This year's meeting coincided with the NATO 50th Anniversary celebration on Friday April 23. The Library of Congress was shut down and Thursday's schedule was rearranged to accomplish the major agenda items on that day. However, there was insufficient time to fully discuss all of the issues. Further discussion on CONSERLST will be encouraged.
The opening and closing sessions were held in conjunction with BIBCO,which held its meeting during the same time period. Jean Hirons, CONSER Coordinator, and Ana Cristan, BIBCO Coordinator, opened the session. Sally Sinn (NAL), Chair of the PCC Policy Committee, provided a welcome from the PCC Policy Committee. Also attending was Mike Kaplan (Indiana), chair-elect.
Jean Hirons chaired the CONSER meeting. Bill Anderson, CONSER Specialist,assisted and took notes. Representatives from all but two of the CONSER full and associate member institutions (Pittsburgh and New York Public) were in attendance, as well as CONSER Enhance and Affiliate institutions, and additional catalogers from local institutions, LC, and UCLA. CONSER discussions focused on four major topics: seriality and possible changes to AACR2, use of the CONSER single record technique for online versions, the potential of adding publication patterns and related holdings to CONSER records, and a report on the progress of the task force working with aggregator databases.
Anderson reported on the next update to the CONSER Editing Guide, which will reflect a number of changes, including LC's implementation of a new ILS. The update will include instructions on linking to monographs, previously forbidden, and use of the new form subdivision $v in subject headings. One major change resulting from the ILS is that LC catalogers will no longer "swap out" sn-prefixed LC control numbers for unprefixed numbers. The change is due to the difficult of deleting records in the ILS.
Anderson also discussed instructions that will be added to the CEG and CCM for serials with multiple relationships to another manifestation. When a serial is issued in both print and online versions and subsequently, the print is discontinued, there are two potential linking relationships and linking relationship notes: 530/776 for the additional version and 580/785 for the continuation. Since the 780/785 linking relationship is considered most essential, this is the link to be used. Either a 530 (if already in the record) or 580 note is appropriate.
The JSC report recommends the use of "continuing" as an umbrella to group successively-issued and integrating resources. In discussing how this might translate into MARC 21 in regard to the bibliographic level codes, Reynolds suggested opening records to all PCC participants, but it was noted that there are many OCLC Enhance members that would also be affected. Adam Schiff (Washington) and others said that in many institutions catalogers are working on both monographs and serials and that changing a category such as databases to serials would disenfranchise those currently working on them. However, Kaplan agreed that databases are a key issue because they are serials in print and CD-ROM formats but monographs online. He noted that this is an area that can't wait for rule revision. There was also discussion about the use of code "s" in the bibliographic level and its use in systems to separate journal titles in searches.
The report recommends that both earliest and latest publishing information be included in the publishing area. In MARC 21 this might be accomplished by defining new subfield codes to be used in a single 260 field, by using two 260 fields, or by defining a new 26X field. Ed Glazier (RLG) found the 260 options intriguing and suggested the use of new indicator values to identify the earliest and latest publisher. Others found the multiple 260 fields preferable to a single, more congested field.
In regard to code organization, Hirons explained that the report supports Tom Delsey' suggestion of an area-by-area approach. Glazier, who is on the CC:DA 0.24 Task Group, which has been studying this possibility, suggested that more attention needs to be given to the principles for each area and that the principles should be the key to any reorganization of the code. Judy Kuhagen (LC) noted that the code needs to be more explanatory for purposes of training. Hirons noted the recommendation for an introductory chapter.
A separate CONSER discussion followed. One of the concerns was the recommendation not to use uniform titles on latest entry records. This would impact databases and e-journals cataloged according to latest entry. Hirons admitted that this was one recommendation she wasn't terribly comfortable with and that it needed more thought. However, Beth Guay (Maryland) said she liked the simplification.
Concerning latest entry and ISSN, Regina Reynolds (LC) said that she would be working with ISSN directors to decide how the use of latest entry could be accommodated by ISSN. She will be meeting in June with a group revising the ISSN Manual and will address the issue at that meeting.
Davis raised the question whether unnumbered series would be impacted. Kuhagen explained that the current practices for dealing with unnumbered series wouldn't really change, but they would be better accommodated by the code.
Hirons then reviewed all of the recommendations in the report and also discussed "considerations" also included for ideas that need further exploration. The JSC will meet in the fall to discuss the report and the reactions to it and will determine next steps. A harmonization meeting will also take place in the fall for constituents of ISBD(S), ISSN and AACR.
The draft suggests three rules of thumb: the single record would be appropriate when either the online version has sufficient content to serve as a substitute for the print, or when the online version lacks sufficient content to be considered an equivalent (and perhaps as a result does not warrant separate cataloging). Separate records would be preferred when the online version contained significant additional material. A question is how to judge the degree to which the content is different from that of the print. Can we rely on statements from the publisher? In some situations the cataloger doesn't have access to the print.
Reynolds raised the issue of using an 007 in records for the print, saying that it is confusing the matter of what the record represents. The issue will be further explored to see if we want to discontinue using the 007 in print records.
