Monday, February 20, 2017

Copyright: the immoveable barrier that open access advocates underestimated

In calling for research papers to be made freely available open access advocates promised that doing so would lead to a simpler, less costly, more democratic, and more effective scholarly communication system. 

To achieve their objectives they proposed two different ways of providing open access: green OA (self-archiving) and gold OA (open access publishing).

However, while the OA movement has succeeded in persuading research institutions and funders of the merits of open access, it has failed to win the hearts and minds of most researchers. 

More importantly, it is not achieving its objectives. There are various reasons for this, but above all it is because OA advocates underestimated the extent to which copyright would subvert their cause. That is the argument I make in the text I link to below, and I include a personal case study that demonstrates the kind of problems copyright poses for open access.

I also argue that in underestimating the extent to which copyright would be a barrier to their objectives, OA advocates have enabled legacy publishers to appropriate the movement for their own benefit, rather than for the benefit of the research community, and to pervert both the practice and the concept of open access.

As usual, it is a long document and I have published it in a pdf file that can be accessed here

I have inserted a link to the case study at the top for those who might wish only to read that.

For those who prefer paper, a print version is available here.


Peter Suber said...

Hi Richard: In the text accompanying notes 83-86, you refer to "the legal uncertainty some see in the Harvard [OA] policies." In support of this you cite only one brief and uninformed letter by a non-lawyer, and overlook the careful legal scholarship supporting the soundness of the Harvard approach (Frankel and Nestor, and Priest ). You also overlook a few relevant facts, for example, that the faculty of Harvard Law School adopted its OA policy by a unanimous vote, and that Harvard has never been sued or threatened with a suit for carrying out its OA policies. Today we have more than 34,000 works in our repository, which have been downloaded more than 10 million times, with no legal claims from publishers.

In note 84 you suggest that my 2012 debate with Mark Seeley (Elsevier VP and General Counsel) supports your thesis about the legal uncertainty of the Harvard approach. But this is incorrect. First, the debate was about the NIH approach, not the Harvard approach, although they are similar. Second, Seeley conceded in the debate that the NIH approach did *not* violate copyright law. (The same concession would apply to the Harvard approach.) See my blog post at the time for pointers to the places in the video where this concession appears .

You seem to think that it wasn't enough for Harvard faculty to make a voluntary grant of non-exclusive rights to the university. You seem to think that the university should "appropriate faculty copyright all together" in a "top-down diktat." We can argue about the wisdom of that later. (I'm not the only academic to find the idea appalling.) Your only argument in context is that this total appropriation of faculty rights is the only way to stop publishers from taking the same rights. But that is simply untrue. When faculty don't get waivers, publishers cannot take the non-exclusive rights possessed by Harvard and Harvard faculty. Which leads to the next point:

Finally, you suggest that because publishers can in principle demand that Harvard authors get waivers, either they do so in large numbers or "the objective of the Harvard OA policies is in any case subverted." This too is untrue. Fewer than half a dozen publishers systematically demand waivers from Harvard authors, and one of them will soon stop. Our waiver rate is below 5%. We don't think that subverts the objective of the policies. On the contrary, we're delighted to have permission to distribute 95% of the new scholarly articles by Harvard faculty.

Finally (and less importantly) I hope you can fix footnote 27. The article of mine that you mention was OA at the time it was published, and is still OA in the Harvard repository. I wish you'd asked me for a link to an OA edition before depicting me as a hypocrite.

Unknown said...

With regard to your case study: the right of fair quotation not only applies to small fragments of text but also to figures or images. Publishers claim too much when they demand you have to ask permission for quoting figures. See for example some of the comments on this blog post: This blog post is also extensively commented by Kevin Smith here:

Richard Poynder said...


Hi Peter, thanks for your input. I appreciate you taking the time to respond.

Let me preface my reply by saying I think you may have misunderstood some of what I said or (perhaps more likely) I did not express it well.

