Digital Libraries: Bibliographic Control Future

By Roy Tennant -- Library Journal, 4/15/2007

Standing in front of Google's main lobby, I'm unimpressed. The building looks like any other corporate building in this Mountain View business park. But I also feel like Dorothy standing before the portal to Oz. Here, I think to myself, is where all the magic happens. Here, for today at least, is where we will talk about the future of bibliographic control.

This meeting, “Users and Uses of Bibliographic Data,” is the first of three organized by the Library of Congress's Working Group on the Future of Bibliographic Control. The next meeting in Chicago, May 9, will cover structures and standards, and the last, in northern Virginia on July 9, will address economics and organization. If you're interested in our bibliographic future, you may want to consider attending, or submit your written comments to the Working Group by July 15, 2007.

Longtime readers of this column know that, in the past, I have taken our existing infrastructure for bibliographic description (i.e., MARC and AACR2) to task for being a barrier to achieving the type of bibliographic infrastructure we now require (see “MARC Must Die,” LJ 10/15/02). What I think we should achieve can be found in “A Bibliographic Infrastructure for the 21st Century” (Library Hi Tech, Jun. 2004, p. 175–181). The meeting at Google did nothing to dispel these notions, and there was much to support it.

Bernie Hurley (UC-Berkeley) and Andrew Pace (North Carolina State) brought up the overcomplexity of MARC. In fact, very few of the fields are actually used. And now, we have more than anecdotal evidence that this is the case.

With over 2000 fields and subfields, the MARC21 format has grown into a complex set of options for recording bibliographic information. Bill Moen and his research team at the University of North Texas evaluated the entire WorldCat database (over 56 million records, provided courtesy of OCLC) and discovered only ten fields and about 20 subfields were commonly used.

“When only 10–20 percent of the available MARC fields and subfields are used by catalogers,” Moen bluntly states, “what's the point of such a complex metadata scheme? Has MARBI (Machine Readable Bibliographic Information) produced something that addresses idiosyncratic cataloger needs rather than the needs of users to find, identify, select, and access resources that may be relevant to them?”

The implications of the group's work and the changes they will bring are many and significant. How we describe our collections and the resources to which we wish to point our users to will be different and in some cases dramatically so. The tools we will require and the procedures we will need to use are even now only beginning to be created. These times are challenging but also inspiring to those of us who believe that our bibliographic infrastructure is overdue for an overhaul.

I've already voiced some misgivings that the Resource Description and Access (RDA) process will make the kind of changes that are necessary (see LJ 3/15/07). We may carry too much useless baggage into our bibliographic future if we try to make the legacy data transition easy. We must take care not to make ourselves irrelevant to a growing base of users who need our advice and assistance with bibliographic description tasks.

I saw many metaphors in Mountain View that day. The meeting was held in a room that had recreational equipment (ping pong anyone?) and a kitchen area. Google staffers constantly walked around the perimeter, making espresso and talking to colleagues or on their cell phones. “You may be discussing your future,” Google seemed to say, “but we couldn't care less. You will have very little or nothing to do with the future we envision.” If we can't get our bibliographic act together, and make bibliographic control easy and effective without being overly painful for too little return, Google may, unfortunately, be right.

For more on the wired library, see the netConnect supplement mailed with this issue and with the January, July, and October 15 issues of LJ