On Mon, 6 Mar 2000, [iso-8859-1] Keld Jørn Simonsen wrote:

> For library use, I don't understand that it is not important to
> record the language. Bokmål and Nynorsk is quite different,
> and if I were an American reader with knowledge of Bokmål,
> and the book was in Nynorsk and also available in Danish,
> I would prefer the Danish version,
> as Danish is closer to Bokmål than Nynorsk is.

The big question in the application of these codes is the fact that there
are 3 codes, not 2. Yes, people can make the distinction between these two
forms, but what would be coded as Norwegian (nor)? And if nothing is left
to code Norwegian, then we have more or less have declared it to be
obsolete and that will adversely affect retrieval. LC alone has about
20,000 records in its database coded as "nor". In the past it had been
argued that just Nynorsk should be added and anything in Bokmaal could be
coded as Norwegian (this includes a request from several years ago from
the University of Oslo library to just add Nynorsk). But some on the
committee insisted there be 3 codes instead of 2. That is the problem
here, not distinguishing Nynorsk.

This is why we suggested a mechanism to code for both the general and the
specific form of the language. That would enable systems that have used
"nor" in the past to avoid splitting the records in their databases into
separate codes. Adding the more specific code to the general would enable
those who want to distinguish to do so and those who don't to drop the
second part.
 
Rebecca