In my comments about the Harvard Policy, for instance, I was referring not just to the policies introduced at Harvard University, but to the Harvard-style policies introduced at the other 60 odd research institutions that have cloned the Policy. So, while I am encouraged by what you say about the low waiver and high download rates at Harvard, my interest (and concerns) relate to the other universities that have introduced Harvard policies too (and of course all the other universities around the world).

Can I also say that I am not (as I think you imply) advocating for universities to assert ownership of faculty copyright, but pointing to the danger (as I see it) that this could occur if researchers don’t become copyright literate, and use that knowledge to prevent publishers appropriating their works. I too find the possibility appalling; my fear is that it may be where we are headed.

Specifically on your four points:

1. When I referred to legal uncertainty over the Harvard policies I was not referring just to the question of whether these policies violate copyright law, but to the uncertainty and concern they generate within the research community. Clearly not everyone will be as confident about their legality, and how they work legally, as the copyright lawyers you cite. So, when you say that I cite “only one brief and uninformed letter by a non-lawyer” you perhaps overlook the point that this non-lawyer (a mathematics professor at the University of Illinois at Urbana-Champaign) is one of the growing number of researchers now subject to a Harvard-style policy. It is apparent from reading his letter that he for one is uncertain about the legal situation. Moreover, he complains that he has been unable to obtain advice on the legal implications of the policy from his colleagues, his librarians or from publishers. In short, the policy creates legal uncertainty.

You are right that I did not mention that Harvard Law School adopted its OA policy by a unanimous vote, but I was not unaware of it. As I say, I am also interested in how the policy has been implemented and received in the 60 other institutions that have introduced such policies, many of which I believe have introduced variations in the wording in doing so (stoking further confusion perhaps?). Moreover, I argued that since (I assume) these votes are a one-time event whose result then persists for ever, it would seem that any new staff arriving after the vote will be presented with a fait accompli. So, talking about a unanimous vote may not fully express the degree of buy-in from faculty over time. And as the OA landscape develops and changes it is not unlikely that any unanimity existing at the time of a vote may later fade.

I am pleased that Harvard has never been sued, and has received no legal claims from publishers. But is this proof that were a lawsuit to emerge it would inevitably fail? In these situations, only a court can provide the necessary certainty as to legality, and I suspect it would take a brave publisher to sue Harvard. The consequent bad press from doing so would likely also be a real disincentive. Be that as it may, I make no claim as go the legality of the Harvard Policy (and how could a non-lawyer?) Again, my point was that these policies do seem to introduce legal uncertainty, not least for those subject to them, so in order to make the situation clearer, and avoid requests for waivers, universities might start to assert ownership of faculty copyright – particularly given the current trend for weakening and removing rights that academics have enjoyed historically. And that is just universities with Harvard policies. There are many others universities who might be looking for a way of guaranteeing open access.

Richard Poynder said...


2. Concerning the debate with Mark Seeley: Yes, it was a debate about the NIH Policy, but as you say the NIH Policy is similar to the Harvard Policy, and I would expect the issues to be similar. Again, this is not just a question of whether these policies violate copyright law, but the extent to which they create legal uncertainty. That it was felt necessary to hold a debate demonstrates that there is uncertainty. And the degree to which the debate settled matters is not clear to me. I am conscious, for instance, that in the comments below the Google+ post you link to above Mark Seeley ends the discussion by saying, “That isn't the end of the debate”, although it is not entirely clear what he meant by that.

3. To repeat, I am not suggesting that universities ought to appropriate faculty copyright. Indeed, I believe it would be a very bad thing were that to happen. You say that publishers would not be able to acquire exclusive rights if faculty don’t get waivers. That might seem to be all the more reason why publishers would press authors to opt-out. You also point out that that this is not a static situation, and as we have learned, the more threatened publishers feel by open access, the more they seek to co-opt and control it. And as I say, when I refer to the Harvard Policy I am talking both about the policies at Harvard as well as those introduced at the other 60 odd institutions.

