Sign languages (was: Re: additions to ISO 639 and the IANA language subtag registry)

[Doug Ewell]

John Cowan <cowan at ccil dot org> wrote:

> I propose that in 3066ter, the 'sgn' subtag should be treated as sui
> generis: partly a language, partly a macrolanguage.  For the national
> sign languages, it is a language and the country subtag serves to
> discriminate; the corresponding 639-3 code element is suppressed from
> the subtag registry.  For the non-national sign languages, it is a
> macrolanguage and the 639-3 code would be used as an extlang.
>
> I agree that this is messy, but we know that people have used codes 
> like
> sgn-US, and there can be no assurances that they have not used 
> 3066-valid
> codes like sgn-CR even though they were never registered with IANA.

We have no assurances that people have not done lots of things that 
aren't valid.  (NOT to reopen this dead-end thread, but I'm sure 
someone, somewhere has used "en-UK".)

"sgn-CR" is perfectly 3066-valid for "sign languages as used in Costa 
Rica," but if anyone has used it to mean specifically "Costa Rican SL" 
on the basis of Michael's page, they have no assurance that it will be 
interpreted as such by 3066-conformant processors.

> Here's my specific proposal:
>
>    Adamorobe SL  [sgn-ads] (Ghana)
>    Algerian SL  [sgn-DZ] (Algeria) (suppress asp)
>    American SL  [sgn-US] (USA) (suppress ase)
> ...

In cases like Algerian SL, American SL, and so forth, where the sgn + 
region syntax would be syntactically legal, why don't we add *both* to 
the registry and make one Preferred over the other?

Type: extlang
Subtag: asp
Description: Algerian Sign Language
Added: 200x-xx-xx
Prefix: sgn
...
Type: grandfathered
Tag: sgn-DZ
Description: Algerian Sign Language
Added: 200x-xx-xx
Preferred-Value: sgn-asp
Deprecated: 200x-xx-xx
Comments: replaced by ISO 639-3 code asp

"sgn-DZ" could be considered grandfathered because it can already be 
composed from existing subtags, although not with that exact meaning. 
(Again, the whole reason we are having this discussion is that "Algerian 
Sign Language" is not guaranteed to be identical to "sign languages as 
used in Algeria.")

In cases like Adamorobe SL, where the tag "sgn-GH-EP" is not legal under 
RFC 3066bis, we would pretty much have to do as John suggests.  The only 
alternative would be to amend the syntax to allow such stuff, which 
would break just about all of the new 3066bis syntactical features.

I suspect this discussion really belongs on LTRU, and I've cross-posted 
it there.

--
Doug Ewell
Fullerton, California, USA
http://users.adelphia.net/~dewell/