Catalogers Discussion Group
New York City

The ARLIS/NY Cataloger's Discussion Group convened on June 11 at the Pierpont Morgan Library. The first hour of the meeting was devoted to a discussion of the bibliographic utilities RLIN and OCLC. Those in attendance sought to compare them as cataloging environments, and to share their experiences as users.

NYU provided a model for the procedures of a large university library. NYU uses both utilities for source copy, seaching OCLC first, then RLIN. (They feel that OCLC gets CONSER records first, since it is the parent database for CONSER). Instead of searching RLIN directly, they use MARCADIA, RLIN's batch searching software.

One participant asked whether other libraries were making use of the "third, invisible bibliographic utility," i.e. individual library OPAC's accessible through the Internet. One cataloger working on specialized material regularly searches 4 OPACS which contain records for similar material. But RLIN and OCLC are still the main source for copy. It was generally observed that RLIN's cluster display results do not offer the clarity of the single "master record" approach used by OCLC. But the master record display results in the disapearance of local information, and there can be transcription problems here as well--these may be particularly telling where rare books are concerned.

Many participants have noticed a general decline in the quality of records in both utilities. Several factors appear to be at work here. One is the records from European libraries that do not conform to required standards. These records are good for acquisition purposes, but require extensive revision. This may take longer than creating an original record, but catalogers are required to use existing copy. Some libraries entirely rewrite the records; some merely English the notes and accept the rest, including foreign-language subject headings.

Inaccurate coding is another problem. Increasingly, non-standard records from many sources are being coded as standard. It was agreed that there are so many ways to code records now that code-based evaluation becomes muddled. But the library which is the first to change the coding from non-standard to standard should make an effort to ensure that the record really is up to par, or leave it coded as non-standard.

It was also observed that as more people use shared copy the overall quality of copy will decline over time. The source pool is diluted as catalogers retire and there are fewer new catalogers to replace them. More para-professionals are used in the cataloging process, and sometimes it is hard for them to recognize the best available source copy.

The difficulties of using EUREKA were observed from the perspective of reference librarians and those concerned with inter-library loan. "Open" records for multi-volume sets and serial records pose challenges, as its often difficult to tell which libraries have a particular volume when using EUREKA. Often it is necessary to combine several records in order to ascertain the true extent of a multi-volume title.

It was observed that over time the concept of RLIN as a union catalog has altered with the addition of RLG's extra databases, such as the German national library catalog, ESTC, and SCIPIO. One-stop searching is affected by the fact that one must pay RLG for these databases, thus eroding RLIN's utility as a union catalog.

The second half of the meeting was devoted to an in-depth demonstration of Cataloger's Desktop, which was performed for the benefit of the group by V. Heidi Hass, Head of the Morgan Library's Reference Collection. A modifed version of her presentation will be posted on Sherman Clarke's Geocities home page ( in the near future.

The next meeting of the CDG will be held at the Cloisters on a date to be announced.

Peter Gammie
Cataloger, Reference Collection
The Pierpont Morgan Library