No Discussion Discussion on RDA

Saturday 10:30-12:30: ALCTS/ Electronic Resources Interest Group

[NOTE: I’d promised to blog this session for the new ALCTS NRMIG Metadata Blog, but since one person at ALCTS must approve all potential bloggers and has no sense of urgency about it, I’m posting it on the LITA Blog. Maybe ALCTS will get their act together for Annual?]

The description for this session was pretty tempting:

“If RDA is attempting to be “all things for all,” is it accomplishing its goal with Dublin Core, LOM, and visual resources communities? We have heard from many perspectives, commentaries to RDA from “traditional cataloging” communities, but what is it to different metadata communities? We have very little direct information from those communities how they think this content standard will work for them. This discussion will address these issues.”

I was originally asked to participate in this panel, but declined because of a conflict with MARBI. Once the MARBI agenda came out and I realized that the issues I was interested in wouldn’t be discussed on Saturday, I decided to come to the presentation and see what came up at the discussion.

The first presenter was Barbara Tillett, from the Library of Congress. A little bird told me that she was not originally invited but asked to be on the panel (and who could turn down Barbara?) [NOTE: Barbara has asserted to me that this isn’t so.] Her presentation was pretty standard RDA high-level marketing, which anyone who has been attending the regular RDA Forums has heard before. The problem is that like all of the RDA marketing materials, it is very well crafted and its points are extremely difficult to argue with at the level presented. It’s essentially “comfort food”—without sufficient nutritional content to address the questions that the IG intended to discuss, designed to make the masses feel better about what’s going on and have faith that it will all turn out well in the end. One would think, had one been living in a cave for the past six months, that there had been no sharp questions about the RDA effort asked on multiple discussion lists, listening to Barbara’s presentation. And it was far too long.

Among the last slides in Barbara’s talk was one that was very telling about the approach the RDA effort is taking. The statements on that slide:

“No significant changes to existing records will be required”
“Need for retrospective adjustments when integrating RDA and AACR2 records will be minimal”

This indicates fairly clearly that the more significant calls for change in approach to RDA are not being heard—very discouraging. Barbara’s reassurance that this change would be nothing like the upheavals associated with AACR2 over 25 years ago was very interesting. Could it be that the conservatism we’re still seeing in the RDA work harks back to that historic period?

Murtha Baca (from the Getty) spoke about trends in the art and cultural heritage communities. These communities have cobbled together a group of metadata standards (both “content” standards like Cataloging Cultural Objects (CCO) and metadata schemas like CDWA Lite and VRA Core) to accommodate the wide variety of materials they must describe. It’s important to note that one reason that there are so many standards in this context is that AACR2, traditionally focused on books and printed works, has never met the needs of this community. As a result, they have done what many specialist communities have done—given up and moved on. That they’ve accomplished this task so effectively is both laudable and ultimately problematic, since they, too, are finding it necessary to emphasize efficiency and user focus, goals not so easy to accomplish without coordination with the larger community.

Murtha noted some differences with the library community in regard to FRBR definitions of works and the fact that this community is often dealing with unique items rather than published items available broadly. She also emphasized the life cycle aspects of metadata creation in cultural institutions, but this suggests that the issues of functional conflict when information acts as both “inventory” and “user access tool” are also inherent in their approaches. Murtha’s last slide was a winner, one she got from a colleague: “Standards are like toothbrushes, everyone agrees that they’re a good idea but nobody wants to use anyone else’s.”

Sarah Shreeves (University of Illinois at Urbana-Champaign) spoke primarily about shareable metadata, and her work with others on guidelines to enhance the reusability of metadata. Her definition of shareable included:
* Is quality metadata (she references the metadata quality chapter written by me and Tom Bruce in “Metadata in Practice”—for which I thank her!)
* Promotes search interoperability
* Is human understandable outside it’s local context
* Is useful outside of its local context
* Is machine processable

These criteria are really important—and the ideas behind them are quite new to most catalogers, who tend to have an entirely different notion of quality based primarily on experience with MARC databases. These notions of quality have yet to penetrate the RDA effort, though I’m not sure how many people noticed that disparity, and Sarah didn’t flog the issue.

She shifted focus to the DLF effort, which now mandates MODS. I was puzzled by the fact that she said that the DLF effort decided to mandate use of MODS because of frustration with the use of OAI with Simple Dublin Core. Since I know that Sarah of all people understands that Dublin Core is far more than the Simple DC 15 elements which OAI mandates as a base, I found this statement disturbing, as it tends to proliferate the confusion about Dublin Core that has been problematic for those of us attempting to teach librarians about metadata standards.

Mary Woodley (Cal State Northridge) had the unenviable task of wrapping up the presentations prior to the promised discussion. She approached her task from the point of view of the traditional SWOT model (Strengths, Weaknesses, Opportunities, Threats) and another model suggested by Joseph Busch: ROT (Redundancy Obsolete Trivial). Mary spoke from the viewpoint of a cataloger who also spends a lot of time training students in research methods and helping them at the reference desk. She described the information seeking behavior of her clientele and suggested that if we don’t take a more drastic approach to change we will become irrelevant—a concern that has been expressed frequently. As part of her focus on user needs, she suggested that the RDA effort would benefit from attention to research studies on user expectations and behavior. Barbara tried to argue that standards don’t develop from user studies, but, on the other hand, the FRBR model talks about user needs but not in enough detail to provide much guidance for RDA development (a point I would have made, if I’d had the chance).

But at the end, the discussion never happened—there was no time, given that an extra presentation had been added at the last moment. There were a few comments and questions asked in the remaining 10 minutes or so, but during the last few minutes, a disgruntled public librarian took the floor and ranted about changes she didn’t like, and the rest of us packed our bags and left.
Addendum: The following was received from Allene Hayes, the coordinator of the session:

I want to correct the misinformation about Barbara Tillett. As the Chair of the ALA ALCTS ERIG, I personally invited Barbara to be on the panel. Barbara accepted, even though to do so added to her already full Midwinter schedule.

I think all panel members did a great job and I’ll take this opportunity to thank Barbara Tillett, Murtha Baca, Sarah Shreeves, and Mary Woodley for a job well done!

I would also like to thank you and everyone else who attended this info packed session.

For those interested, all presentation slides will be posted on the ALCTS ERIG Web page at:


  1. Sarah Shreeves

    Diane –

    About my reference to the DLF Aquifer initiative using MODS because of ‘frustration’ with simple Dublin Core, I did originally mean to talk a little more about qualified Dublin Core and application profiles on the previous slide (my slides are available at: http://ideals.uiuc.edu/handle/2142/204), but with the limited time I cut that. Perhaps I assumed that the audience had more familiarity with the range of options in Dublin Core than I should have.


  2. Diane Hillmann

    Thanks for the comment Sarah–it puzzled me because I knew this was something you understood quite well, but the explanation makes me feel much better … 😉 I guess I’ve learned over time that it never pays to assume that people know that Dublin Core is more than 15 elements, so probably indulge in a bit of overkill making that point over and over …


Comments are closed.