Bibliographic Control Webinar Q&A

Below are listed questions that were submitted during the October 14th NISO webinar, "Bibliographic Control Alphabet Soup: AACR to RDA and Evolution of MARC." Answers from the presenters will be added shortly. Not all the questions could be responded to during the live webinar, so those that could not be addressed at that time are also included below.

Feel free to contact us if you have any additional questions about library, publishing, and technical services standards, standards development, or if you have have suggestions for new standards, recommended practices, or areas where NISO should be engaged.

Bibliographic Control Alphabet Soup: AACR to RDA and Evolution of MARC
Webinar Questions & Answers
October 14, 2009

  1. With RDA, will MARC records eventually disappear? If so, how much longer will they be used?

    Diane Hillman: I think it will take a long time for MARC records to disappear entirely (as I mentioned in the webinar, I think there will be a place for a lossy format for specialized exchange for some time to come). But I think it's critical that we start to look forward to the really good outcomes we can accomplish with the RDA elements and vocabularies as a basis for data exchange that can be generalized beyond the MARC world.

    Barbara Tillett: RDA is intended to be usable with any schema or display, so having RDA does not mean anything about the MARC format itself. However, RDA does provide mappings to MARC 21 and to other schemas so people can see where the elements should go in each schema. The first release of RDA will have just a few mappings (ISBD and MARC 21 in RDA Appendices D and E, possibly MODS/MADS and Dublin Core mappings as links), but we expect more to be added in the future.

    Most ILS vendors today do not use MARC 21 internally, but just for import/export of bibliographic and authority (sometimes holdings) data. As long as OCLC and ILS vendors still require that format for exchanging data, it will remain, but already many are using ONIX XML, MARCXML, and are able to handle MODS and MADS in XML. IFLA is creating a schema for the ISBD elements in RDF XML, and it is likely that the RDA: Resource Description and Access elements also will be available as a schema in XML. The mapping of MARC 21 data to future XML schemes, as Diane says will undoubtedly lose some data in the conversion/transfer process, but mostly will still be usable. The MARC 21 structure has limits as to what relationships and links can be made.

  2. How are vendors being encouraged to better take advantage of RDA and FRBR?

    Diane Hillman: My consulting partner (Jon Phipps) and I are working individually with as many vendors as we can who might be interested in exploring the possibilities with RDA and FRBR. The difficulty we have is one that extends to the library community as a whole: they hear "RDA" and they think "cataloging rules" and " extra costs," not necessarily "new possibilities." We need to change that before we start seeing progress.

    Barbara Tillett: The ILS vendors have been invited to follow the activities of the US RDA test in 2010 and the US national libraries also have been encouraging the vendors to discuss future plans and ideas to take better advantage of RDA and FRBR. Each of you may also wish to pursue such efforts with your own vendors.

  3. Are RDA and FRBR currently being taught in any Library Science programs? How much longer will MARC need to be taught?

    Diane Hillman: I know I'll be teaching an online class for the iSchool at the University of Washington starting winter quarter, but I don't really know much about what other schools are doing. I think MARC will continue to be taught so long as we're exchanging MARC data, but not, perhaps in the same way we do now.

    Barbara Tillett: FRBR has been taught for nearly a decade now in many library schools (it appeared in 1998) and RDA isn't even published yet, so it's a bit early to teach it, but there are several US library schools who will participate in the US RDA test next year.

    As for continuing to teach MARC 21, it's tied to an extremely important part of cataloging history, so I expect there will be coverage of it for a long time to come, but hopefully catalogers won't need to continue to learn the coding in MARC 21 for too much longer.

  4. Is OCLC Connexion going to be modified so that catalogers are inputting RDA and FRBR elements into their catalog records? Is there a timetable for when that would happen?

    Diane Hillman: You might want to ask OCLC those questions directly at the free Webinars they're sponsoring later this month!

    Barbara Tillett: OCLC is also participating in the US RDA test, so will be preparing for that soon.

  5. [For Bill Moen] How do the 8 areas of description figure in your research? How was it accounted for?

  6. [For Bill Moen] Did this study take into account the create date of the records compared to the publication of the core standards? It seems to me a lot of the records were created before the core standards were created.

