I'm renaming this thread, since we're off-track on another issue, which
is MADS.

On Feb 21, 2005, at 2:21 PM, Mark Hart wrote:

> I can see how that is helpful as a cataloging tool, particularly with
> titles; but I think that some of the difficulty that people are having
> with MADS in this thread is that they are trying conceive of MADS as a
> formatting tool for bibliographic work.

Just a minor correction: I wasn't really couching it in this way.
Better markup makes for generally increased flexibility.  It has its
costs too, of course.

> My impression is that bibliographic references to published resources
> are supposed to identify the resource by the information that was
> presented in the resource.

No, this is a common misconception, I think.  If I publish in one
journal under "B D'Arcus" and cite myself in another journal that
requires the full name, I'd put the full name.  Otherwise, it'd be
impossible to reconcile different formatting.

> Some other concerns.  If a user is trying to associate an author, say
> John Smith to an authority source such as a MODS list, how can the user
> be sure that a particular John Smith in the authority list is actually
> the one that wrote the paper: often a resource will not include all the
> information that a user (particularly one without intimate knowledge of
> the resource realm) will need to match it unambiguously to an authority
> record. This is not a problem if the authority list already contains a
> name record with and associated title, but rather just a name authority
> record.

Right; why your bank uses social security numbers.

> An aspect that Bruce brought up.  MODS itself does not seem to have any
> method of recording the ordinal position of an author in a resource
> with
> multiple authors.  Does it strictly rely on position within the record
> for author ordination? It seems an explicit method of indication that
> would be needed. Does the ID attribute do that? Nowhere in the
> documentation that I have read is there a description or example that
> would indicate how that might function (unambiguously) if the ID
> attribute was used.

No.  The only way you can deduce who a first author is is to assume
it's the first name with an author role.

The ID attribute is to able to link across names.

Bruce