Criteria for languages?

Doug Ewell doug at
Thu Dec 3 05:44:20 CET 2009

John Cowan <cowan at ccil dot org> wrote:

> We have our first problem of constitutional interpretation under our 
> fourth Constitution.  (We are not the Fourth Republic, because we are 
> a monarchy, and a monarchy under a single dynasty -- nay, a single 
> monarch.)
>> 2.  'Extlang' records SHOULD NOT be created for languages if other 
>> languages encompassed by the macrolanguage do not also include 
>> 'extlang' records.
> I interpret the main clause in this sentence as applying if there are 
> as yet no other co-encompassed languages; you and Addison interpret it 
> as not applying in that case.  Consulting my local Talmudist produced 
> a definite maybe.

This looks straightforward to me:

"1. If a language has a macrolanguage mapping, and that macrolanguage 
has other encompassed languages that are assigned extended language 
subtags, then the new language SHOULD have an 'extlang' record assigned 
to it as well.  For example, any language with a macrolanguage of 'zh' 
or 'ar' would be assigned an 'extlang' record."

This rule does not apply to 'lv', since it has no encompassed languages.

"2. 'Extlang' records SHOULD NOT be created for languages if other 
languages encompassed by the macrolanguage do not also include 'extlang' 
records.  For example, if a new Serbo-Croatian ('sh') language were 
registered, it would not get an extlang record because other languages 
encompassed, such as Serbian ('sr'), do not include one in the 

This rule also does not apply to 'lv', since there are no other 
languages encompassed by the macrolanguage (since it was not, until now, 
a macrolanguage).

Since neither rule 1 nor rule 2 applies to a newly designated 
macrolanguage and its newly created encompassed languages, it is 
completely up to the Reviewer, with the usual help of this list, whether 
to admit the proposed 'ltg' and 'lvs' as extlangs.  The criterion should 
be as described in RFC 5645, Section 2.2: the macrolanguage 'lv' should 
be "determined... to have been used to represent a single dominant 
language as well as the macrolanguage as a whole, making the extended 
language mechanism suitable for languages encompassed by the 

As Addison points out, either (a) 'ltg' and 'lvs' both get extlangs, or 
(b) neither one does.

Of course, all of this presumes ISO 639-3/RA takes the requested actions 
on Latvian, and the same decision process (possibly with different 
results) applies to the new Lithuanian subtags as well if those are 

> What is (or rather, will be, if the RA acts) an issue for us is 
> "Should lvs and lgt be registered as extlangs as well as individual 
> languages?" And logically prior to that issue is the issue at hand, 
> namely "Do we have the power to do so without overriding an RFC 5646 

I don't see it this way at all.

Doug Ewell  |  Thornton, Colorado, USA  |
RFC 5645, 4645, UTN #14  |  ietf-languages @ ­

More information about the Ietf-languages mailing list