But can I put a couple of questions to you: if the Harvard Policy grants the institution an upstream non-exclusive licence why does Harvard go along with publisher-specified embargoes? (Or does it perhaps not abide by publisher embargoes?). I also believe DASH has not implemented the “copy-request button”. I assume this is because there is some uncertainty about its legality, but if Harvard has an upstream non-exclusive licence to the works deposited in DASH, why does that matter? It might seem to suggest that—to all intents and purposes – publishers do have exclusive rights in the works, since it would imply that they can set the rules as if they did.

4. On the issue of one of your texts being behind the ScienceDirect paywall. I was certainly not trying to depict you as a hypocrite, and I apologise if it seems as if I was. My point was that even when a paper is available as open access in a repository (or elsewhere) it can still also be behind a paywall elsewhere. I used the example of your paper to make the point that even someone 100% committed to open access can end up with their work sitting behind a paywall. What is significant about this is that (as with the Apple Academic Press controversy) potential readers may be unaware that an OA version of a work is freely available elsewhere. So they may feel they have no option but to pay $35.95 to read a work. After all, it is not as if publishers like Elsevier indicate if a free copy exists, or link to it if there is one. And should that free version disappear for any reason, the only copy available may be stuck behind a paywall.

Please note I am happy to link to your comments from the relevant places in the pdf should you wish me to.

David Groenewegen said...

Great piece and not just because I am quoted in it. In your conclusion you ponder why researchers keep giving away their copyright. In my view, it's because the benefits they accrue by doing so outweigh the benefits of not doing so. This has alway been the weakness of the OA movement, which sees scholarly publishing as a communication process, when a huge part of traditional publishing's continued power lies in its role as providing an assessment and evaluation mechanism. Researchers (and their employers) compare themselves to others based on where they publish and how often the work is cited. The ARWU university rankings give points based on how many Science and Nature articles the institution has published! This is what brand name journals, be they Elsevier or BMJ, offer - access to the symbols of merit that the research community (and its funders) values. If the price of getting that is giving away a right you will probably never need or use, why wouldn't you? What's the incentive to resist?

Just for clarity's sake: I'm the Director, Research at Monash University Library, not the university as a whole.

Stevan Harnad said...

FOA: Free Online Access

Let’s simplify:

1. Stick to peer-reviewed research articles: that’s all FOA is or was ever about.

2. Copyright and re-use rights are and always have been a red herring in the FOA age.

3. All that’s needed is an FOA version of the peer-reviewed research article.

4. That’s the author’s peer-reviewed final draft.

5. The only thing needed from “journals” (or equivalent) is the adjudication and certification of the peer review.

6. That’s a service, not a product: Nothing to “copyright.”

7. To make the current house of copywrit (sic) cards collapse, all authors need do is make 4 (the author’s peer-reviewed final draft FOA (freely accessible online)

8. FOA immediately (upon “acceptance”) and permanently (“FIPATRAFTO”).

9. For the faint-hearted and superstitious, there’s the Copy-Request Button during any (bogus) publisher “embargo” on “OA.”

10. Like Copyright worries (2), Button worries are red herrings.

11. FOA is all that’s needed or was ever needed.

12. Once researchers, their institutions and their funders get round to providing FOA (= “Green, Gratis OA”), Fair Gold “OA” (peer-review service fees) and all the re-use rights researchers need will be within trivial reach.

(Waiting for researchers, their institutions and their funders to get their heads around this and to set their fingers in motion continues to be yawningly boring; please wake me when they get round to it...)

Charles Oppenheim said...

Stevan is not quite correct. Copyright IS relevant because the authors foolishly agree to assign (or give an exclusive licence) it to the publisher. For Stevan's proposal to work, they must stop doing that. If a researcher has foolishly given away that right, Stevan's No 7 onwards in principle cannot work. So amend Stevan's 12 point plan by adding a final sentence to his point 6, "UNDER NO CIRCUMSTANCES SHOULD THE AUTHOR ASSIGN COPYRIGHT TO THE PUBLISHER, OR GRANT IT AN EXCLUSIVE LICENCE TO THE WORK". Then his plan is legally watertight. I do agree with Stevan's final thoughts in brackets. I too am waiting.....

Rick Anderson (editor) said...

