So OCLC has released a new version of the new policy for record sharing.
Just like CC and Open Source? Still “No.”
Difference in intent
Karen Calhoun suggests that the new policy was based on the Creative Commons CC-BY-NC license, suggesting that OCLC was going for more open-ness along the lines of CC.
But there remains one very big difference between the CC-BY-NC license you used as a model, and the actual policy. Your actual policy requires some recipients of sharing to enter into an agreement with OCLC (which OCLC can refuse to offer to a particular entity). The CC-BY-NC very explicitly and intentionally does NOT require this, and even _removes_ the ability of any sharers to require this.
This is a very big difference, as the entire purpose of the CC licenses is to avoid the possibility of someone requiring such a thing. So your policy may be like CC-BY-NC, while removing it’s very purpose.
Legal ability to enforce based on different assumptions
There is another thing to wonder about. The very ability of a CC-BY-NC to be legally binding is predicated on the assumption that the entity offering that license holds copyright to the work being licensed. If I don’t hold copyright on a work, the legal justification for being able to require the terms of CC-BY-NC is gone. This applies both to things that someone else might own copyright on, or to things that nobody might own copyright on.
The data held in Worldcat may be either or both of those things. It’s unclear if copyright can be held on cataloging records in the first place. If it can be, then it’s held by those who did the original cataloging, since as far as I’m aware OCLC members have never transferred copyright or other IP rights to OCLC.
So while the CC-BY-NC license takes something that you would have no rights (or rather, only fair use rights) to use without a license, and gives you (very broad) rights to use it—the OCLC policy takes something which OCLC has questionable Intellectual Property ownership of in the first place, and which third parties might have been able to use as they like without permission, and applies requirements (much more strict than CC-BY-NC) to them by virtue of a (click-through style) contract which OCLC members must agree to use OCLC.
Very different thing. Apart from whether we like the new policy (some parts are good, other parts less so), I think comparing it to CC licenses in an attempt to garner some of the good will that CC licenses have–is a mis-direction.
Changes from last draft
What had been a requirement to include attribution and license restrictions in any records shared becomes instead a reccommendation/encouragement. This is good. It changes things to the spirit of partnership of OCLC members working together to meet OCLC goals, rather than what before seemed like an attempt for OCLC to aquire psuedo-ownership controls by contract that they didn’t have by actual ownership.
The new contract creates new categories/classes of people. OCLC members, libraries/archives/museums that are not OCLC members (called ‘non-OCLC members’, but confusingly this only applies to library/archives/museums), and everybody else (‘third parties’). The previous draft placed less emphasis on whether an entity was a library/archive/museum or not. So, while before, it seemed pretty clear to me that this policy authorized sharing with the Internet Archive, now it’s less clear. The Internet Archive considers itself an archive and a library, but does OCLC? The policy doesn’t have definitions of who OCLC considers a library/archive.
Of course, whether or not it’s a library/archive, it could still be restricted if OCLC believes the Internet Archive or it’s project(s) are an “unreasonable” threat to OCLC’s business. The difference is that if it’s not a library/archive, they require pre-approval before transfer from OCLC in the form of an agreement, in which OCLC may demand payment or concessions, or refuse to grant permission altogether. If they are not a library/archive, no such pre-approval is needed, although it seems OCLC could still decide after discovering it that it was an “unreasonable” use and forbid OCLC members from sharing with them.
Before this new policy, my understanding is that OCLC members asked for clarification from OCLC as to whether sharing with the Internet Archive would be allowed, and OCLC pointedly chose not to give it. They didn’t say ‘no’, but they wouldn’t say ‘yes’ either.
Also, as ecorrado notes, the new policy seems to say that for us to allow our patrons to connect to cataloging records for inclusion in EndNotes or RefWorks etc (both have features that allow automatic import of catalog records via z3950), each individual patron would need to sign an agreement with OCLC! After all, each individual patron is neither an OCLC member, nor a library/archive/museum capable of being in the confusingly named “non-OCLC member” class. Presumably this was unintentional on OCLC’s part, and they could fix it by making clear that not only staff, but also patrons/members of any library/archive/museum are considered within it, not an external party. But, so I have to forbid people at another university from pointing their EndNote or Refworks at my catalog? Still not so good.
These confusions are inevitable as long as you’re trying to put such restrictions in the licences (or ‘policy’). This is not a Creative Commons style license, as CC licenses are pointedly written to eliminate such confusions by removing restrictions not adding them.
More open questions of legal basis
The remaining open question is on what legal basis OCLC can enforce this ‘policy’ anyway. As mentioned, if copyright can be claimed on cataloging records at all (very unclear), most of the copyright for records in WorldCat would not belong to OCLC (and would have very confusing line of ownership in general, for records touched by more than one library–if copyright exists in cataloging records at all.). (I believe that OCLC employees or contractors do do some amount of original cataloging. Anyone know what percentage of WorldCat might be such?).