    Barbara Tillett: Very important point! It also seems odd that we'd want to base a decision on what future elements are important to our users on how we did things in the past (that's why I asked if they were planning user studies to test the FRBR decisions, which were based on knowledge of the user studies conducted prior to 1998, on expert cataloger input, and on ISBD decisions). I also am not clear why some of the fields like the 656 (occupation) only occurred once - when many of LC's special collections put that field in most of their records - the statistics from our "BooksAll" Retro file for 2008 shows 11,732 total occurrences of field 656 in 6,501 LC records - so it's not clear what the universe of records was in 2005 or whenever this study was conducted. Hopefully Bill can enlighten us.

  7. One reason MARC is so large - it has grown over the years to accommodate changes and desires of many communities of catalogers that use it. How will the RDA vocabulary respond to inevitable requests for changes?

    Barbara Tillett: As I responded in the Webinar, the governance structure for the Committee of Principals, which oversees the work of the Joint Steering Committee for Development of RDA and the publishers, will be changing this coming year, and it is not yet clear how we will manage the involvement of a more international participation in the future development of RDA, but it is hoped with a Web tool that we will be able to be much more responsive. Most of the RDA vocabularies are not closed lists and intended to be adjusted over time as new things emerge that need to be described, so we hope the Web tool will incorporate a mechanism for submitting suggested changes. I just don't know yet what will be in the first release for submitting changes.

  8. [For Bill Moen] What are the 7 fields that appear in every book record in WorldCat?

  9. [For Bill Moen] Any plans to do user studies? They are very difficult, but desperately needed to guide our "core" elements needed.

  10. Is RDA going to be better suited for cataloging realia than MARC and AACR2?

    Barbara Tillett: I hope so, but also RDA defers to some specialty instructions like CCO (Cataloging Cultural Objects) when libraries feel they need further data for their descriptions than the general descriptions that RDA would provide. We hope the instructions would be compatible and will work with those special communities to maintain basic principles in all of our standards.

  11. What's the relation between RDA and the Bibliographic Ontology?

    Barbara Tillett: None so far that I know of, but I'm interested in work in this area. As I understand it, Bibliographic Ontology is a grassroots effort that Bruce D'Arcus started up, which some other people got interested in -- including Ivan Herman of the W3C. Bibo basically pulls together some other vocabularies DC, FOAF, and some others, and creates a few new vocabulary pieces to allow people to encode bibliographic citations in RDF.

  12. Did Dr. Moen's research control for EncLvl when gathering and analyzing the data?

  13. Three questions for Diane Hillmann:

    1) In "Why Not MARC" are you talking about MARC 21 syntax (i.e., ISO 2709) or MARC element set? Note that the MARC element set can also be modeled as RDF/OWL.

    Diane Hillman: Certainly the first, but also the second-I think it's difficult to deal with the two questions as fully separate issues. One difficulty is that MARC is designed for flat records, and doesn't translate easily into the kind of data structures and encodings used outside libraries. We've seen lots of attempts to add URL's into MARC, as an example, but it's not been entirely successful as a strategy. There are other limits (number of subfields per tag, for instance) that also bedevil MARC. With the legacy of MARC data out there, overhauling the MARC element set is a difficult process, kind of like gutting an old house while still living in it (which I've attempted on a room by room basis, so I'm not just speculating here!) I do take your point about RDF/OWL, but I don't think it's a particularly satisfying strategy, and there's not much enthusiasm for it in the community, as far as I can see.

    2) What does the bibliographic format of the future look like for institutions NOT using RDA? MARC has always been independent of a particular cataloging code (although it was developed originally and often associated with AACR). An example is CCO (Cataloging Cultural Objects).

    Diane: I think that there will never be "one" bib format that works for everyone, but the really interesting future includes the idea that we don't have to be limited to just one-we can pick and choose what we use depending on what our needs are. Libraries already use a variety of formats (and variations) but still haven't addressed data exchange issues around that variety, which the strategy for RDA elements and vocabularies is designed to do. CCO is not strictly a bib format, of course, but I think it could really benefit from the kind of future-oriented strategy (registered and openly available RDF vocabularies, primarily) that RDA has adopted.

