:
> >And that's no different from, and no more revealing than, the typology
> >that was in place from Taylor to Gelb and all those who have since
> >followed Gelb. Sampson, expanding that system, noted that Hangul has
> >structural elements that represent features, viz. the iconicity of the
> >basic shapes and the successive addition of marks in a systematic way
> >(from the point of view of 15th-century Korean phonological theory) to
> >arrive at the other letters. (The vowels, of course, were a lot more
> >regular before 550 years of sound change.)
> After reviewing the facts about the script, it doesn't seem to me that one
> can say that Hangul has structural elements that represent features. If
> that is what they represented, for instance, then a distinction between
> (say) /s/, /t/ and /n/ or /p/ and /m/ should require pairs of structural
> elements. Since these are not represented in the script in that way, the
> structural elements must represent bundles of features.
> There are certainly featural aspects of the script. The iconicity in
> shapes is a featural characteristic, but it is irrelevant for the basis of
> classification / typing that I am proposing. The use of doubling of shapes
> to indicate lenis / fortis (voiced, voiceless) is more relevant, but it
> still isn't consistent with a view that structural elements represent
> features.

You seem to have straitjacketed your brain into a single definition of
"feature" that isn't consistent with that used by Sejong's scholars (or,
as Young-Key seems to still insist, Sejong himself).

You can find the original description in Gari Ledyard's dissertation
(1966); available from UMI for a lot cheaper than if it had ever been a
published book. (I don't remember how much of the 1443 text is included
in his article in Kim-Renaud 1997, U HI Pr.)

> Actually, perhaps there is some analogy here with the abugida (your
> definition) that could be of interest: there are structural elements that
> represent points of articulation with default values for other
> articulatory features; then there are structural modifications (e.g.
> doubling) that correspond to a deviation from those default values. It's
> not entirely systematic, though. E.g. /m/ has the basic shape for labials,
> but /k/ has the basis shape for velars. I don't recall all the details and
> don't have materials with me to study the sound / shape patterns, so I
> can't pursue this further.
> *Perhaps* this may point to a valid structural basis on which to build a
> definition of "featural" that I'd go along with. But a definition for
> "featural" that makes reference to the *shapes* of jamos and their iconic
> relation to points of articulation does not belong in the same taxonomic
> system that otherwise defines types / categories on a structural basis. It
> belongs in a two-way taxonomy of systems for which there are iconic or
> metaphoric relationships between shapes and linguistic objects, and
> systems for which there are not. That is a different taxonomy that is
> orthogonal to one that has classes for things like alphabet and syllabary.

Well, if you think a classification that has separate classes for unique
objects is a good thing, more power to you!

The crucial property of an abugida is the inherent unmarked vowel, which
of course Hangul doesn't have.
Peter T. Daniels grammatim@...