A stake in the ground

[Randy Presuhn]

Hi -

> From: "John Cowan" <cowan at ccil.org>
> To: <ietf-languages at iana.org>
> Sent: Wednesday, December 09, 2009 9:57 PM
> Subject: Re: A stake in the ground
>
> I finally found the original criteria that LTRU used to decide which
> macrolanguages were made usable with extlang tags.  The details are
> at http://www.ietf.org/mail-archive/web/ltru/current/msg08904.html ,
> but in summary, Peter Constable divided macrolanguages into seven groups:
> 
> 1) Macrolanguage tag is deprecated.
> 
> 2) Macrolanguage code has been widely used; a single encompassed language
> is dominant.
> 
> 3) Macrolanguage code has been widely used; no encompassed language
> is dominant.
> 
> 4) Macrolanguage code has been widely used; more than one encompassed
> language is dominant.
> 
> 5-7) Macrolanguage code has *not* been widely used; same division as 2-4.
> 
> LTRU decided to place languages encompassed by macrolanguages in group
> 2, and only those in group 2, into the registry with extlang subtags.
> It seems clear to me that of the five new macrolanguages being proposed,
> Latvian and Lithuanian fall into group 2, Blang and Central Bontoc fall
> into one of groups 5-7, and Low German (if that even flies) into group 3.
...

The discussion focussed on cases of group 2 because they provided the
compelling arguments for retaining the macrolanguage subtag.  Groups 3
and 4 didn't get any significant discussion as far as I recall.  The distinction
between group 2 and either group 3 or 4 is useful only if one heads down the
path of treating the macrolanguage code as a synonym for the dominant language
variety, a path repudiated by the WG, even though there are applications where
this kind of under-specification heuristic can be useful.  Consequently, I'd be very
reluctant to claim that the distinction between 2 and 3 or 4 should carry much weight.
The sense of the discussion was clearly that the deciding factor was the existing
widespread use of a code that was later given macrolanguage status.

Randy