Variant subtag proposal: Høgnorsk variety of Norwegian

Thorgeir Holm thorgeirholm at
Fri Jan 1 23:35:51 CET 2010

<hermer Kent Karlsson frå 01.01.2010 22:57>

>>>> The macro 'no' is simply a political creation, and people fill it
>>>> with whatever ideas they have about this concept.
>>>> It is vital that 'no-hognorsk' be valid, the practical
>>>> circumstances in Norwegian language tagging being as chaotic as
>>>> they are.
>>> Vital?
>> Of course: if 'no' is valid for 'nb' and 'nn', for whatever reasons
>> people might have to tag them thus (and they do), the same should apply
>> to -hognorsk.
> If the tagger knows that the text is in Høgnorsk, then the tagger also
> knows that it is nynorsk, not bokmål. There is no reason to ever use
> "no-hognorsk".

This is not to my point. I perfectly agree that 'nn-hognorsk' would be 
the natural tagging if the other recommendations from this forum were 
followed in practice. However, they are not. It is the practical use of 
'no' today, as we see it, that makes it desireable to have 'no-hognorsk' 
as a valid tag as long as 'no' is possible for 'nn' and 'nb'.

> The only use for the collection codes and "und" in language tagging is
> to mark "partial knowledge". Basically, macrolanguages fall in that category
> too. Again, if you know a document (or a part of a document) is in Høgnorsk,
> then you also know that it is in Nynorsk.

I wonder? Who knows that Høgnorsk is Nynorsk? It is Norwegian, yes, but 
is it a part of Nynorsk, or a separate part of Norwegian? This is a 
matter of definition, and, indeed, neither 'no' nor 'nn' are defined in 
more detail, so it is not to expect that users without special knowledge 
can reproduce this attribution.

> But disregarding that, I find your reasoning here nonsensical. You are
> right that "no" is heritage (but an ISO 639 coding mistake heritage),
> but that in no way implies that "no-hognorsk" should in any way be
> recommended.

My concern is that 'no-hognorsk' should be possible as long as 'no' is 
possible for 'nn' or 'nb'. That makes perfectly sense to me, but I 
didn't see anyone else discussing this particular concern.


More information about the Ietf-languages mailing list