In response to Barry: the definition of Open Access has always been, and continues to be, to some degree contested. I discussed this a couple of months ago in the Scholarly Kitchen, and there was some very good further discussion of the issue in the comments. You can see that posting here, if you'd like:

Rick Anderson

Richard Poynder said...

Interestingly, it started out as Free Online Scholarship, or FOS, rather than FOA, or OA. See here

Unknown said...

What puzzles me about the case study is that there was a very simple solution: just insert a link to the source at one or another of its several OA locations. No permission is required to make a link, and no copyright issues are involved.

OA zealots have treated BOAI's definition as though it created the concept of OA, but that simply is not true. The idea was floating around for decades before the BOAI existed. It was the key to a proposal made by what is now called the Big Ten Academic Alliance about a collaborative online monograph project in 1996, and the idea was being discussed in the Big Ten for at least a decade before then. What BOAI did do is stake a claim that only CC BY is true open access. But BOAI represents the interests of, at best, the STEM community, not the scholarly community overall, and certainly not the humanities. As I have argued repeatedly, there is very good reason for humanists to be concerned about, and have control over, the quality of translations and also over where any republication of their writings appears (what Richard refers to as "reputational risk"). How OA applies to monographs is also very different from how it applies to journal articles.

On the question of Harvard's policy, at least in the Arts and Sciences Faculty it was adopted "unanimously" only to the extent that all faculty who showed up for a meeting voted for it, which was a small subset of the overall faculty. I'm not sure whether that was not the case also for the Harvard Law faculty.

One reason that copyright has been transferred is that traditionally in scholarly publishing publishers have served the role that literary agents serve for trade authors, with respect, for instance, to knowing how to negotiate subsidiary rights. Faculty have no clue about how to negotiate a translation license, for instance, much less an option contract for film rights, which can run to 20+ pages of legalese. Publishers know the territory and can negotiate these deals on authors' behalf, sharing the income, usually 50/50. This may happen rarely for STEM journal articles, but certainly can happen for monographs in the humanities and social sciences--another reason that monographs differ greatly from articles. Publishers also need exclusive rights; they cannot go to court to defend against infringement if they only have nonexclusive rights.

Stevan Harnad said...

Charles’s formal caveats are quite correct — for pedants and poltroons. But think: if the physicists in 1990 (or the computer scientists even earlier) had had the slightest inclination to give such reservations a single moment’s credence — instead of going ahead and doing exactly what they did (self-arxiving their preprints, and then their refereed final drafts, immediately, and not bothering with the publisher’s “version of record”), FOA would never have made even the modest strides it has made:

So let’s stop fussing about copyright and start self-archiving all our refereed final drafts immediately upon acceptance for publication! It’s 2017, almost 3 decades since good sense first erupted and prevailed in some intrepid subsets of the research community. Coraggio!

Apart from that (which should be done in any case), Charles is also right that it is foolish (and unnecessary) to assign copyright or grant an exclusive license to the publisher. The limited right to publish and sell the version of record, online and on paper, is more than right enough — and as much reward as they deserve for their one and only “value-added,” which is the administration of the peer review (which the peers provide for free, just as the author provides the text for free).

Heather Morrison said...

This is an important and timely piece, Richard, thanks. It is helpful to have examples of existing issues with CC licenses and OA gathered in one place. There are some specific points that I see differently. I think copyright as an immovable barrier to OA is an overstatement,but I see the equation of OA with the CC-BY license as a significant barrier to progress. Here I see hopeful signs: I asked the open science list (well populated with CC-BY purists) if we now have consensus that CC licenses on their own are not sufficient for OA, and I gather that, in large part due to Ross Mounce's work exposing Elsevier pay walls on paid for OA articles, we do now have consensus. This is important because it means that now we can enter a new phase of discussion to ensure ongoing OA. The Harvard style policy is not universities grabbing copyright, it is nonexclusive rights. I consider the MIT version to be an improvement as it specifies open access in comparison to Harvard's "not sold for a profit" (which opens a loophole for future cost recovery). Specifying free of charge would be even better. I think institutional repositories are the way to go (in addition to other methods). The concept completely fits with other major tasks for academics. We have to report what we publish for evaluation and to promote our work and our university on our websites. When evaluating other people's work or refereeing articles, we need to make connections between bibliographies and works. Universities are very tradition-bound and slow to change but the superior functionality of clickable links that work for everyone will win in the end.p

Heather Morrison said...

