I’m a bit late in writing up my report about the 2011 ALA in New Orleans, because I’ve been trying to find the best way to explain a statement that profoundly affected my thinking about cataloging. I heard it at the MARC Formats Interest Group session, which I chaired and moderated. The topic of the session was “Will RDA Be the Death of MARC?” and the speakers were Karen Coyle and Diane Hillmann, two very well-known cataloging experts.
Coyle spoke first, and elaborated a devastating critique of the MARC formats. She argued that MARC is about to collapse due to its own strange construction, and that we cannot redeem MARC, but we can save its data. Coyle argued that MARC was great in its day, it was a very well developed code for books when it was designed. But as other materials formats were added, such as serials, AV materials, etc., additions were piled on top of the initial structure. And as MARC was required to capture more data, the structure of MARC became increasingly elaborate and illogical. Structural limitations to the MARC formats required strange work-arounds, and different aspects of MARC records are governed by different rules (AACR2, the technical requirements of the MARC format itself, the requirements of ILS’s, etc.). The cobbled-together nature of MARC has led to oddities such as the publication dates and language information being recorded in both the (machine readable) fixed fields of the record and in the (human readable) textual fields of the record. Coyle further pointed out the oddity of the 245 title field in the MARC record, which can jumble together various types of data, the title of a work, the language, the general material designation, etc. This data is difficult to parse for machine-processing. Although RDA needs further work, it is inching toward addressing these sorts of problems by allowing for the granular recording of data. However, for RDA to fully capture this granular data, we will need a record format other than MARC. In order to help develop a new post-MARC format, Coyle has begun a research project to break down and analyze MARC fields into their granular components. She began by looking at the 007/008 fields, finding that they have 160 different data elements, with a total of 1,530 different possible values. This data can be used to develop separate identifies for each value, which could be encoded in a MARC-replacement format. Coyle is still working on breaking down all of the MARC fields.
After Karen Coyle, Diane Hillmann of Metadata Management Associates spoke about the developing RDA vocabularies, and it was a statement during her presentation that really struck me. The RDA vocabularies define a set of metadata elements and value vocabularies that can be used by both humans and machines. That is, they provide a link between the way humans think about and read cataloging data and the way computers process cataloging data. The RDA vocabularies can assist in mapping RDA to other vocabularies, including the data vocabularies of record schemas other than the MARC formats. Also, when RDA does not provide enough detailed entity relationships for particular specialized cataloging communities, the RDA vocabularies can be extended to detail more subproperties and relationships. The use of RDA vocabulary extensions means that RDA can grow, and not just from the top-down. The description of highly detailed relationships between bibliographic entities (such as making clear that a short story was adapted as a radio play script) will increase the searching power of our patrons, by allowing data to be linked across records. Hillmann argued that the record has created a tyranny of thinking in cataloging, and that our data should be thought of as statements, not records. That phrase, “our data should be thought of as statements, not records,” struck me as incredibly powerful, and the most succinct version of why we need to eventually move to RDA. It truly was a “wow” moment for me. Near the end of her presentation, Hillmann essentially summed up the thrust of her talk, when she said that we need to expand our ideas of what machines can and should be doing for us in cataloging.
The other session I went to that is really worth sharing with everybody was the RDA Update Forum. Representatives from the Library of Congress and the two other national libraries, as well as the chair of the PCC (Program for Cooperative Cataloging), discussed the results of the RDA test by the national libraries. The national libraries have requested that the PCC (the organization that oversees the RDA code) address a number of problems in the RDA rules over the next eighteen months or so. LC and the other national libraries have decided to put off implementing RDA until January 2013 at the earliest, but all indications were that they plan to adopt RDA eventually. As the PCC works on revising RDA, the national libraries are working to move to a new record format (aka schema or carrier) to replace the MARC formats. They are pursuing a fairly aggressive agenda, intending to, by September 30 of this year, develop a plan with a timeline for transitioning past MARC. The national libraries plan to identify the stakeholders in such a transition, and want to reach out to the semantic web community. They plan for this to be a truly international effort that extends well beyond the library community as it is traditionally defined. They plan to set up communication channels, including a listserv, to share development plans and solicit feedback. They hope to have a new format developed within two years, but the process of migrating their data to the new format will take at least several more years after the format is developed. Needless to say, if the library world is going to move post-MARC format, it will create huge changes. Catalogs and ILS systems will have to be completely re-worked, and that’s just for starters. If some people are uncomfortable with the thought of moving to RDA, the idea of moving away from MARC will be truly unsettling. I for one think it’s an exciting time to be a cataloger.