I'm just reading a colleague's recently-completed MA thesis on electronic representation of writing system descriptions, and that led me to an insight that pertains to this recent thread:

Michael Everson wrote in response to Lars Marius Garshol:

>>| Yes, it's my principal point that Hangul is an alphabetic script
>>| because Jamo is an alphabet.
>>I can sympathize with that point of view, and certainly agree that
>>Jamo could have been used as an alphabet like all the others. That is
>>not how it is used, however.
>Of course it is. The principle feature of an alphabet is that it has
>symbols denoting consonants and vowels. Linear presentation of these,
>whether horizontal or vertical, is not the underlying feature.
>Syllable clustering of alphabetic letters in Hangul is a typographic
>feature of the script. It takes nothing away from its alphabet-nature.

As we all know, in Daniels' classification, he put Hangul in its own class, calling it featural. Now, there is a disagreement between Daniels and Bright over Indic scripts: Daniels considers Devanagari (e.g.) to be an abugida. Bright, on the other hand, considers them to be alphasyllabaries, a classification which Daniels rejects.

In reading my friend's thesis, I was reminded of DeFrancis' argument that writing systems can represent language only on the levels of the syllable or the phoneme. It struck me then that Hangul is a prototypical alphasyllabary.

I find some other interesting aspects to this. Daniels and others argued that study of writing systems needs to examine, and only needs to examine, the relationship between written symbols and the linguistic forms that they represent. This was in opposition to an earlier structuralist view that said that the written form could be studied independently of any relationship to linguistic forms. It has seemed to me that, in fact, study of writing systems needs to consider both structure of written forms and relationships to linguistic forms.

Applying this to Hangul, the jamos represent things on the level of the phoneme (further comments below). But then we can describe purely structural characteristics of the writing in relation to how jamo forms get arranged into cells. That on its own is an interesting structural characteristic of Hangul writing, shared by SignWriting. But then we get a relationship again to the relationship to linguistic forms since the cells correspond to syllables.

As to the relationship betweens jamos and things on the linguistic level of phoneme, Martin Duerst wrote in response to Ken Whistler:

>>Hangul is structured from an alphabet (the jamo). That alphabet is
>>so tightly coupled to the phonology of Korean that it can be
>>considered a phonemic alphabet -- it is very regularly related to
>>the sound of Korean.
>Oh well. I guess you never heard about all the liaisons
>and things that go on with final consonants and consonant
>clusters. Hangul was order-made for Korean, but Korean has
>changed. In many ways, it's similar to the current state
>of French. It's definitely not as nice as Finish or Italian,
>but not as bad as English.

DeFrancis argues that writing systems should be distinguished according to the basic units that they relate to rather than precisely what each written unit represents. I think there's no question that jamos represent language on the level of the phoneme. There is an orthogonal characterisation of writing systems, discussed by Sproat and I'd expect earlier by others, as to how concrete or abstract the relationships between written forms and the linguistic forms they represent. Thus, Spanish and English writing systems both represent their respectic languages basically on the phonemic level, but Spanish does so in a fairly concrete way while English writing has become abstract as the result of several hundred years of significant phonological changes. It shouldn't surpise us that Hangul has become more abstract over several hundred years -- what is surprising is that it hasn't become more so. I think there's no question that jamos represent t! hings on the phonemic level, and that they still do so in a relatively concrete way, though as Martin points out something more concrete than English but less concrete than Italian.

- Peter

Peter Constable

Non-Roman Script Initiative, SIL International
7500 W. Camp Wisdom Rd., Dallas, TX 75236, USA
Tel: +1 972 708 7485
E-mail: <peter_constable@...>