Fair use / fair dealing (copy of message sent to GOAL list for discussion here or there):
Thanks Richard - this brings together many of the major problems, issues and cases on this topic. I encourage everyone to read it, comment on it and begin building from this understanding of the complexity of copyright.

As Fair Use / Fair Dealing week is wrapping up, I would like to take this opportunity to argue that open access scholarship requires strong fair use / fair dealing, not only for open access works but for all types of works. As a researcher, I make my own work open access but I also need to draw from works with a variety of copyright statuses including All Rights Reserved, and sometimes scholars and journalists need to work with material that the creators did not intend to publish at all. When we work with these kinds of materials, we need to be able to point to the originals, sometimes in whole and sometimes by copying portions. Even if ubiquitous open licensing were possible for scholarly works were possible, this would not address non-scholarly materials with which we conduct research. For this, we need fair use / fair dealing.

When I re-use portions of works that are clearly not intended for broad-based blanket downstream re-use rights, I cannot release these portions of my work under open licenses. If I am subject to overly strong open licensing requirements, this is a disincentive to conduct many types of research.

Fair use / fair dealing cannot be taken for granted. These rights are recognized by some countries (fair use in the US, fair dealing in the UK and Canada), but are not as widespread as copyright. These rights are often contested in courts and a subject of discussion in developing laws and treaties touching on copyright.

I argue that fair use / fair dealing is essential for open scholarship and that working cooperatively with advocacy efforts for fair use / fair dealing is in the best interests of open scholarship. In Canada we currently have a good copyright law with respect to user's rights. This is up for review later this year. It is wise to assume that those who wish to profit from IP are already lobbying to weaken or eliminate user rights, both through our national copyright law and in international treaty discussions. Advocacy in the interests of readers and creators who need content to build on will be necessary.

Individual readers are largely invisible when it comes to policy discussions on copyright. For this reason, librarians and library associations (such as IFLA) have been the major champion of user's rights, speaking out for all of the readers served by the profession.

The ARL fair use / fair dealing website is available here:

Some libraries may be planning events at a later date. For example, at the University of Ottawa it is currently study week, so our library is having workshops next week.

Heather Morrison said...

Correction re Harvard / MIT policy: on review thanks to a comment from Peter Suber, I would like to note that both policies contain the same text and both are flawed in that neither specifies that works covered under the policy will be free of charge to the reader.

Heather Morrison said...

For Barry (and Richard): one item that I do not agree with from Richard's case study is the idea of asking publishers for clarification of rights with respect to preprints and to items included within preprints. Merely asking the question gives the publishers even more power. Much of copyright law (under common law) is determined through case law. If people get in the habit of always looking to publishers for rights, this can become evidence to support such rights in a court of law. This is one of the reasons why I strongly support the Harvard model (even though I wish it would be fixed to clarify that works will be free-of-charge) as this establishes that no publisher has exclusive rights to these works. The Harvard-style model can be combined with the top-down institutional requirement approach. Both have benefits and are not mutually exclusive.

Heather Morrison said...

Another reason not to ask permission from publishers is that it forces the pro-enforcement publisher - not the librarian or the OA advocate - to take on the role of enforcer themselves. Every take-down notice sent to an actual author is a wake-up call. I wonder how many discussions that led to the development of new OA journals, or conversion of existing journals to OA, resulted in part from this kind of enforcement on the part of a publisher?

For this reason, I advocate that repositories stay out of the role of copyright enforcer. If takedown notices are sent to the repository, forward them to the authors and ask them what to do. Take-down notices might be sent in error; publishers likely do not keep careful track of every single instance of modification of their default license. The repositories' license from the author needs to address this, i.e. ask authors to confirm that they are responsible for copyright.

