--- In qalam@yahoogroups.com, Marco Cimarosti <marco.cimarosti@...>
wrote:

> Chinese shape-based method are based on a the *pre-existing* methods for
> analyzing Chinese characters in smaller units: either some hundreds
> "components" (such as the well-known "radicals"), or a handful of

"strokes".

> By "pre-existing" I mean that these methods have been invented

centuries ago

> (mainly for lexicographic purposes, i.e. sorting dictionaries), and that
> knowing these methods of analysis is a (hard-acquired) part of Chinese
> literacy.
>
> What Chinese shape-based input methods do is simply to *exploit* this
> pre-existing skills that Chinese user acquire in school, and they do

this by

> mapping the available keys to these elements. On a computer

keyboard, keys

> are loosely mapped to "components"; on a cell-phone keypad, they are

mapped

> to strokes.
>
> Going back to Liberia, AFAIK, no corresponding shape-based analysis ever
> existed for the Vai script. So, again, what kind of "widgets" would

you map

> to keys? Mind that, whatever these 30-odd "widgets" are, you cannot

expect

> potential users to learn them and how they map to Vai characters just to
> learn your input method: they must something that the users

*already* know.

This is one reason to be interested in the reported 60-character core
set of Vai characters. A lot of characters are formed as
modifications from others by means of unsystematic diacritics,
basically double dot and bar. One description of Vai describes double
dot below as marking nasalisation, but that is actually the one place
the double dot is almost never placed - unless it's actually the place
to mark the rare and unattested nasalisations. KPEN from KPE would
be the prototype of this exception, as opposed to regular examples
such as MGBE, which is KPE with an overstruck double dot, just as GBEN
is with an overstruck double dot. (I'm using the proposed Unicode
names to designate the *characters*, and not the sounds they
represent.) Adding a bar tends to indicate voicing. Sometimes the
diacritics just mean similar but different.

Historically, there has been variation in the meaning of base form
plus a diacritic - presumably one reason for having the 1962
standardisation!

Richard.