A look at the Policy for Use and Transfer of WorldCat Records revision

OCLC has posted a revised version of it’s new policy for use and transfer of WorldCat Records.  This is a revision of the short-lived version of this document posted Nov. 2nd and taken down early Nov. 3rd.  Additionally, OCLC has significantly revised the FAQ’s for this document (which is important, since this is what most people will use to interpret this agreement). 




When OCLC released their initial revision of this policy (Nov. 2), a number of concerns were raised by myself and the library community.  However, of all the concerns, there were two really, really big concerns that I consider to be deal breakers with the policy as proposed on Nov. 2.  These were:

  1. The embedding of this policy into the OCLC record in the form of a 996 field essentially placing a license within the library’s bibliographic records.  While I know that OCLC didn’t feel that this was a big deal — it was.  It’s a tremendously big deal because it takes ownership of metadata away from the library, and gives it to OCLC.  And it’s dangerous — as a link to a policy, terms and rights can be changed (as we have seen between the Nov. 2 and the Nov. 5th revisions).  This is one of the reasons why licenses have versions…so rights granted to users at one period of time won’t change if a license undergoes revision.
  2. The non-compete clause that essentially barred use of the WorldCat data for the creation of services that competed against OCLC.  And while this was meant primarily to discourage Commercial development, it could very easily have applied to non-commercial research projects — if such a project began to encroach upon the size or function of the WorldCat database.

So those were my big concerns.  With that in mind, I want to look at the revised policy specifically to see how these two issues have been addressed.

The 996 field:

While OCLC’s policy and FAQ’s have been significantly revised, there are few places where this is as evident as in the areas of attribution.  In the Nov. 2nd version of the policy, OCLC’s policy stated that transferred records “must be clearly identified” and that the OCLC record number, if available, “may not be removed from any WorldCat Record”.   The Nov. 5th revision is significantly different in this area — changing “must” and “will” to “OCLC encourages”.  In essence, this part of the agreement has been made optional.  The revised FAQ’s further spell this out by saying: “It is optional for OCLC members and nonmember libraries, archives and museums to implement or retain the 996 field for attributing WorldCat as the source of WorldCat-derived records.” 

For this change, I’m definitely pleased.  Though, now that this has been fixed, I’m going to be honest, I want to see OCLC to push this even further.  In the FAQ, OCLC makes a case for attribution.  That’s fair.  However, there is still a fundamental problem with the approach being used:  that being that this license is unversioned.  This policy will change again at some point and it’s likely that some previously legal uses of the metadata records may fall outside of the boundaries of the new conditions.  By having a versioned license, previous uses of the metadata would be grandfathered.  So, if OCLC continues to want to move down this path of attribution and linking to it’s policy, then I would like to see the 996 field revised to something like the following:

Field: 996  WorldCat Record Use Policy Link
1st indicator: undefined
2nd indicator: undefined

$a: Function identifier
$i: Display Text
$v: Version Number
$u URI

By versioning the license, it would at least close this future ambiguity.  Were the versioning component added to this field, I’d feel a little bit better about it.  Good enough to let this licensing info into my library’s catalog — probably not, but at least I’d feel better about it.

The non-compete clause:

Well, this is still pretty much in the policy — though, the policy has been modified to reflect that this is primarily geared towards commercial (or third party) use.  Of course, there still remains the clause that defines “Reasonable Use” as uses (commercial or non-commercial) that:

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

2) substantially replicates the function, purpose, and/or size of WorldCat. Please see the FAQ for a discussion of Z39.50 for cataloging using WorldCat-derived bibliographic records.

While OCLC may argue that this would allow libraries to participate in projects like the Open Library, I have a feeling that that’s not the case, for two reasons.  First, Open Library does substantially replicate function, purpose and size — but more importantly, Open Library couldn’t accept metadata records that include the OCLC policy of transfer (the licensing terms of currently incompatible with those that Open Library uses last I checked).  But more importantly, I believe that this non-compete clause likely makes projects like the one being undertaken by some ILS vendors to create a type of cooperative catalog for users of it’s ILS systems (since here, we are not talking about Z39.50, but an actual creation of a network based service that would provide like a Napster-like functionality allowing users to select records from their ILS network) or stop the creation of a like-minded service by a group of open source developers that may wish to create or build a cataloging service from records from other libraries (say, from the records on Open Library for example).  Of course, if that’s not the case, it would be nice if OCLC would provide examples of this type of use within it’s FAQ documentation.

Of course, in many respects, the problem that I have with the OCLC non-compete clause is in many ways philosophical.  Libraries have always been about unfettered access, including access to our metadata, so I have a very difficult time believing that we need to somehow quickly move to protect our metadata from unscrupulous commercial uses.  I’ve said it a number of times, if the library community (and OCLC) want to view WorldCat as the Library Commons then it should be open, to anyone, for any reason…period.  These types of non-compete clauses are difficult for me because they are contrary to many of our core library values.   But I won’t rehash this since I did write an earlier thought piece on the subject ( a few days ago.

Anyway — as far as the policy goes.  I think that OCLC has done a good job working to answer some of the initial criticisms that I know that I personally had with the new policy.  Would I like to see them make additional changes…yes.  If OCLC is determined to add a license to the Commons (I guess I’ll have to stop calling it the Commons then), then they need to version the license to reduce potential future ambiguity.  Do I like that OCLC wants to make themselves a not so silent partner in my library’s 3rd party dealings…no.  Would I recommend that my library (or any other library) remove the 996 licensing when downloading data from OCLC…at this point, you bet.  OCLC can certainly make this available in the WorldCat database, but I still don’t want it in mine…but I guess there’s those darn philosophical differences again.



  1. Thank you for doing this excellent work. It has all become so contractual. There is no mention of public review, which I think was until Feb. 2009. Has that changed, do you know? I also have been very concerned about the future of library services and how OCLC is becoming the bibliographic monopoly of the world, having gobbled up any and all other services (UTLASS, RLIN, WLN, etc.). I worry that OCLC intends on becoming a for-profit agency that will eventually provide most library services, thus eliminating local control, local needs — WorldCat local is a big step into this. It’s not a wise move to allow OCLC to gain so much power over our library services. One size, the OCLC size, does not and should not ever fit all. We have have to realize the quickening steps OCLC has taken to start compiling all of our services into their products, many of which are for a fee. For-Profit OCLC in 10 years, with our implicit permission because we didn’t stop this, and their ownership of all our records and monopoly of our service needs, is something I’m afraid of.

  2. Ed,

    Yes — in the URI as well. I was thinking in the MARC subfields simply because it would make it much easier to parse on the MARC side of things — but for portability outside of MARC — you are right.

Comments are closed.