CC-By documents cannot be re-enclosed if their publisher is acquired
October 29, 2014
Just a quick post today, to refute an incorrect idea about open access that has unfortunately been propagated from time to time. That is the idea that if (say) PLOS were acquired by a barrier-based publisher such as Taylor and Francis, then its papers could be hidden behind paywalls and effectively lost to the world. For example, in Glyn Moody’s article The Open Access Schism, Heather Morrison is quoted as follows:
A major concern about the current move towards CC-BY is that it might allow re-enclosure by companies […] This is a scenario suggested by assistant professor in the School of Information Studies at the University of Ottawa Heather Morrison. As she explains, “There is nothing in the CC BY license that would stop a business from taking all of the works, with attribution, and selling them under a more restrictive license—not only a more restrictive CC-type license (STM’s license is a good indication of what could happen here), but even behind a paywall, then buying out the OA publisher and taking down the OA content.”
This is flatly incorrect.
Reputable open-access publishers not only publish papers on their own sites but also place them in third-party archives, precisely to guard against doomsday scenarios. If (say) PeerJ were made an offer they couldn’t refuse by Elsevier, then the new owners could certainly shut down the PeerJ site; but there’s nothing the could do about the copies of PeerJ articles on PubMed Central, in CLOCKSS and elsewhere. And of course everyone who already has copies of the articles would always be free to distribute them in any way, including posting complete archives on their own websites.
Let’s not accept this kind of scaremongering.
Sauropod Vertebra Picture of the Week 
October 29, 2014 at 9:55 am
I’d not heard that objection but it seems strange to me: like refusing food on the grounds that there might be a famine later.
Conversely, if a paywall publisher is acquired by an open source buyer, what happens then? Are the legacy articles protected by the original intent?
October 29, 2014 at 10:05 am
It’s not entirely wrong, I am afraid. It may be true that the publishers cannot control the content of the work, but under these circumstances, they would control the identifiers. If the work is cited, it uses text to say “PloS 2014” and that is how you find it. Or, it might use a DOI. In either case, after acquisitation these would be owned by some one else. And these could be behind a paywall.
Now, of course, you say, that’s not a problem, because it will be in other repositories. But, the reality is that many papers are already in institutional repositories. And very few people find them there.
So, they would be closed, off, and no longer accessible in a very real sense.
As an aside, your reference to CLOCKSS is a red herring, I think. CLOCKSS is a dark archive, and under the circumstances you describe, the publisher would still exist, albeit with a different owner and different policy. The content would remain in CLOCKSS, but would remain dark.
October 29, 2014 at 10:08 am
Thanks, palaeosam, for an interesting and important question.
In general the answer is that restrictions on distribution can never be tightened up but can always be relaxed. This is true because whenever a item is relicensed on new terms, the old terms are still also available. (That is because the CC licences are deliberately written that way, and may not be true of all other “open” licences.) The upshot is that you can do anything that any of the licences allow.
So if, say, PLOS became so huge it could buy Elsevier, then it would, among other assets, acquire the copyrights to all those papers that authors over the years have transferred to Elsevier. In keeping with its mission (“to accelerate progress in science and medicine by leading a transformation in research communication”) we’d expect them then to relicense those papers as CC By.
October 29, 2014 at 10:13 am
Thanks, Phillip, interesting thoughts.
I am not too bothered about the old reference (“PLoS 2014”) — in practice I always search for papers just by title, and I suspect that’s a very common pattern of behaviour. It certainly would become so in the scenario you suggest!
Regarding CLOCKSS: the archive remains dark only until a trigger event occurs. The candidate trigger events are defined by the publisher that registers, so in the case of PeerJ, you’d expect one of those trigger events to be “papers enclosed behind paywall”.
And of course CLOCKSS in this scenario is really only the braces that goes with the belt of PubMed Central, which already has freely available copies of all the PeerJ content.
October 29, 2014 at 6:20 pm
Isn’t this what happened when DeGruyter acquired BEPress?
October 29, 2014 at 7:41 pm
Morrison’s argument that
“There is nothing in the CC BY license that would stop a business from taking all of the works, with attribution, and selling them under a more restrictive license …”
is a truly bizarre misreading of CC BY, which specifies that
-“You are free to […] copy and redistribute […] The licensor cannot revoke these freedoms”
-licensees “may not apply legal terms […] that legally restrict others from doing anything the license permits.”
http://creativecommons.org/licenses/by/4.0/
Thus, neither the licensor nor licensee can place works under a more restrictive license. I don’t see how they could make it any clearer. And this is not hidden in the fine print – I’ve just quoted the lion’s share of the summary.
