639-3 updates

Kent Karlsson kent.karlsson14 at comhem.se
Mon May 14 18:40:23 CEST 2007

John Cowan wrote:
> Michael Everson scripsit:
>> (It seems to me that if one knows what something is, one should tag 
>> it accordingly, and that if one does NOT know what something is, one 
>> should not tag it at all. Tagging it with an "I don't know right now" 
>> tag seems to me to be peculiar at best, and temporary in the long 
>> run. That is my personal opinion.)
> Quite so.  Furthermore, we already have the tag "und" to represent,
> when it is necessary to do so, complete ignorance.  Furthermore,

True. Or better, when allowed, the empty language tag (as opposed to
an absent language tag, which in at least one context means 'inherit').

>we have the language-collection tags to represent partial ignorance,
>as when one knows that a document is in a Chamic language but
>not which one.

That is a conceivable use, but not one currently supported by ISO 639,
nor by RFC 4646. Use of a collection code is not recommended if there
is a more specific code available, according to ISO 639-2. RFC 4646 is
silent on that issue, but if that use is to be allowed/recommended,
that would need to be made explicit (in ISO 639, and/or 4646bis).

		/kent k

More information about the Ietf-languages mailing list