RDA Test Final Report

Posting to RDA-L

One of the most remarkable findings I have found in the RDA Report is on p. 103:

“Intriguingly, a majority of test institutions thought that the U.S. community should implement RDA while at the same time, a majority believed that the implementation would have a negative impact on their local operations.

In general, formal test partners reported that they needed more time to create or update an RDA record than a record using their current rules; this was clear from responses to IQ question no. 2, “Please provide any general comments you wish concerning the test, the RDA Toolkit, or the content of the RDA instructions.” Comments showed major concern over the initial costs of an implementation that would be evident in reduced production for an unpredictable length of time. Several commenters also stressed the need for some kind of bridge document, possibly as part of the workflows in the RDA Toolkit: “The RDA instructions are organized according to FRBR and FRAD principles while the descriptive cataloging process remains linear by format. We found RDA to be a collection of rules that are ordered without respect to our existing workflows ….””

I will grant that once people are trained, they will probably be able to make RDA records more or less as quickly as AACR2 records. But it seems obvious that an increase in productivity is not a reason to move to RDA, therefore, we must look elsewhere for reasons. Other reasons are addressed on p. 111, where people mentioned the benefits of RDA. Analyzing the benefits is also quite fascinating. They mention changes in “perspectives” and “how we look at the world”. RDA “encourages” developments, is felt to be more “user-centric”, and other highly nebulous benefits that make very little difference in a business case. I disagree with much of this, but it goes on to mention some specific benefits:

“RDA testers in comments noted several benefits of moving to RDA paraphrased as follows:

  • using language of users rather than Latin abbreviations,
  • seeing more relationships,
  • having more information about responsible parties with the rule of 3 now just an option,
  • finding more identifying data in authority records, and
  • having the potential for increased international sharing – by following the IFLA International Cataloguing Principles and the IFLA models FRBR and FRAD.”

My remarks:

– The inescapable fact is that the public will still be seeing Latin abbreviations until the end of time so long as there is no recon project. Therefore, merely changing the rule solves *nothing whatsoever* for our users, and we must be very clear about that.

– Seeing more relationships, that is, making the current relationships between records more explicit, may be useful to the public or not. According to the rest of the report, figuring out the exact relationship of an entry seems to be much more difficult for the cataloger than just making the entry in a 7xx field. I would say that this needs to be studied further to find out what the utility is to our users.
This reminds me of an actual instance in my own practice where an organization was considering adding “cause-effect” and other relationships for catalogers to encode in their subject descriptors. I gave an example where a cataloger assigns the terms “deforestation” and “desertification” as simple keywords, and comparing this to having to add them as cause-effect relationships. Either can be a cause or effect of the other, and the only way to find out is to spend additional time actually reading the document much more closely than you would otherwise, and even then, I found cases that were very unclear. I felt that the result on productivity could be quite large as much more time would be needed reading, considering, consulting, and so on, which would be in effect, making judgments that other could very easily disagree with. The ultimate utility to the user appeared quite negligible, and I argued, potentially even harmful since someone searching this could relate in entirely different ways to the information in the document. To repeat, studies need to be made to decide if the supposed gain of users seeing the relationships more explicitly is worth the necessary drop in production.

– Eliminating the rule of 3 “cuts both ways”. As it stands now, it could just as easily wind up making fewer access points available. This also needs to be made very clear.

– More data in authority records is another one that “cuts both ways”. If we add more data to authority records and actually increase the number of responsible parties in each record, total productivity must go down. To be fair however, if the additional data in authority records could be crowdsourced or even better in the current world: work with tools such as dbpedia, perhaps through tools such as Worldcat Identities, then this one could actually work. This has little to do with RDA however, and everything to do with changes in MARC format.

– having the “potential” for international sharing: a nebulous benefit once again. It would appear to make more sense that if we are to achieve increased international sharing, we should follow ISBD more strictly than less so. The authorized forms can be shared through the VIAF project, which, as Bernhard points out, is one place for interoperability. Still, in the VIAF we see very clearly how the actual *form of the name* becomes much less important than the concept (where we can see that one concept can have many forms of name), therefore the rules for “establishing the authorized form of the name or concept” as RDA does, are becoming anachronistic, since the actual textual forms are far less important than the concept URI, e.g. see http://viaf.org/viaf/35605/#John_Paul_II,_Pope,_1920-2005