October 29, 2014 at 8:08 pm
The Morrison quote actually appears to be correct, pedantically speaking, except for the characterization of this as a “major concern”. Someone could buy out peerj and set up a paywall around all of their articles and re-publish them with more restrictive licenses (assuming peerj currently owns copyrights?). I don’t know why anyone would have a “major concern” about this possibility, though.
Since the DOI links etc would point to the now-paywalled site, I suppose someone might be a little less confident that the other copies of your giraffe paper, such as the one at nih.gov, are “official” – I can usually find non-typeset versions of various paywalled papers for free and I never know if they’re the final versions, etc. Big deal.
October 29, 2014 at 11:08 pm
Thanks, all, for comments.
moominoid, I know nothing of DeGruyter and BEPress — I think this must predate my involvement with OA, or even really with awareness of academic publishing at all. Do you have a handy link that would bring me up to speed?
Thomas, I do agree that 90% of the fears people have about the CC By licence would be erased if only they would take the trouble to 108 words of the licence. See also: The CC By licence does not let people distort and misrepresent your work.
ech: it’s hard to mind a more “official” source than nih.gov!
October 30, 2014 at 9:58 am
This is the announcement of the acquisition:
http://www.libraries.wright.edu/noshelfrequired/2011/09/16/de-gruyter-acquires-67-journals-from-bepress-details-from-de-gruyters-sven-fund/
If you visit the journals now, they are behind paywalls, when they were OA before the acquisition.
October 30, 2014 at 10:47 am
That is disturbing. I will investigate.
October 30, 2014 at 4:30 pm
In reply to ech – no-one can legally take CC-BY papers and re-publish them with more restrictive licenses. That would be a clear breach of the license; see my link above. CC licenses are legally binding, have been tested in court in many countries, and to CC’s knowledge have never been found unenforceable in any jurisdiction.
https://wiki.creativecommons.org/FAQ#Are_Creative_Commons_licenses_enforceable_in_a_court_of_law.3F
Peerj and all other real OA publishers never acquire copyright; it remains with the authors, who are therefore the licensors. But more importantly, not even the licensor can impose new restrictions on copying and distribution later. See the same link. So even fauxpen-access publishers like Elsevier, who demand copyright transfer first then publish CC-BY as the licensors, cannot legally impose new restrictions.
On the BEPress journals – they didn’t use CC licenses, but retained exclusive copyright. They were not OA even before the sale:
https://web.archive.org/web/20101226112133/http://www.bepress.com/journals/subscriptions.html
They offered gratis access, but only for academics who registered, under a weird system that they acknowledged was not OA:
https://web.archive.org/web/20101227090123/http://www.bepress.com/journals/guest_access.html
De Gruyter bought them and withdrew that offer. They didn’t re-license the papers; they were “all rights reserved” already, so they could do what they wanted with them. Any author who fell for that bait-and-switch should learn from it and use CC-BY.
Another apparent example of re-enclosure is the “Journal of Visualized Experiments”, which was founded as OA in 2006, then went TA in 2009. Even the formerly free videos are now paywalled on the official site.
But note:
1) Again, they didn’t use CC-BY, but required copyright transfer and reserved all rights.
2) They deposited in PMC, so those videos are still available free. In fact, the free links come up before the official paywalled version on google scholar, e.g.:
http://scholar.google.com/scholar?cluster=1283084329396825520
So neither of these examples reveals a flaw in CC-BY, but rather confirm the importance of it, as well as archiving.
October 31, 2014 at 12:47 am
Thomas, the human-readable summary (“the deed”) of CC-BY 4.0 International that you quote glosses over some fine distinctions found in the legal code ( http://creativecommons.org/licenses/by/4.0/legalcode ). Specifically:
a) What is irrevocable is the license by the Licensor to an individual or entity, not to the world. So the rightsholder is entitled to remove the CC license from the copy they’ve published online, but if others have already exercised their rights under the license in the meantime, that right can’t be revoked. So, as ech notes, Heather Morrison is correct in the possibility that the original copies would be paywalled.
