I'm interested in how you avoid duplication in this cooperative
effort. In other words, if I were sitting some place and had ten
items that I wanted to catalog, I might want to devote some effort to
making sure that my ten entries were into the system. On the other
hand, if there were ten of us doing the same thing, I might be way
ahead of the game by waiting for the other nine to do the work for
me.
Beacher Wiggins:
There are components of everything you say in the answer to the
question. One of the ways duplication is controlled is the fact that catalog
librarians at many institutions have fairly immediate access
through the bibliographic utilities, so they see what has already been
cataloged. The workflow procedure for new material at most institutions
incorporates searching to see
if cataloging copy has already been created by another institution.
The fact that we have the mirror authority files, so you can see
whether authority headings have already been created for the work, is another
way to make sure that there isn't any duplication. Even
with the fast turn-around time of 24 hours there is still, as you
might suspect, some window for duplication, and that happens
periodically. We have built into our national bibliographic
infrastructure ways for having those duplicate catalog records
reported to us, and generally the Library of Congress takes that on.
As we have become increasingly "on-line", the incidence of duplication
has certainly decreased over the years. With more and more libraries
getting on-line systems, we hope to reduce that amount of duplication
even further.
Roy McDiarmid:
In setting up bibliographic reference points to taxonomic names,
which many of us do, there are various kinds of numbers that are
associated with books, and serials, and whatever, including LC
numbers. From your perspective, which of those should we pay
attention to?
Beacher Wiggins:
There are several numbers involved, and depending on what you're
using it for, the main controlling number for a bibliographic record,
for the Library of Congress and the library community, would be the
Library of Congress Control Number. For authority records there is a
similar control number that you would use to link
that record then to any other authority record you would create. Presumably
the system then would be able to match those numbers and be able to
display a hierarchy, or even an encoding listing, of some of the headings that
you might have created. It gets back to some of what Karen was
beginning to describe with authority control on a global basis, where
we're beginning to look at having one number, and that number would
control a number of headings each for a particular language or
cultural background. If you use the control number for any authority
records we've created, and mapped that to any
authorities you create, in a taxonomic sense, I think that
would ensure some level of integrity.
Barbara Stein:
You mentioned at the very end of your talk that you hope now to be
able to turn more attention to electronic media. Would you
briefly state what you think the differences are in dealing with
information you get electronically versus in paper form.
Beacher Wiggins:
By and large for us, at least at this juncture, it is simply a matter of
having the staff resources to focus in a way that would give us true
benefits from dealing with electronic resources. So much of what we
have to deal with now—speaking from the perspective of the Library
of Congress, and I think it's true for other major research libraries—
is the on-going influx of print material, and the limited staff
that we have still have to maintain control over those. There are
inherent problems in dealing with electronic resources, of dealing
with the transient nature and determining how much of it is worth
expending our time to provide the kind of cataloging that we've been
doing in the past. Does the full-fledged MARC record serve the
purpose for electronic resources or should be we finding different and
more streamlined approaches for doing that? In order to take this on we
need to have staff engaged enough in that kind of processing to see
what the problems are, to see what the pros and cons are. As we've had
a hundred more years to devote to the print world to come up with
our procedure for print materials, we simply have not had that luxury with electronic
resources. So we're hoping that if we can get more authority records and more
bibliographic records created collectively, then each of us can have
the leisure of turning some of our staff to focus in a more direct way
on these new issues. That's what I had in mind. We've started that process at
just about every major research institution, we've started it at the
Library of Congress and it's overwhelming. We're nowhere near where
we want to be.
Adam Schiff:
I just wanted to comment on Roy's question about linking to
bibliographic records. The vast majority of the source citations that
you would include in a taxonomic authority file would be to journal
literature rather than the kinds of bibliographic records that would
be included in the database that libraries manage. Most of what we
are cataloging are whole books, or journals themselves, but not the
individual contents, and so that would be a problem.
Beacher Wiggins:
But might not there be also some of the headings that are created
that could serve a purpose for the subject side of what we
create— subject authorities? There should be some way we should
explore those kinds of links. That was the thrust I was hoping you would
pick up on.
John Mitchell:
To follow on that, since I am in the Cooperative Cataloging Team
and a proponent for SACO, we would highly welcome your contributions
to the SACO program, as long as you submit appropriate authority
control and citations.
Gail Hodge:
Nice plug, John! In response to what Roy was saying and some of
the other comments, there is also work within the primary publishing
community and the secondary publishing community to develop a system
for article level identification, primarily within journals. There is
some in use within Elsevier's publications I believe, the Primary
Publisher Item Identifier (PII), which tracks an item from manuscript
all the way through to publication, and then incorporation of that
into a larger standard called digital object identifier, which can
then break a whole manuscript down into pieces that can then have some
level of rights management, for example copyright of a photograph within
the span of an article. So those are the kinds of things to look out
for. They're not here, yet, but there are several primary publisher who
are involved in major tests at this point. I think that it's
going to come.