I see from the early responses to my "guidelines" document that there's a basic principle the document doesn't discuss.
We're not (yet) creating nonroman HEADINGS. In the LC/NACO Authority File, there will be, for the foreseeable future, no 1XX that is not in roman script. What we'll begin to do in July (I hope it'll be July) is add nonroman 4XXs to records, and these will have no more "authority" than roman 4XXs. I know that in some of your catalogs, you DO have "nonroman headings," and it makes sense there to say "they should be constructed according to AACR2." But we're not talking about that yet in the national authority file. The fact that there's no such thing as a nonroman heading has consequences. When a subordinate body, for example, is established in a $b, the $a of the heading has to be the established form of the name of the parent body--therefore it has to be roman. And when you add a reference for the nonroman version of the subordinate body's name in a $b, the $a will still have to be in roman, because that's the way the file works. If you're adding a nonroman reference for the title of a work, say the Yiddish translation of Singer's 'Eved, the $a of the reference will still have to be the established, i.e. the roman, form of Singer's name. This accounts for some of those ugly-looking mixed-script fields in the examples in the document. In other places, AACR2 directs us to add a qualifier "in English." Will that rule have to be followed in references where the entry-element is nonroman? Just some things to think about before we get into our discussions. Joan