b) As for “apply[ing] legal terms [. . .] that legally restrict others from doing anything the license permits”, the legal code says, “You may not offer or impose any additional or different terms or conditions on, or apply any Effective Technological Measures to, the Licensed Material if doing so restricts exercise of the Licensed Rights by any recipient of the Licensed Material.” So you can’t attach a CC-BY license with the additional provisio that the Licensee, say, follow the Licensor on Twitter. But the Licensed Material is defined elsewhere in the agreement as the “work [. . .] to which the Licensor applied the license.” While “work” is not defined, I think Creative Commons thinks of it as the original work, not derivative works downstream. If it included derivative works downstream, Creative Commons wouldn’t offer a separate ShareAlike license since all of their licenses, which contain this clause, would in effect be “share alike” licenses!
c) You assert that “Peerj and all other real OA publishers never acquire copyright; it remains with the authors, who are therefore the licensors.” First of all, I don’t know of any definition of OA that actually requires that copyright remain with the authors, though all definitions restrict use of copyright just to ensuring the integrity of the work and to credit to the author. So an author might transfer copyright to the journal (thus making it easier for someone 50 years from now to track down the rightsholder to make a use not allowed by the CC license) but have done so only under the condition that the work be published under a liberal license and under the condition that the author retained a large bundle of rights allowing for their own use of the work.
Or even if the author retained copyright, it might be that they didn’t actually attach the CC license but rather gave the publisher the right to publish the work with a CC license attached. This may seem like splitting hairs, but it’s determined by the agreement between the author and the publisher, and it could matter if the license is violated.
October 31, 2014 at 10:53 am
kshawkin, thanks for this contribution.
You are certainly right that the CC By-SA licence provides additional protection against re-enclosure — that is, it prevents enclusure of derived works. (If I licence a photo of a vertebra under CC By-SA, and you make a composite figure of several vertebrae including mine, you must also use CC By-SA for that composite.)
Nevertheless, CC By does indeed provide everything required to prevent re-enclosure of the original licenced work. You say “the rightsholder is entitled to remove the CC license from the copy they’ve published online, but if others have already exercised their rights under the license in the meantime, that right can’t be revoked”. Precisely. And included among those irrevocable rights is the right to distribute futher copies, also under CC By. In short, there is no way to put the genie back into the bottle.
You are right open access does not by definition entail copyright remaining with the original author. But Thomas is equally right that this is in practice how all the major reputable open-access publishers have done things.
October 31, 2014 at 10:55 am
In short, a licensor may stop offering a work under CC By, so that no further copies may be downloaded from their site under that licence. (To do this would be unethical but not illegal.) What they can’t do is go back in time and remove the CC Byness from all the thousands of already downloaded copies — whether downloaded by archiving organisations like PubMed Central or by individuals.
October 31, 2014 at 1:04 pm
Yes, agreed on all counts. We’re just disagreeing over what “re-enclosure” means. I think Heather Morrison and I are focused on the original copy, whereas you are focused on copies made subsequently.
November 1, 2014 at 3:29 pm
[This is long. TL;DR summary: If you don’t transfer copyright, CC-BY is enough to keep your work from being relicensed under more restrictive terms. If you do transfer it, no license will save you.]
In reply to Kevin:
a) I agree that a CC-BY journal could legally be paywalled (but I also agree with Mike that this would in practice not re-enclose them, because they’re available elsewhere under the original license). But that’s not the claim I was objecting to. It was this:
Ech: “Someone could buy out peerj and […] re-publish them with more restrictive licenses (assuming peerj currently owns copyrights?)”
Morrison: “There is nothing in the CC BY license that would stop a business from taking all of the works, with attribution, and selling them under a more restrictive license […] Picture Elsevier buying out Hindawi […]”
This is false. Peerj, Hindawi, PLoS, BMC (i.e. all the major commercial OA publishers) and countless society publishers do not acquire copyright. Copyright remains with the authors, who therefore control the license (they’re the licensors):
“Our licenses and legal tools are intended for use by anyone who holds copyright in the material.”
https://wiki.creativecommons.org/Frequently_Asked_Questions#Who_gives_permission_to_use_material_offered_under_Creative_Commons_licenses.3F
Real OA publishers are licensees, who cannot “legally restrict others from doing anything the license permits.” They can’t impose a more restrictive license, because it’s not their copyright to license.
