suzmccarth wrote:

>
> --- In qalam@yahoogroups.com, "Doug Ewell" <dewell@...> wrote:
>
> > This was where I fell off the trolley, when it turned out that a single
> > script could be one type of writing system or another depending on how
> > it was used. At that point I knew the thing being discussed was not of
> > practical use to me.
>
> I agree. If the distinction between alphabets and abjads is of the
> same order as the distinction between alphabets and morphosyllabic
> systems (Han Chinese) then the classification is not useful. A

You haven't grasped a single thing I've been saying for the past month,
have you.

> script can transform from alphabet to abjad because they are of the
> same fundamental type - they are phonetically segmental systems, as
> is Korean. What will be said about Tifinagh, which has vowels in
> some countires and not in others?

???????

> Syllabic systems are of a different basic type and therefore it is
> possible to make comparisons between Cree, Ethiopic and Japanese in
> terms of coding, keyboard input, psycholinguistic effect and so on.
> Even Chinese is a type of a basically syllabic system and
> phonological processing can be compared between Chinese readers and
> readers of syllabaries.

Do you STILL not understand that your concerns are orthogonal to -- and
irrelevant to -- my typology?

> Tamil has been described by Richard Sproat in this paper
>
> http://catarina.ai.uiuc.edu/L403C/paper.pdf
>
> as being "almost a core syllabary like kana". "Tamil is moving
> towards being an alphasyllabic version of kana." I agree
> wholeheartedly with this.

It's absurd.

> There must be a way to indicate that there are similarities between
> Tamil and other syllabic systems. Tamil is processed like a
> syllabary with very little phonetic segmentation below the syllable
> in the minds of readers. This inhibits phonetic input. Readers of
> syllabic scripts do not develop the same phonemic segmentation
> skills that readers of alphabets do.

That has NOTHING TO DO WITH the relation between graphic and linguistic
substance.

> Therefore, there is a significant population in Japan, Hong Kong and
> China who prefer to input by glyph not by transliteration (Pinyin or
> Ramnji). Tamil needs input by visual sequence (comparable to glyph-
> based input) as an option, because readers of Tamil do not develop
> phonemic segmentation skills to the same extent that readers of
> alphabets do.

SO WHAT??????????? That has NOTHING TO DO WITH the classification.

> There is a significant literature and research on the effect of
> literacy in syllabic languages, which refers to readers of Japanese,
> Chinese, and Tamil. These scripts belong to the same basic group-
> syllabic. See Richard Sproat and Alice Faber. I don't agree with
> their classification entirely but it is at least somewhat useful.

Alice doesn't say anything of the sort, and Sproat isn't the sharpest ax
in the forest.

> I hope that Unicode members will consider these issues and indicate
> that there are two major classes:
>
> A Phonemically segmental systems: alphabets and consonantal alphabets
>
> B Syllabic systems - syllabaries, visually segmental syllabaries
> and morphosyllabaries

It seems clear to me that for Unicode's purposes this is also a useless
dichotomy. They might find that the break between 8-bit and 16-bit
characters is important, for instance.
--
Peter T. Daniels grammatim@...