On Wed, 2004-09-08 at 06:43, Keith Jenkins wrote:

> Somewhat related to the recent "mods:name" discussion, I would be interested in
> hearing people's ideas about the international/multilingual potential of MADS.

Some libraries do have multilingual authority records, i.e. Nat'l
Library of Canada, but I don't know more about it than that. Hopefully,
someone else on the list does. I believe that it only applies to subject
headings, though, since names in English and French both use the latin
alphabet. As far as I know, names only get "translated" when the script

> What if the MADS schema were modified to allow different authorized forms for
> different languages (although no more than one per language)?  Besides the
> transliteration/script issues, for certain qualified names there are
> significant differences between different languages.  For example, from three
> different national catalogs, we have:
>   [LOC] Catherine II, Empress of Russia, 1729-1796
>   [DDB] Ekaterina <Rossija, Imperatrica, II.>, 1729-1796
>   [BNF] Catherine 02 (impératrice de Russie ; 1729-1796)

As you show here, it's not so much different languages, but different
authorizing agencies. It's easy to imagine that there will be different
forms of the name in the same language when used in different countries
with different name rules. (Think of all the countries with Spanish as
their language.) Presumably all of these libraries have their own
authority files, but the key thing is that there is no combined file
that shows that the names above represent the same human unit. That
would be interesting... big job, though -- someone(s) would actually
have to bring together the names for the same person, it couldn't be
done automagically.

But this is still a different case than the one Bruce wants to solve. He
has the same name in different *scripts*. And he doesn't really have an
authoritative form, or at least that's not an issue. He has whatever the
form was in the citation, which I assume he'll render faithfully.
Libraries, as your Catherine example shows, make up their own forms of
the name, which may not be what was on the piece. I'm not sure what role
an authority file plays when you aren't creating authorized forms of the
name in the bibliographic metadata. You could say that an authority
record could be a record of equivalent names, but as you know that
doesn't resolve the ambiguity between all of the John Smith's out there.
Maybe its purpose wouldn't be to resolve ambiguity -- maybe it would
just be a way to store records compactly -- maybe it would be a way to
expand searches rather than narrow them -- maybe ??


> =================================
> Keith Jenkins, Metadata Librarian
> Mann Library @ Cornell University
> [log in to unmask]    (607)255-7953
> =================================
Karen Coyle
Digital Library Specialist
Ph: 510-540-7596 Fax: 510-848-3913