The converse of this is that if you do transfer copyright to the publishers, then they can later switch to a more restrictive license no matter which one you agreed on to begin with. This applies to Morrison’s preferred SA sharealike licenses just as much as it does to CC-BY. Indeed, ironically enough, the example of re-enclosure that Morrison gives in her thesis was originally full-blown CC-BY-NC-SA, although she doesn’t mention that.
http://summit.sfu.ca/item/12537 (See p. 60)
The “Flat World Knowledge” textbooks Morrison mentions that used to be given away under the ultra-copyleft CC-BY-NC-SA are now paywalled, sold under pretty much the exact opposite license, with hilariously draconian terms promising your firstborn to Sauron:
“This Content is licensed, not sold, to you … You may not distribute the Content … This License is non-transferable. Any purported assignment or transfer of the Content or this License shall be null and void. You are also prohibited from any rental, lease, sale, redistribution and/or sublicensing of the Content.”
http://www.flatworldknowledge.com/student-terms-use
How did they pull off that neat trick, you ask? Simple! Copyright transfer! Further down we read: “the intellectual property rights in and to the Content are the sole and exclusive property of FWK.”. Under each textbook we find “©2014 Flat World Education”.
That’s why, in answer to Kevin’s point c, I say that purportedly open journals demanding copyright transfer, like Elsevier’s, are fauxpen access. You’re handing them a blank cheque. Note that all three of the examples of attempted re-enclosure we’ve discussed demanded copyright transfer. If you don’t transfer copyright, CC-BY is enough to keep your work from being relicensed. If you do transfer it, no license will save you. So the take-home message is, as Mike said, properly done CC-BY cannot be re-enclosed. Don’t transfer copyright, feel free to use CC-BY, and archive elsewhere.
p.s. To finish on a cheerful note, this is also an example of Mike’s point – the old textbooks are still available elsewhere under CC-BY-NC-SA!
https://archive.org/details/flatworldknowledge
November 2, 2014 at 12:03 pm
With the usual caveat that I am not a laywer, this is not legal advice etc etc.
Just to confirm a point that is implicit in much of the above. PLOS, like PeerJ, BMC, Ubiquity and others generally obtain from authors a non-exclusive license to publish. In the case of PLOS we have a non-exclusive right (or if you prefer contractual obligation) to “publish the article under a CC BY license”.
Those sets of rights are an asset that in principle could be transferred to a purchaser. But PLOS at least as far as I understand does not actually have the rights to publish under a more restricted license. Now this raises an interesting question – should our contract with the authors require us to publish without a paywall? I don’t believe that’s explicit at the moment but it could be added to the agreement.
Of course the real point as Mike and others have noted is that the articles are all available at PMC under a CC BY license as well so no re-enclosure is really possible in practical terms.
More generally this does raise some interesting points about best practice with license to publish agreements for OA. I do believe that there would be real value in making these consistent across publishers and getting good practice in place. There’s a bunch of things that PLOS has learnt (like the value of being able to update license versions) and I’m sure there are other experiences out there that could be usefully shared. The idea that the license to publish only applies to copies made free on the internet is one I hadn’t thought of before but could be worth exploring.
[As an aside I’m not necessarily convinced its a smart idea – what if PLOS wants to distribute print copies in sub-saharan Africa at cost, would we want to exclude that? But its definitely worth exploring how we can address these kinds of concerns]
November 2, 2014 at 1:50 pm
I don’t think there would ever be a problem with distributing print copies, at cost or above. Doing so doesn’t in any way restrict people’s access to the content, it merely offers an optional additional route.
November 2, 2014 at 2:09 pm
@Mike – sure, but it illustrates how we should be careful with obligations we put into the system. Its easy to say a throw away “lets just prevent the publisher from charging ever for anything” but there would be downstream effects. And of course the publisher can always use content under the terms of the CC BY license anyway.
Core point – we should think these agreements through and share experiences on how to make them serve community needs.
November 6, 2014 at 8:06 am
[…] PeerJ paper on sauropod neck anatomy, it would in some sense go away if PeerJ folded or were acquired by a hostile entity. But then the practical truth is of course that we’d just make it directly available here on […]
November 16, 2014 at 10:01 pm
[…] the end of October, we published a short piece called CC-By documents cannot be re-enclosed if their publisher is acquired. In an interesting discussion in the comments, moominoid […]
March 27, 2015 at 10:46 am
[…] Since CC-By documents cannot be re-enclosed if their publisher is acquired, even if PeerJ were acquired by a predatory barrier-based publisher such as Elsevier, the articles […]