ISO 639 JAC decision re mo/mol
doug at ewellic.org
Tue Nov 4 06:10:49 CET 2008
"Lang Gérard" <gerard dot lang at insee dot fr> wrote:
> I completely agree with Doug's surprise !
> Moreover, the alpha-2 ISO 639 code element "mo" for the language name
> "moldavian" was created with the publication of ISO 639 (1988), and
> the alpha-3 ISO 639-2 code element "mol" for the same language name
> was created with the publication of ISO 639-2 (1998).
To be fair, I wasn't expressing surprise at the deprecation of a code
element assigned 20 years ago, but of one that had just been modified
(i.e. maintained) five months ago. I suspect that a decision to
deprecate a code element is not taken lightly, and that the possibility
of deprecation might be under discussion for several months before a
decision is made. Given that, I would think any updating of the names
associated with the code element would be put on hold until the question
of deprecation had been resolved.
> Between this time, Moldova became an independant state, and the
> Constitution adopted on 29 july 2004 states that:
> "The language of the State of the Republic of Moldova is the moldovan
> language, written with the Latin script".
They can say that if they want. It would also be possible for the
United States Constitution to refer to "the American language," or the
Mexican Constitution to refer to "the Mexican language." That does not
make them distinct languages from English and Spanish according to the
criteria of ISO 639.
It's completely understandable for a newly independent country to want
to take steps to demonstrate its independence -- that's why the U.S. had
a national mint before it had a federal constitution -- but
unfortunately for Moldova, their assertion of a "Moldovan language"
distinct from Romanian is not shared by most professional, non-partisan
> How is it possible that this language name, that passed all criteria
> to be recognized inside ISO 639[-1] and ISO 639-2 during 20 or 10
> years, has now to be deprecated ?
Better judgment? Re-examination of the facts?
In any case, it's the code element that is being deprecated, not the
language name. Håvard specifically stated that the names "Moldavian"
and "Moldovan" were used to refer to a "variant of the Romanian
language," and in fact it would be reasonable to assume that these might
be added to ISO 639 as alternate names for 'ro' and 'ron/rum'.
> Maybe, but this does not explain why "mo" and "mol" have been created
> and used during this time. In particular, if "mol" has been created,
> there must be some terminological or bibliographical entries (notably
> at the LC ?) having used this code element ? Because "mol" was created
> in 1998, after the 1994 Constitution of Moldova.
Håvard did say that "recordings using these identifiers will not be
invalid." The same is true for uses of 'mo' in the Language Subtag
Registry: subtags are never removed, only deprecated, and not only will
material tagged with 'mo' retain its meaning, but users can even
continue to tag material as 'mo', though it will not be the preferred
Ihar Mahaniok <mahaniok at gmail dot com> wrote:
> How is the official language of Transnistrian Moldovan Republic to be
> identified from now on?
> They use Moldovan in Cyrillic script, and both people and authorities
> of this de-facto independent territory would definitely object calling
> this language "Romanian".
The ISO 639 code element for "Romanian" should have the alternate names
"Moldavian" and "Moldovan," as mentioned above. This is just like the
existing code element 'es' which has two names, "Spanish" and
"Castilian." Most Spanish-speaking people in Spain would refer to their
language as "castellano," not "español." I would certainly encourage
these names to be applied to 'ro' in the LSR. Using the subtag 'ro'
will not make the Transnistrians into Romanians, any more than using the
subtag 'fat' will make speakers of Fanti fat.
> the fact is that the content in this language is not going to be
> tagged with this tag, as creators of this content won't do it.
They should, if they want their content to be identifiable and
searchable. "ro-Cyrl" might help distinguish Moldova-dialect content
from Romania-dialect content. But calling a single language by two
different names doesn't make it two different languages, no matter how
strong the nationalistic fervor may be.
Doug Ewell * Thornton, Colorado, USA * RFC 4645 * UTN #14
More information about the Ietf-languages