The Future of Series in Bibliographic Control
David Banush
July 2006
I. What is happening with series?
The Library of Congress (LC) recently implemented a significant change in its cataloging policy. Effective June 1, LC is no longer tracing series statements for those monographic titles issued as part of a series. (A tracing is the creation of a searchable access point under authority control which allows users to collocate all items issued in the series with one search. In other words, it?s a title added entry.) Although there are some exceptions being made (for law series, for example), new cataloging coming from LC will no longer have series statements recorded in fields that are searchable directly by title in the Voyager catalog. (They will remain keyword searchable). LC is also discontinuing the practice of classing monographic titles in numbered series together, although this was a small subset of new LC cataloging. Therefore all new monographs cataloged by LC that are part of a series will be classed separately, fully analyzed (in other words, have individual bibliographic records for each volume), and will lack series added entries.
This policy applies only to records created at LC. However, other libraries may elect to follow LC?s decision, or they may continue to trace series and to class items together if they so choose. They may edit LC records accordingly. Those libraries that participate in the Program for Cooperative Cataloging (PCC) must continue to follow the guidelines of that program if contributing their records to OCLC as PCC cataloging. The guidelines currently state that all established series be traced fully under authorized forms. If an item appears with a series statement that has not been entered in the national authority file, libraries have the option of tracing or not, depending on their policies. CUL does participate in the PCC and records we create encoded as PCC continue to have series tracings for now. However, PCC cataloging represents only about 6% of our total cataloging output for the year.
II. Considerations
Before we make any decisions about what to do with series in the wake of LC?s decision, we need to ask ourselves how important access to authority-controlled forms of series headings truly is to library staff and users. We have little quantitative evidence from which to draw strong conclusion, but there are hints that series do not represent as important an access point for most users as librarians may believe. The library literature includes two transaction log studies that analyze information about series searching. In each case, series heading searches accounted for less than one percent of all catalog searches.
Because CUL does not have a separate series title search available in the OPAC, we cannot say how many series titles are searched in Voyager. The CUL transaction logs for the fiscal year 05/06 record that 47% of all OPAC searches were title searches, but we have no way of knowing exactly how many of these might be series searches. However, we can infer that the high percentage of title searches means that known item searching accounts for the most use of the catalog. Indeed, when combined with author browsing ?another typical search for known items?these searches made up over 61% of the 3.3 million OPAC searches. (Keyword searching of various kinds accounted for most of the remaining 40%). Users searching for series are less likely to know the title or author of a particular work, and those searchers would probably be most likely to use keyword searching to find materials issued in series. One final point worth making about series is that they are rarely included in bibliographies, so the access point may not be as critical as some may think.
III. Our options and their potential costs
As many of you know, LC catalog records support LTS fastcataloging operations, in which records are briefly reviewed by students or classified staff and the books are quickly sent on to the stacks. Many of these titles are new North American imprints which are more likely to be in high demand. Because of the availability of LC records and our fastcat policies, turnaround time is short and cataloging costs are extremely low for materials with LC-supplied cataloging copy. We believe these policies have served CUL users well by moving higher demand titles to users more quickly and at lower cost.
In the wake of the LC decision, we have the option of pulling items with series statements out of the fastcataloging stream and sending them to original or copy catalogers to complete the series authority work. This would allow us to retain series collocation for all titles represented by LC-supplied records. During this past fiscal year (through May 31), we fastcatted nearly 27,000 LC-supplied records. Of those, about 8700, or roughly 32%, had series statements. If LTS no longer fastcatalogs LC records with series headings on them and instead hands them off to higher-level staff for additional handling, we estimate that we would need the equivalent of nearly 3 full-time employees in cataloging to support that change.
LTS is very unlikely to obtain additional funding for this purpose. To support it, we would almost certainly have to divert LTS employees from other work. This would slow turnaround time for new library materials of all types and could result in a build-up of a new cataloging arrearage. However, Sarah Thomas has clearly stated that the library should not backlog new materials. LTS has gone to considerable lengths to eliminate our arrearages and ensure that new ones do not develop. A new backlog would be met quite unfavorably by senior administrators and would be a significant step backward in our efforts to get materials to users in a timely fashion.
LTS recently completed a turnaround time study to measure how quickly we obtain, catalog, and physically process new items for the collection. The data show that over 80% of our new receipts are currently cataloged and ready to circulate in one month or less. We would like to improve upon that figure.. If we were to change procedures to keep our series tracings intact, however, that number could drop by 20 to 30 percentage points, if not more.
Our other option is to follow the Library of Congress' decision and continue to accept LC catalog records for fastcataloging. This would mean that for a portion of CUL cataloging, we would see a gradual reduction in the users? ability to search by authorized forms for series titles. Staff and users could still search for series by keyword, and we do have some options to further mitigate the loss of direct series title searching. For example, the OPAC settings could be changed to allow a narrower kind of keyword search covering only series statements. But the most direct method of series collocation would erode away as fewer statements were traced or placed under authority control.
We can also explore the use of automation to help us keep many of the series accessible by series titles in our database. Gary Strawn of Northwestern University is developing a software tool for the Voyager system that will use algorithms to compare series statements in LC records with the authority-controlled form and flip the untraced statements to traced headings. Although the software is not yet ready, it should be available for testing soon. The approach is promising. However, no software will catch every case, and it cannot do anything about new series for which there is no authority record in the national authority files. Maintaining collocation will require additional manual effort under any circumstances.
Let?s assume that librarians in public services and collection development feel that giving up series access is an unacceptable diminution of access. As noted above, the only way for LTS to pick up the series work would be to drop or reduce other work we perform. But what other work would be acceptable to give up? Would slower turnaround time be appropriate? Would less authority work for authors? names or corporate body names be an acceptable tradeoff? Should we do more minimal level cataloging of materials not found in series? What other suggestions might there be?