There were also comments on the list of pros and cons. Carroll Davis (Columbia) said he didn't think we could tell institutions what to catalog and thus, the paper could only be suggestive, not policy. GPO will continue to use the single record for most cases according to Steve Utoff. Jeanne Baker (Maryland) noted that use of the single record technique can make it difficult to decipher what one has. Others said we should be more welcoming to its use.
Hirons will work with Bross to determine where to go with the guidelines.
Hillmann's proposal is to include publication pattern and accompanying holdings data in the USMARC holdings format in CONSER bibliographic records. The group's presentation explained what is being proposed, and gave suggestions on why, why now, how it would be accomplished, and by whom.
Hirons led discussion. She first asked, what is needed? Do we need both current and retrospective patterns in records since a primary use of patterns is for predictive check-in. Kaplan thought only the current pattern is needed. Riemer advised against throwing away information that we may need. In addition to the patterns, do we really also need the holdings? Hillmann wants the group to take the long view and try to see the value of holdings data in records. However, Robert Bremer said that, to his knowledge, OCLC had no plans of increasing record length. The point of supplying holdings is to tie the patterns to specific holdings and also to build a generic holdings record to show what has actually been published. If holdings are not included at this time, might we use $3 to indicate the period of coverage to which a pattern applies?
Hirons asked about institutional commitment. If FAXON was to give us their publication pattern data they would expect us to keep it up-to-date. There was little discussion.
The technical issues are significant. One of the basic problems is that systems vary in their application of patterns to the degree that for manyserials there may not be a single generic pattern that could be used across all systems. Furthermore, system application of the holdings format and export/import capabilities seem to vary widely. Kristin Lindlan (Washington) expressed concern about automatic overlay of bibliographic records and that local notes might be overlaid when changes to patterns/holdings are made.
There was also the question of how systems would distinguish between generic and local holdings if the data were in the bibliographic record. Anderson suggested that we might better look to the OCLC LDR (local data record) model and try to attach the CONSER generic holdings and patterns in an LDR-type record.
Other questions regarding technical issues were how a library would find out that a pattern has changed and how sequencing numbers could be controlled.
Staffing raises other concerns. Who in the institution would make changes to the CONSER records? How would authorizations be handled? Would staff working with holdings also make changes to the bibliographic information, such as the fixed field and frequency fields?
Finally, Hirons asked whether the idea of a stand-alone publication pattern database (to be used as a source for patterns in CONSER records) should be considered. It would seem that both are needed. LC is anticipating having over 3,000 patterns and has established a name convention scheme that might be used more broadly. Patterns in the database could be coded for use in bibliographic records.
The following seem worthwhile for the task force to explore:
The assumptions that the group is working with are: 1) for many libraries the title in the aggregation will be the only format held; 2) records need a field for de-duping with record for the original; 3) a flag of some type is needed so that all records can be deleted if a subscription is dropped; 4) data in records will be a subset of that found in the CONSER record for the print; 5) a single record reflecting multiple aggregations will be possible; 6) loading records will require prior local customization; and 7) summary holdings data, if supplied by the vendor, could be included in $3 of field 856.
Karen Calhoun (Cornell) then walked the group through sample records created by Oliver Pesch of EBSCO, based on the guidelines established by the task force. The records use an encoding level 5 and include the ISSN of the print in field 776. Both the encoding level and placement of related ISSN are issues needing further discussion. The task force is sponsoring a discussion paper for MARBI that proposes the definition of a new subfield in field 022 to accommodate aggregators.
Calhoun then gave a demonstration of how other libraries are creating records for titles in aggregations. Using PERL scripts and the software MARC Maker, the University of Tennessee Knoxville accessed descriptions from the Dow Jones and ProQuest web sites and created brief records for 4,271 titles in the Dow Jones set and 1500 titles in ProQuest with only a few hours required to create each set of records. University of Tennessee Knoxville staff report their library users love the records, despite the fact that there is no subject data. They are including ISSN in 022 subfield 9, a local subfield. They also use a local 945 field for the aggregator and contain a control number for the set so that it can be added or deleted. However, the records cannot be shared with others. The University of Illinois, Chicago is creating a set of records for H.W. Wilson Select full text journals and OCLC TechPro has created records for Elsevier Science full text journals and Academic IDEAL full text journals. Both OCLC and the University of Illinois Chicago are creating standard MARC records. Illinois is placing the ISSN in 022 $y; TechPro is using 022 $a.
One problem for the task force is to discern what fields the vendor can supply when they don't have access to the print records, as EBSCO does.
Glazier noted that using the 773 field for the vendor may be stretching the definition of the field and that any URL used in the field should be for the overall aggregator and not for an individual title within.
Calhoun noted that UMI has expressed interest in working with us. EBSCO will be placing the records on their Website for downloading. As to maintenance, a revised set might be reissued quarterly using the date of update in the record to replace them.
Jeanne Baker reported that she will be giving several presentations on the topic. At NASIG in June, she will join Yumin Jiang (Cornell) in a workshop on access to serial titles in aggregator collections, and at the American Association of Law Libraries in July, she and Hirons will discuss the group's progress.
Go to:
Library of Congress