Re: [Ltru] Extlang stability (was: extlang)

First, I don't want to belabor the point about scripts, but it is easy to
recognize when a particular language is written in a script. The harder part
is determining usage. That is different from the language case, where as you
know the boundaries between languages are *FAR* less well defined; as you go
from one village to another you get accumulating differences that finally at
some point you say constitute separate languages.

As to macrolanguage, it appears that we are in general consensus on the
usefulness of having the Macrolanguage field in the registry.

The open issue is whether we also bake it into the extlang structure. I'm
hoping that in the telecon we can make some progress on resolving that
issue.

Mark

On 9/13/07, John Cowan <cowan@ccil.org> wrote:
>
> Mark Davis scripsit:
>
> > I disagree. The base information of whether an instance of written
> language
> > is in a given script is pretty clear, and the number of scripts are
> quite
> > bounded. While it may take some effort to determine whether a script is
> in
> > overwhelming usage for a language, it is a relatively straightforward
> task.
> c
> If it's so straightforward, why has nobody done it?  We have some
> Suppress-Script entries for 639-2 languages, but by no means all,
> and the 639-3-only set is going to make the job ten times bigger.
>
> > On the other hand, the boundaries between languages (languages vs
> dialects
> > and so on) are always the subject of dispute. According to a
> conversation
> > with Ken Whistler, there is a considerable controversy among the
> linguistics
> > community over the language model used in SIL which then went into
> 639-3.
>
> It doesn't matter to us, the IETF; we are committed to doing whatever
> 639 does.
>
> > Now, this is not at all to say that they did a bad job -- they did a
> great
> > job, especially with the huge number of languages that they had to deal
> with
> > -- however, it is inherently a much more difficult task. And the whole
> > notion of macrolanguage is not something that exists "in the wild"; it
> is a
> > construct designed to enable certain usage patterns -- and it is up to
> us to
> > determine whether it fits best in 4646bis as the Macrolanguage: field or
> > hard-wired into extlang.
>
> That's why I'm proposing making the macrolanguage information fully
> available in the Macrolanguage: header, as mutable as SIL decides it
> needs to be, and then using the current state (plus the sign
> language case) for use in the extlang pattern.  We have macrolanguages
> at all because of the vagaries of 639-2, but 639-2 is our core
> language set: we can't just deprecate all those languages in favor
> of obligatorily precise tagging.
>
> --
> No,  John.  I want formats that are actually       John Cowan
> useful, rather than over-featured megaliths that
> http://www.ccil.org/~cowan
> address all questions by piling on ridiculous      cowan@ccil.org
> internal links in forms which are hideously
> over-complex. --Simon St. Laurent on xml-dev
>

-- 
Mark