Questioning Authorities: Adapting Authority Control to the Changing Needs of Library Users: LITA/ALCTS CCS Authority Control Interest Group (part 2)
Manon Theroux, Authority Control Librarian at Yale, gave a thorough and deliberate presentation on the vernacular field authority control issue which was a bit of a relief to this note taker after trying to keep up with the fast-talking Mr. Singer. The issue involves changes made to headings (indexed access points) in records sent from an institution (in this case, Yale) to an authority control vendor (in this case, MARS from Backstage Library Works, whose Authority Control Product Manager John Reese followed Manon’s presentation with the vendor’s eye view) and the complexity introduced by parallel “romanized” and “vernacular” encoded fields for non-roman alphabet language controlled access points such as names, series and subjects.
When the record contains “vernacular coding” so that languages in non-roman alphabets can be displayed, and in some cases indexed in a catalog alongside the romanized version, the vernacular version of a heading is not updated when the romanized heading is – because updates are based on authority records (from the National Authority File) which currently only contain romanized forms of headings. Catalogers must find out about these changes and manually update the vernacular form if readers of that language have correct information. Corrections cannot be applied in the same way to indicators (a field which describes the size of initial articles to allow them to be ignored in browsing) because the vernacular article may differ in size from its roman counterpart. When a series heading must change, (from a 440 to 490 _1 / 830 split, for you catalogers) so that the statement of what’s on the work differs from the authorized form, the tag numbers of the romanized headings change, and the vernacular fields (880 tags) and subfields in both tags describing their linkage, need to change so that the link is maintained. The vendor has made numerous changes to its processing due to problems reported by the customer (Yale) and is looking at how to extend services in this area.
I have to note that series authority updates, which dominated the examples here, are one of the toughest and most labor-intensive parts of authority control as practiced in most library systems — not to indicate approval for the LC decision discussed in the next presentation, but that’s how it is. But, is this difficulty the fault of the pesky series themselves and their publishers’ propensity to vary/change title or use common titles that need to be “disambiguated”? Should we blame the MARC format and its application that requires a tag number change in order to keep straight the published form and the controlled form of a series when those start to differ (automated tag changes based on authority changes aren’t handled well or at all by most ILS systems)? Or is an authority control strategy that’s based on string matching rather than, say, linking with identifiers, really the problem?
Beecher Wiggins (Director, Acquisitions & Bibliographic Access, Library of Congress) arrived just in time to go on before the question/answer session was starting. His presentation had the formality of a government official giving a briefing or annoucement. The Library of Congress has a lot of new initiatives and changes in the works, beyond the series decision mentioned in the agenda. Points mentioned:
LC has created documentation and refined existing documentation related to the series change, which in case you missed it, involves the LC decision to cease creating authority records for series or record controlled headings in bibliographic records. It will continue to record series statements – the form of series as it appears on the work cataloged (in 490 _0 tags, for the catalogers among us) in original and copy cataloging. LC had studied and consulted with the cataloging community on the possibility of this change a decade ago, decided against it then, but felt that the environment had changed, that indexing and keyword access were sufficiently powerful to provide access to content through the series statement.
This decision, which was implemented June 1 (delayed one month due to concerns expressed by the cataloging community), has been heavily discussed on discussion lists and in several other ALA forums; little discussion occurred at this meeting though there were a couple of questions at the end, one about whether there’ll be opportunity for input from the library community on decisions about LCSH (answer, LC managers have yet to report recommendations; there will be opportunity for wider comment and input after that) and whether LC had developed or documented alternative searching strategies to compensate for lack of controlled vocabularies and cross references (not yet, but they are meeting with public services and acquisitions staff and if anything is developed to address their questions that would be useful to share, LC will do so).
Whew! Did I miss anything?