Re: Why ISO language codes shouldn't be trusted

> I think that also linguistics community have the
> responsibility to help
> as much they can international standardization
> bodies like the ISO

Sadly, linguists (at least the linguists who I know)
don't pay much attention to the ISO language codes.
Neither of my two friends who are linguists in Bolivia
had even heard of the ISO language codes. My guess is
that most linguists don't bother looking at the
websites for SIL International, Ethnologue.com, and
the ISO. It is only when you try to use different
languages in software that the ISO codes become an
issue--the linguistics field certainly doesn't use
them.

> I'm going to commit soon the languages that have the
> ISO code until Amos
> gets back to us and we make a decision about this.

Jordi, I have submitted forms to SIL/ISO. They might
change the codes in January 2008, but they also might
wait for years. So my suggestion is to use the ISO
language codes that are available today and don't wait
for the SIL/ISO.

The Bolivian government is eager to start using
AbiWord in Quechua and Aymara today. They are
currently building 1000 telecenters in rural
communities (about 350 are already up and running) and
they keep asking me when they can start installing
AbiWord in their telecenters. The Bolivian Ministry of
Education has already handed out 400 CDs with AbiWord
in Aymara, but they currently are using Northern
Aymara by selecting "Spanish" from the language list
and Southern Aymara by selecting "Italian", and they
can only these languages by replacing the strings
files for Spanish and Italian. People in Bolivia keep
asking me when they can download AbiWord in Aymara and
install it without fiddling with the strings files.

Likewise, in Peru, they are making plans to buy the
educational laptops from the One Laptop Per Child
program, and they are very interested in using AbiWord
in Quechua and Aymara. I talked with the director of
Bilingual education in the Peruvian Ministry of
Education, and he told me that they might distribute
AbiWord for Windows in their training packs for
Bilingual teachers.

So there is serious demand here in the Andes to use
AbiWord, but only if it comes in their native
languages.

So let me reiterate, my earlier suggestion to get
around the ISO code problem:

http://www.runasimipi.org/abiword/ayn-BO.po
becomes "aym-BO.po" and is listed as "Aymara (La Paz)"

http://www.runasimipi.org/abiword/ays-BO.po
becomes "ayc-BO.po" and is listed as "Aymara (Oruro)"

http://www.runasimipi.org/abiword/qu3-BO.po
becomes "quh-BO.po" and is listed as "Quechua (3
vowels)"

http://www.runasimipi.org/abiword/qu5-BO.po
becomes "qul-BO.po" and is listed as "Quechua (5
vowels)"

(Our previous Quechua translation file will also have
to be removed, which is "qu-BO.po" and was submitted
when AbiWord was in version 2.10)

Hopefully this is an acceptable solution.

Cheers,
Amos

--- Jordi Mas <jmas@softcatala.org> wrote:

>
> Hello Omer,
> > I read with interest the account of the various
> Aymara languages and
> > dialects, and how ISO got them wrong.
> >
>
> I think that also linguistics community have the
> responsibility to help
> as much they can international standardization
> bodies like the ISO to
> get the facts properly. Unfortunately, there is
> always a gap between the
> standard and reality.
>
> > However, I disagree that ISO language codes should
> not be relied upon to
> > decide whether a language is included in AbiWord
> or not.
> > The AbiWord developers are not, in general,
> linguists. So they need to
> > follow some acceptable authority. ISO is a
> reasonable choice of
> > authority. So is the Unicode consortium (when
> dealing with glyph shapes
> > and properties).
> >
> > However, we do need few mechanisms:
> > 1. A mechanism for integrating into AbiWord
> modifications to ISO
> > language codes (like the mechanism used by the
> Unicode consortium to
> > export updates - by a set of data files, which can
> be semi-automatically
> > incorporated into projects which rely upon them).
> This mechanism will
> > also include a mapping from old to new language
> codes, so that
> > applications will know how a language+dialect was
> renamed.
> > 2. An escape mechanism for adding languages and
> dialects not currently
> > blessed by ISO. Such a mechanism will accommodate
> also languages and
> > dialects, which are under dispute even by
> linguists.
> > And it would be cool to have also a Klingon
> translation for AbiWord, if
> > one does not already exist.
>
> In my opinion, and as you point out, we are
> developers not
> linguistics.The complexity of the situation of some
> languages in some
> regions is even difficult to clarify even for
> linguistics, since
> languages are moving targets and they change
> everyday. That is the main
> reason why I'm in favor of sticking to the ISO
> codes.
>
> I do not like the idea of modifying an already
> existing standard
> (modification to ISO language codes). Standards are
> a common agreement
> between people, if you modify them unilaterally they
> lost their point.
>
> Extending an already existing standard is another
> issue. Alan for
> example was proposing to use the @ sign to extend
> the ISO standard for
> non local dialects not recognized. This looks
> reasonable to me (even if
> my favorite option is to stick to the ISO codes).
>
> I'm going to commit soon the languages that have the
> ISO code until Amos
> gets back to us and we make a decision about this.
>
> Jordi,
>
> --
>
> Jordi Mas i Hernàndez, HomePage
> http://www.softcatala.org/~jmas/
> Bloc personal http://www.softcatala.org/~jmas/bloc/
> Planeta Softcatalà:
> http://www.softcatala.org/planet/
>
>
>

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around
http://mail.yahoo.com
Received on Fri Oct 19 02:20:23 2007