Clearly, as a legal matter, even without copyright, OCLC can impose obligations on parties it has entered into contract with, as a matter of the contract. While they call this a ‘policy’, the only way it seems legally enforceable is if it is in fact a contract. Whether it’s a legally enforceable contract, whether OCLC members have actually entered into that policy as a contract, is an open question too, but OCLC could easily fix that if they wanted to admit it was a contract/license, and require members to agree to it.
The original version of the new policy, with it’s obligations to include attribution and licensing in shared records, copying from the ‘viral’ nature of open source/CC licenses was an attempt (intentional or not), to spread these obligations to third-parties who had no such contract with OCLC. an OCLC member has a contract with OCLC, and when sharing records with a third-party, that OCLC member is required to make that third-party agree to the same contract with OCLC regarding record distribution.
But after removing those and making it optional, that “viral” license isn’t there anymore. So on what legal grounds does OCLC think it can enforce this policy against third-parties who got records from a non-OCLC source? “Regardless of the source from which WorldCat Records are received, Use and Transfer of WorldCat Records is authorized solely by OCLC pursuant to this Policy.” Why does anyone need OCLC authorization in the first place, if they have no copyright claims to this data?
It goes without saying that OCLC’s attempts to make ‘third parties’ enter into agreements with OCLC to receive any OCLC records from anyone, is another attempt to enforce by contract what OCLC does not have rights to by ownership. A perfectly legal thing to do, mind you, content and software vendors do it all the time. Note that OCLC can legally require those who’ve entered into a contract with it (members) to share records with third parties only if the third party has a seperate contract (“agreement”) with OCLC, but thats’ a legal requirement on the _member_, not on the third-party. If the third party does get the data in some other manner, why do they need OCLC authorization? What’s the legal basis? ]
OCLC’s Business Model, and the Future
I think OCLC is trying to do the right thing here. They are trying to provide for more open sharing of records to meet 21st century internet needs, while still doing what they think they need to do to protect their business interests. And they think their business interests protected only by aquiring psuedo-ownership like rights of the data in WorldCat (which they do not actually have ownership rights for). And indeed this isn’t new, the 1986 policy does the same thing.
The problem is these two motivations conflict. There’s really no good way for OCLC to protect these psuedo-ownership rights without interfering with the kinds of uses of their records that we (OCLC’s member/owners) need–and have it all be legally enforceable too! One hint to this can be seen in the Creative Data Commons effort. Originally Creative Data Commons wanted to include “BY” (attribution) and “NC” (non-commercial) restrictions in their license, but they realized, after much work, that the strange (and internationally inconsistent) nature of intellectual property over _data_ (arguably like cataloging records are) made this pretty infeasible to do in a way that was both reliably legally enforceable and didn’t interfere with the planned use cases for open access data. So the Creative Data Commons license ends up being a dedication to the public domain, and relinquishing of all rights.
But do OCLC’s business interests really require controlling psuedo-ownership rights over WorldCat in the first place? I’d argue no, and, more than that, that if that was neccesarily OCLC’s business model–then OCLC would be doomed, because it’s not tenable. And us libraries would be doomed, because it doesn’t allow us to do what we need with the data.
OCLC keeps insisting that the reason their high fees are justified is all of the value-added services they offer, and the huge infrastructure neccesary to support these services. They never mention a monopoly on the data as the reason to justify their fees. So, start acting like that. Free the data. If you’re right, we’ll keep paying you, because your infrastructure and services are worth it. If someone else can provide the infrastructure and services we need at a cheaper cost–well, that’s competition. You know, the thing you have when you don’t have a monopoly.
[ To be fair, many of our libraries, including my hypothetical fantasy library in my head I’m in charge of, would be willing to pay a reasonable premium to OCLC precisely because it’s a cooperative, in theory owned and controlled by it’s customers. There’s a value to that. But the value depends on it actually being true, on OCLC acting like it. The more OCLC squanders that good will by not acting like a cooperative, the less anyone will be willing to pay that hypothetical premium. Hint: Actually publisizing a draft of this new policy for comment from OCLC members would be one way to actually act like a member-controlled cooperative. Drafting it in secret and releasing it only as a fait acompli is not.
Although the tough financial times don’t help either.]
What can libraries do?
Well, for starters, libraries that are OCLC members can start making the above case to OCLC about what it’s business model should be, what business model serves the interests of it’s members. Do any of our administrators get it?
But also, very much of the data in Worldcat that OCLC is trying to exersize pseudo-ownership rights on to preserve it’s monopoly–came from libraries, the members, you know the same ones it’s promulgating restrictions on data use to.
Any library that does original cataloging can make it a matter of policy to release this original cataloging under an Open Data Commons dedication/license. Whether submitting the cataloging to OCLC, or anywhere else. Add your own 996 explaining that the data is released under Open Data Commons.
Open Data Commons license is written specifically to acknowledge that it’s unclear if copyright holds for the data being licensed, or if other intellectual property rights might, and that it might vary in different jurisdictions–but to release the data into the wild as public domain open access regardless of how those questions may be resolved in a given jurisdiction.