OCLC Defends Records Policy, Faces Questions, Suggestions, and Criticisms

Norman Oder -- Library Journal, 1/27/2009

• Lively discussion at Midwinter Meeting
• OCLC's Karen Calhoun defends intent, apologizes about communication
• Others question OCLC’s path

On one level OCLC’s recently revised—and suspended—policy regarding record-sharing, aiming to "modernize record use and transfer practices for application on the Web, foster new uses of WorldCat data that benefit members and clarify data sharing rights and restrictions,” was simply a matter of bad communication, a cooperative behaving in top-down rather than consultative fashion.

That, OCLC VP Karen Calhoun could easily acknowledge. However, in much of her presentation Monday at the American Library Association (ALA) Midwinter Meeting in Denver, she firmly defended the intent of the policy, suggesting that critics in the blogosphere had an unrealistic view of the library ecosystem. In response, some panelists suggested that OCLC itself was failing to modernize.

She apologized, on behalf of the study group and OCLC, for not taking the “value of participatory decision-making nearly seriously enough” and said the review process will address not just the policy and process behind it but also what community norms should be in place.

A revision is expected by the third quarter of 2009. Calhoun acknowledged that “OCLC is caught, along with many other organizations, in this painful transition,” one in which new business models emerge and potential competitors like the upstart ‡biblios.net provide similar services.

Shared community asset
Calhoun said the factors driving the policy change included documented efforts to systematically download and copy the entire database; commercial entities reselling records without helping support the shared data creation system; and the need for a legal mechanism to encourage organizations outside the cooperative to negotiate with OCLC. 

She likened OCLC records to a swimming pool, a shared community asset supported by the community at large but open only to those who contribute. “Up to now, the value of the shared system and data and methods of recovering costs have been viewed as fair and reasonable,” she said. “But many organizations, including OCLC, have reached a painful time of transition,” she said. “They have business models based on value in creation and control… [but] need to move to a model based on value in the exchange of linking of data.”

She cited the WorldCat API and WorldCat.org, both aimed to link OCLC partners like Google Book Search to OCLC members’ collections. The intent behind the revised guidelines was not to “change so much what OCLC members do but to ensure that use of records benefits the OCLC cooperative as a whole,” notably reaching agreements with organizations outside the cooperative.

No heroes, no villains
OCLC’s study group noted that the prevailing opinion in the blogosphere is that data should be free and open. However, she noted, “nearly every organization has terms and conditions for data sharing.” (She cited a presentation at IFLA.)

“Some thought leaders would have us believe that there are heroes and villains,” she said, citing the maxim, “Other people’s data should be free.” She instead argued “that there are few heroes and fewer villains… There are merely organizations responding to the reality of the situation.”

A typical comment she’s heard, she said, is “I paid for the records and they are mine to do as I please.” That, she said, suggested “relatively low awareness and support for the norms of the OCLC cooperative.” Rather, records belong jointly to the members that contributed them, and to OCLC, she said, suggesting that, of records in a typical OCLC library’s catalog, the member library contributed about two percent.

The OCLC policy aimed to reduce operational costs, expose library data and collections as much as possible, and drive to libraries from popular web sites. She said that collaboration, as with Google Book Search or Yahoo, is rarely about exchange of funds but usually about the exchange of value.

“So the record use study group faced a very tough balancing act,” she said, aiming to make WorldCat as open as possible while assuring that use outside the cooperative would provide a fair return and protect members' investment in OCLC data, infrastructure, and services. “This has been very, very hard to do.”

The revised policy is framed as a legal document, Calhoun noted, stating that the “intent is not so much to change what OCLC members do but to ensure that use of records outside the cooperative benefits members. Unfortunately, the voluntary nature of present guidelines has impeded accomplishment of that goal.”

Missed opportunities
John Mark Ockerbloom, of the University of Pennsylvania Libraries, suggested the policy posed another cost, that of missed opportunities for librarians to create new features, for example, expanding on work he’s done on subject maps. He noted some tough language in the policy, which cited “reasonable use,” and noted that the policy was subject to unilateral change by OCLC.

He suggested, by contrast, taking cues from the world of open source, which does include terms and conditions for sharing. He acknowledged that he hadn’t done a fiscal analysis. “There is a significant cost to coordinating a high quality knowledge base,” he said, “but there’s also a substantial cost in keeping our metadata closed off from full access and use”—a cost that is hard to quantify.

