Re: Tibetan Ewts variant proposal (and related questions)

Peter Constable wrote:

> Why do we use, say, az-Latn and az-Cyrl rather than az and az-t-...?
> Or why not zh and zh-t-...?

We assume there is a single, more-or-less standard "native" way of
writing Azeri in Latin, and a single "native" way of writing Azeri in
Cyrillic. Same for written Chinese in Simplified vs. Traditional.

If more than one "native" variety exists for a given script, to the
extent they need to be tagged separately, then this calls for variants.
This is what we did with Belarusian. Neither '1959acad' nor 'tarask'
represents a transliteration; both are native ways of writing Belarusian
in Cyrillic.

Tibetan in Latin, even for newly created text, is not in a "native"
writing system. In this case, one is expressly using a scheme that was
developed to convert from the Tibetan script to the Latin script. This
is a transliteration, even if the text did not originally exist in the
"source" script.

> Tagging for transliteration is most appropriate when there is a
> conventional orthography in script A but the content is written in
> another script, B, that is not traditionally used by the primary user
> communities; and when there is a particular convention in use for
> conversion from script A to script B.

Yes. That is one reason why I believe Extended Wylie and other
transliteration schemes should be extension subtags under the T
extension instead of variants. Another is that, once the T extension was
approved in 2011, we stopped registering variant subtags for
transliterations (not orthographies and phonetic alphabets), and it
seems arbitrary to start again now.
 
--
Doug Ewell | Thornton, CO, US | ewellic.org