Re: [OT] ANN: Site about scripts

From: Lars Marius Garshol (larsga@garshol.priv.no)
Date: Fri Oct 12 2001 - 13:58:44 EDT


* Kenneth Whistler (in an earlier posting)
|
| 2. Script B is a de novo design influenced strongly by Script A.
|
| 3. Script B borrowed formal and/or functional characteristics of
| Script A.

* Kenneth Whistler
|
| Yes. Xi Xia is a good example of a de novo design influenced strongly
| by Han. Somebody sets out to create a script for a language that
| had none, and deliberately clones concepts from another script
| to do so. [...]
|
| The kind of thing I am thinking about for 3 could be exemplified
| by the modern Tai Le orthography. It is an evolutionary descendant
| of earlier SE Asian Brahmi-based scripts maybe 800 years ago.

Does anyone know what the nearest ancestor is, BTW? I've put Tai Le
into the site, but would like to connect it into the Brahmi family
tree.

| But the 1954 orthographic reform introduced the systematic
| representation of tones with combining diacritics. And the
| particular combining diacritics are basically a structural loan from
| Latin (as opposed to Thai, which *could* have been taken as the
| model, but wasn't).

Unless I am missing something both 2. and 3. involves a cloning of
concepts, and the difference is that in 2. the design is new, while in
3. it is not.

So it seems these situations can be captured in this way:

  2: Script has no derived-from relationship, but at least one
     influenced-by relationship

  3: Script has a derived-from relationship, and zero or more
     influenced-by relationships

Am I right? (I'm asking because I need to be able to express this in
terms of topic maps, and your original formulations do not work
entirely inside that model.)

* Lars Marius Garshol
|
| I feel that some device is needed to cluster the scripts into groups
| in order to explain how they are related to one another historically.
| If you can think of a better solution I am certainly open to
| suggestion.
 
* Kenneth Whistler
|
| One obvious device you are missing is geographic area. That is the
| mechanism we use to help organize the Unicode Standard. It is
| obviously related to historic connections, but is not everywhere the
| same, and is an easier organizing principle for most people to start
| from.

This is a very good point. Geographic regions are obviously easier for
people to relate to, and the descendant relationships are still there.
I should probably switch to this as the main categorization system,
and either change or get rid of the existing category system.

Thank you very much for all this help!

--Lars M.



This archive was generated by hypermail 2.1.2 : Fri Oct 12 2001 - 12:32:05 EDT