    Barbara Tillett: This question seems to be confusing the format/schema for communicating bibliographic data with the cataloging instructions - the cataloging instructions are independent of any display format of any encoding schema. Even existing AACR2-based data can be encoded other than in MARC. One can put bibliographic data following AACR2, the CCO instructions, or the RDA instructions into MARC 21 records or an XML schema or just about anything. There surely will be opportunities for developers of DACS, CCO, the DCRM's, and other instructions to work with the developers of RDA towards a shared principle-based structure and one that shares the same conceptual model so they are all compatible into the future. Each specialized community will want their own more granular approach for their special audiences.

    3) How will this new RDA format interact with holdings data in libraries? Using MARC for different kinds of information has allowed us to integrate data (i.e., bibliographic, authority, holdings, classification, community information).


    Diane Hillman: Holdings data has always been an interest of mine (I spent my early career as a serials librarian and law librarian). Unfortunately RDA doesn't really cope well with the holdings level we've been used to in MARC. I know that the Extensible Catalog Project has been trying to address that as they build services to transform legacy MARC data into RDA, but holdings is clearly an area ripe for attention. I think the other categories of information you mention will be addressed (or are being addressed) with reference to the broader world of data-I would suggest that the MARC notion of a "family" of integrated bibliographic standards may not be ideal in a world based on broader structure and encoding standards. LC's new website for LCSH, for instance, uses the Simple Knowledge Organisation System (SKOS) for concept vocabularies, as does OCLC for their new top level for DDC, and the NSDL Registry. FRAD is coming along, which will address name authorities. I've always been a fan of the MARC Community Information format, but sadly it hasn't seen much uptake. One issue is that it's not particularly "bibliographic" in nature, which leaves the future for it a bit uncertain. I'd love to see it "re-discovered" and re-conceived for a webbier world!

    Barbara Tillett: RDA is not a format. It's a set of cataloging instructions on the identifying characteristics for the things we wish to control in the bibliographic universe. If you still are using an ILS when you start describing things according to RDA, you'> d still be using MARC bibliographic records and MARC holdings records (and MARC authority records) for the foreseeable future.

  14. If/How is RDA going to increase cataloging efficiency?

    Diane Hillman: Just as the original MARC format increased efficiencies for libraries by broadening the availability and re-use of catalog records, so RDA, by looking beyond libraries for data and sharing opportunities, has the potential to increase efficiency for libraries yet again. I think we've reached the limits of what we can do with MARC and data created almost entirely by humans, and if we've been watching closely to see what communities outside our world have done, we can see that there is potential for us both as providers of data to others and users of data that others are better positioned to provide. Consider the Geonames service at http://www.geonames.org/, where libraries could easily find and use latitude/longitude data to provide mapping services instead of only the human readable geographic name strings currently provided in MARC data. Thinking about how we might use readily available services like this in a more modern data environment, given the fact that we're all facing economic challenges and the need to re-think our use of precious human resources, is what the move to RDA is really all about.

    Barbara Tillett: The major benefits that we will realize with RDA will come with the development of new systems for cataloging in a FRBR-based system - imagine being able to link the name of a creator to the description you are providing for an item by just pointing to the things you want to connect. Imagine getting most of the "transcribed" data directly from the publisher (as we now do with ONIX data) and moving on to machine-assisted suggestions for type of content, type of carrier based on clues in the item or linking to authority data in a system like VIAF where you just link to the identifying URI for the entity and do not have to worry about the form/structure of the name (your system would know which form you wished to use as a default to display to users, based on your preferences and the prioritized distinctive data needed to differentiate one entity from another); or machine-assisted suggestions for subject terms and classification numbers that a human would confirm or adjust. Imagine incrementally adding to a global web of bibliographic description - no more creating individual bibliographic or authority "records" over and over by all libraries, but instead adding data to further identify and connect the things in our bibliographic universe, where everyone has free access to the shared, linked data. We have the technology to do all of these things now…so where are the systems?

  15. Bill: Have you received feedback from catalogers about the data? Do they express surprise to learn how as a group they use MARC? Is it possible that a broad awareness of their decisions will help them look beyond MARC?

  16. Could Bill please comment more specifically on the use/non use of MARC fields in PCC records with reference to the new BIBCO Standard Record?