Friday, September 16, 2011

European Thoughts on RDA

Some European libraries are asking the question: "Should we adopt RDA as our cataloging code?" The discussion is happening through the European RDA Interest Group (EURIG). Members of EURIG are preparing reports on what they see as the possibilities that RDA could become a truly international cataloging code. With the increased sharing of just about everything between Europe's countries -- currency, labor force, media, etc. -- the vision of Europe's libraries as a cooperating unit seems to be a no-brainer.

There are interesting comments in the presentations available from the EURIG meeting. For example:

Spain has done comparisons with current cataloging and some testing using MARC21. They conclude: "Our decision will probably depend on the flexibility to get the different lists, vocabularies, specific rules... that we need." In other words, it all depends on being able to customize RDA to local practice.

Germany sees RDA as having the potential to be an international set of rules for data sharing (much like ISBD today), with national profiles for internal use. Germany has starting translating the RDA vocabulary terms in the Open Metadata Registry, but notes that translation of the text must be negotiated with the co-publishers of RDA, that is the American Library Association, the Canadian Library Association, and CILIP.

The most detail, though, comes from a report by the French libraries. (The French are totally winning my heart as a smart and outspoken people. Their response to the Google Books Settlement was wonderful.) This report brings up some key issues about RDA from outside the JSC.

First, it is said in this report, and also in some of the EURIG presentations from their meeting, that it is RDA's implementation of FRBR that makes it a candidate for an international cataloging code. FRBR is seen as the model that will allow library metadata to have a presence on the Web, and many in the library profession see getting a library presence on the Web as an essential element of being part of the modern information environment. One irony of this, though, is that Italy already has a cataloging code based on FRBR, REICAT, but that has gotten little attention. (A large segment of it is available in English if you are curious about their approach. )

The French interest in FRBR is specifically about Scenario 1 as defined in RDA; a model with defined entities and links between them. An implementation of Scenario 2, which links authority records to bibliographic records, would be a mere replication of what already exists in France's catalogs. In other words, they have already progressed to level 2 while U.S. libraries are still stuck in level 3, the flat data model.

Although the French libraries see an advantage to using RDA, they also have some fairly severe criticisms. Key ones are:
  • it ignores ISO standards, and does not follow IFLA standards, such as Names of person, or Anonymous classics*
  • it is a follow-on to, and makes concessions to, AACR(1 and 2), which is not used by the French libraries
  • it proposes one particular interpretation of FRBR, not allowing for others, and defines each element as being exclusively for use with a single FRBR entity
They recommend considering the possibility of creating a European profile of RDA scenario 1. This would give the European libraries a cataloging code based on RDA but under their control. They do ask, however, what the impact on global sharing will be if different library communities use different interpretations of FRBR. (My answer: define your data elements and exchange data elements; implement FRBR inside systems, but make it possible to share data apart from any particular FRBR structuring.)

* There is a strong adherence to ISO and IFLA standards outside of the U.S. I don't know why we in the U.S. feel less need to pay attention to those international standards bodies, but it does separate us from the greater library community.

(Thanks to John Hostage of Harvard for pointing out the recent EURIG activity on the RDA-L list.)


maryakem said...

So grateful for your blog as we are working through our latest curriculum on cataloguing and classification.

I think it is clear to us, and hopefully soon to our students, that FRBR is the adjustment to make not the adjunct of RDA. RDA will never respond to the multiligualism of new resources.. while it is intrinsix to FRBR.

Anonymous said...

This one: also comes from Europe (France again). Greetings from over the ocean!


Charles Riley said...

"There is a strong adherence to ISO and IFLA standards outside of the U.S. I don't know why we in the U.S. feel less need to pay attention to those international standards bodies, but it does separate us from the greater library community."

It's interesting to note that we do follow at least one ISO standard, 639-2/B for identification of languages in a bibliographic context, for which LC is the registering authority.

ISO 2022 & 8859 as well could be counted, I guess, for character encoding.

Others which should be given at least equal consideration for going forward might include: 3166 (for countries), 639-3 (for more granular language tagging), 15924 (for script identification), 5426 & 6438 (both essentially subsets of 10646), for more universal character encoding.

UNIMARC employs several of these; their adoption (or broader availability of mappings through SKOS) would result in better integration with web services.

Karen Coyle said...

Philippe -- thank you for the link to the French RDA blog. I will try to remember to check it from time to time, but please send me a note if there is something significant that we in the US should pay attention to (like reports from IFLA or from the RDA development group in France).

To all: DO check out the "LOST: 2nd season of RDA" visual for a laugh.