Karen Coyle wrote,
>
> With metadata, the metadata record iself is not the end goal; the end goal
> is the "thing" (and I use that term loosely) that it represents. If we
> start seeing metadata itself as a goal (and I think some are tending in
> that direction) then it loses its value for discovery of anything but
> itself. In that case, I would argue that we have "data" not metadata.
>
Resource Discovery was the goal metadata movements have focused on from the
beginning, and indeed the discussion sometimes seemed to branch off or even
take a u-turn. Efforts to improve resource discovery must keep the end-users
and their goals in mind. End-users seldom look for a "Creator", never for
a metadata record (let alone ITS creator), they look for an artifact, mostly
something one would call a "work" (in library cataloging terms).
Although the "work" is an eminently important notion in cataloging,
there is no agreed-upon definition for it. Not even in the AACR (Anglo-
American Cataloging Rules)!
The 1:1 issue, in my view, has these aspects:
1) Metadata should describe entities that people may want to discover.
More often than not, people are not after one specific embodiment
or manifestation of a "work" (whatever the definition), they want the
work itself or a suitable representation of it. (A bad one is still
better than no one.)
2) Every particular document or resource can be a container containing
more than one manifestation of more than one work. The only appropriate
approach at metadata from the end-users view is, I think, to have one
metadata record for each work. (one cannot know in advance for which
one they will be looking)
3) Every particular document or resource can be a representation of just
one work, but it can consist of any number of parts. It is not the parts
the end-user will want to discover, it will be the whole.
Conclusion: Metadata should be created with a work-oriented rather than
a manifestation-centered view.
Last year's big conference on the future of AACR in Toronto, and the
discussions before and after it, left many librarians wondering if they,
too, had not for all too long been following the manifestation view.
In terms of books, it means they have done one record for one physical
item. When what is needed is one record for one work.
The big question this brings up is, how do you name a work? That's why we
have those elaborate authority control procedures in cataloging, along
with databases of names and titles. Conceptually, only verified and
authorized forms of names and titles should enter a database. Or the bigger
it grows, the more inconsistent it will become, the less useful for
resource discovery. If you wonder what I'm getting at, try to find a
representation of Tchaikovsky's "Nutcracker" in any library or catalog.
If in fact you are looking for one specific manifestation, then indeed
"Contributors" come in, like the conductor or the orchestra, or the
arranger or choreographer. But only then. These questions can best be
answered, if first of all we have one and only one name for the work.
This, I think, is the most important ONE in the 1:1 debate.
B.E.
Bernhard Eversberg
Universitaetsbibliothek, Postf. 3329,
D-38023 Braunschweig, Germany
Tel. +49 531 391-5026 , -5011 , FAX -5836
e-mail [log in to unmask]
%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|