Criteria for languages?

[Randy Presuhn]

Hi -

> From: "CE Whitehead" <cewcathar at hotmail.com>
> To: <duerst at it.aoyama.ac.jp>; <ietf-languages at iana.org>; <iso639-3 at sil.org>
> Sent: Sunday, December 06, 2009 3:47 PM
> Subject: Criteria for languages?
...
> For possible macrolanguages there's not only gem and de but also gmw.

As a germanist, I'd consider gem or gmw to rather inappropriate choices for
Walliserdeutsch.  I believe de would also be wrong, because even though it is the
most likely fallback language choice, this is due to diglossia rather than a "variety of"
relationship.

This is one of the situations where having a macrolanguage *per se* available
provides limited value.  It's relatively easy for a speaker of German to say
"that sounds like gsw."  And that's a useful bucket for classifying material.

However, there are a lot of mutually more-or-less incomprehensible varieties of gsw,
and a typical speaker of German (or even gsw) isn't necessarily going to be able
to tell you which of these a given sample of gsw is.  (Rather like asking an
inexperienced American to identify the dialects of rural England.)  This is
what makes having a gsw tag useful, not the availability of a form of gsw
that would serve the same function as zh-cmn. Knowing that these are
closely-related languages is interesting, but not always helpful.  When
the need arises for a speaker of one communicate with a speaker of a
non-adjacent one, the likely choice will be de rather than gsw.

Randy