As I mentioned in my last post, OCLC’s new policy for the use of catalog records from WorldCat (a policy now in at least its third revision) has generated substantial controversy, particularly among library bloggers.

The controversy focuses on statements that prohibit reuse (without express OCLC approval) of records contributed to WorldCat, unless the use is both “noncommercial” and “reasonable”. What does “reasonable” mean here? OCLC’s policy statement (as of November 20) specifically excludes any use that

a. discourages the contribution of bibliographic and holdings data to WorldCat, thus
damaging OCLC Members’ investment in WorldCat, and/or

b. substantially replicates the function, purpose, and/or size of WorldCat

But OCLC makes it clear that this is not an exhaustive list, and that “OCLC has the sole discretion to determine whether any Use and/or Transfer of WorldCat Records complies with this policy”. OCLC has stated in its FAQ that starting in February, it will add notices to every record downloaded from WorldCat stating that use of the record is subject to OCLC’s new usage policy.

In other words, OCLC claims the right to decide whether any application using WorldCat-deposited catalog data should be allowed to exist.

OCLC admits that their policy is intended to protect their revenue, and claims that they are acting to protect its subscriber libraries. But as plenty of other commenters have pointed out, the policy puts a big OCLC-sized bottleneck in the way of innovative use and reuse of library metadata. This further promotes the marginalization of library resources, which could easily hurt those same subscriber libraries. Ultimately, I think this policy is self-defeating even for OCLC itself, because the policy itself “discourages the contribution of… data to WorldCat”. At least it does for anyone who sees open library metadata as important, and doesn’t want confusion or licensing fights to erupt when OCLC’s terms of use are inserted into the records they contribute.

OCLC Vice President Karen Calhoun likened the new OCLC policy to the Creative Commons Attribution-NonCommercial-ShareAlike licenses, but it’s incompatible in both letter and spirit with that license, and with any other ShareAlike license. The Creative Commons license allows enhancement, adaptation and reuse for any “function, purpose, or size”. (At least in the noncommercial domain. If, as I recommend, the NonCommercial aspect is not used, then both commercial and noncommercial applications– including any that OCLC sells– can draw on and enhance the content.) The terms are straightforward, and not subject to later reinterpretation or revocation if the grantor decides that their business is threatened. Users do not have to apply for permission to innovate.

The ShareAlike aspect, furthermore, prohibits any downstream organization from trying to lock up or impose additional restrictions, such as OCLC’s, on what was given to them freely. Open data is kept open.

Opening WorldCat data might make OCLC a leaner organization that it is now; Microsoft and Elsevier show that quasi-monopolies can sustain bigger margins than most other businesses. But OCLC also sells lots of services on metadata that bring in money, and it employs smart people who have deep knowledge of library practices and needs, and who continue to bring out innovative and useful services. (To take just one example from OCLC’s many services, WorldCat Local could attract lots of subscription fees from libraries that no longer want to run their own ILS, if the service is well enough designed and run.) In the information technology world, companies like Red Hat and IBM do quite well selling services on free software, and I’m confident that folks at OCLC also have the expertise to sustain themselves with services on open data.

As Stefano Mazzocchi points out, OCLC’s new policy might buy them some time, but the Web isn’t going away, and the people (and companies) who might find library metadata useful aren’t going away either. OCLC is hardly the only organization that can create library metadata. (Indeed, they largely get it from their own subscribers.) But OCLC’s ShareAlike-incompatible license effectively draws a line in the sand: a catalog record can support the agenda of open library data, or support OCLC’s agenda, but not both.

If forced to take sides, the choice is clear to me: open library data is ultimately the best way to support libraries and their missions. If possible, though, I’d rather work with OCLC than work around them. In Part 3 of this series (coming soon now posted), I’ll suggest some possibly useful ways to respond between now and February, when OCLC’s new policy is scheduled to take effect.