He suggested some support strategies from other products: keeping overhead low; charging for related services; and gaining direct support from members or partners. “We could acknowledge that substantial portions of catalog records are in the public domain,” he said, and not forbid their use or dissemination.

OpenLibrary competition?
Peter Murray, of OhioLINK (and author of the Disruptive Library Technology Jester blog), pointed to growing competition for records, including OpenLibrary, a project of the Internet Archive that aims to create one web page per book. And discovery methods are changing. “Google is pointing to another way about getting information about these surrogates… relying on the text of the item itself.” 

He suggested ownership of records was a sticky issue, given the distinction between a recitation of facts and evidence of creativity, such as subject headings and abstracts. “Lastly, we also have user-generated content,” he said, citing such things as tags, reviews, and comments. “All these are being mashed up into one very messy surrogate when you’re looking at rights involved.”

One way to handle the messy situation is to put the data out in some kind of open data license, he said, nothing that “anything other puts up roadblocks.” Otherwise new discovery layers, mixing and matching records from various sources, could be stymied, as could subject-specific portals, he said..

Speaking for himself and not his employer, Murray said of OCLC, “Cooperatives need to be less risk-averse in trying new things.” He said he saw OCLC, in the proposed policy, as trying “to mandate that OCLC be in the middle.” However, he said, “that may not be the most effective place for OCLC as we make use of bibliographic data. We need to be able to use these records in ways that make most efficient use of our backroom operations.” 

Cooperative vs. community?
University of California, San Diego librarian Brian Schottlaender pointed to a potential tension between community and cooperative, with the latter a more formal entity, and it may be that “the trust norm” is more typical in a cooperative.

Ockerbloom, he noted, “pointed out quite rightly a distinction between telling people what you’re going to do with stuff you’re attributing to them and asking for permission,” that the request for permission was itself a roadblock.

From the audience
No one in the audience was particularly happy with the policy; one called it draconian, pointing out that the accompanying FAQ was comparatively permissive. Calhoun clarified that the FAQ was indeed part of the policy, though some seemed skeptical.

As for whether the revised policy stifles innovation, Calhoun said that “has been widely misinterpreted.” The aim was to make it easier to say yes to organizations that do business with OCLC. She gave the example of a request to create a portal of everything written about Martin Luther King. Some people have made requests, and OCLC has said yes.

The review board, she noted, is not charged to write a new policy, but to “provide recommendations for principles… how should the policy function as a mechanism for formalizing trust.”

Murray observed that, despite OCLC’s stated good intentions, “there’s no reporting mechanism back to the cooperative. We don’t know what’s been approved, and why.” Calhoun agreed it could be an addition to the policy.

OCLC’s future
Murray said he was “fearful for the continued health of OCLC,” citing the recent emergence of ‡biblios.net, which “may not be what OCLC is, but it’s good enough.” Some customers “may start looking to it as a source of their descriptive surrogates,” putting OCLC at risk. (See story on ‡biblios.net.) 

Where do ‡biblios.net records come from? Universities, mostly. “If OCLC wasn’t there, ‡biblios.net would have to absorb the costs,” she said. “Take it all to the end… the ironic outcome would be the cooperative would collapse, cost of data would go way up, and ‡biblios.net would have to support the cost of services needed to create that repository.”

Murray suggested, “if OCLC wants to sustain itself as business model… make use of the aggregate to make best of breed services, that’s OK… but the underlying records are either facts or member-contributed records” that OCLC may not own.

One audience member noted that her library had upgraded 37 percent of its records, and asked OCLC to consider a redefinition of original cataloging. Calhoun said OCLC is reviewing the issue.

Indeed, OCLC is opening up, debuting an "Expert Community Experiment." “Next month, at the request of members, OCLC is opening the database to much more enhancement,” Calhoun said. 

While previously “a kind of priesthood” governed replacement master records, as the database has grown, collaboration has become more important, creating a “more Wikipedia-like environment” giving experts more of a role in taking care of the database.

Read more Newswire stories:

With Economy Sputtering, ALA Midwinter Attendance Dips Sharply

As ‡biblios.net Emerges, a New Opportunity for Catalogers (and Competition with OCLC)?

LITA and Top Tech Trends, Part 1: Panel Ranges Far and Wide Over Library Tech Issues

LITA and Top Tech Trends, Part 2: Digital Outreach a Model for ALA?