It could still occur that a repository will have to take down works because of a threat of a lawsuit. If this happens, go ahead and take things down - but let people know. Tell authors that the publisher they chose threatened to sue their employer over sharing their work. Tell the world, in the IR, "this article has been removed due to legal threat from publisher X" - post the threatening notice, and point to this.

Not every repository will be able to do this. I am posting this by way of argument against assisting publishers with IP enforcement, as one reasonable approach, not "the" approach that everyone must follow.

Heather Morrison said...

Ethics and research subjects

Richard - to add to your analysis: if authors and journals are using materials (works, images) that are from research subjects, there are ethical issues involved with open licensing. Getting consent to publish something is not the same as getting consent to publish under an open license.

The Virgin Mobile case is a good one to cite, even if this not an academic work. In brief, a photographer took a picture of a young girl and uploaded it to flickr under a CC-BY license. Virgin Mobile's interpretation of the license was that it was an invitation to use the girl's photo in an advertising campaign without having to ask permission, which they did and did not cease even when notified by the girl and her parents that this was not acceptable. The case was settled out of court so we may never know the legalities, however from an ethical perspective informed consent to use material like this under an open license would involve something like telling people about this case as one potential consequence of open licensing before requesting permission to publish with an open license.

Lessig's post on this can be found here:

Elizabeth Gadd said...

This is a great exploration of the issues surrounding copyright and open access. However, implying that Librarians have a "knowledge deficit" in this area is unfair. This suggests that there is knowledge out there, but that Librarians have failed to acquire it. In fact there is no agreed "knowledge" on these matters - a point your piece makes very well. Indeed, even lawyers are often unable to provide definitive answers on copyright or licensing questions, because unless a case goes to court, we can only guess which way a court would decide. And even when a case DOES go to court, we are only left with a decision pertinent to the specific facts of a particular case and no certainty that in a related but different case, the decision will go the same way.

My experience of Librarians, particularly Institutional Repository Librarians, is that they ARE the go-to experts on copyright and open access - and copyright expertise is by no means a new role for libraries seeking to find work for themselves. The problem is not the lack of expertise, it is the confusing nature of the landscape - which is exactly the point the twelve year study of publisher OA policies sought to evidence. I would say as one of the "Librarian" (actually ex-Librarian) authors of this paper (appearing in your case study), I felt pretty confident about the nature of our agreement with Sage. I deliberately chose Sage over other publishers for the wording of their agreement ("you may do whatever you wish with version 1 of your paper") and it was on this basis (not because we retained copyright, which as you rightly point out is a nonsense) that we offered the right to re-use the images in your blog piece. The problem was not a knowledge deficit, I "knew" and still believe this was legitimate, the problem was that the publisher disagreed with my knowledge - albeit not so vehemently that they were prepared to take us to court. And unless it goes to court, we will never know who is "right".

You make a good point about a publisher's OA policy being a different beast to its licensing agreement and, again, the lack of clarity over the relationship between the two. In our case the licence agreement had no 'upstream' clause and to my mind related only to the author final version (version 2 in Sage terminology), which would chime with their OA policy which states that we may do whatever we wish with version 1 (the pre-print). However, confusion inevitably arises where the pre-print is virtually identical to both the author accepted manuscript and the copy-edited version of that manuscript. This was the case with our paper, hence initially submitting the copy-edited version to our IR, and then later replacing it with the author-accepted manuscript to strictly follow the HEFCE REF definition of the AAM. Either, in fact, would probably have done.

But the bigger issue is why I, as the originator and joint owner of the copyright, should be playing to the tunes of EITHER funder or publisher, and why, as you point out, it is critical that more universities seek to assert their stake in the ownership of scholarly works. This is a point I make in some subsequent research which sought to understand whether and how UK universities were responding to open access through their copyright policies. Answer? 28% UK universities now seek to either 'share' copyright with academic authors through licensing or expect to re-use their academics' scholarly works in some way. The challenge facing this activity is the same as that facing Harvard-style mandates, in terms of the reliance on individual academics to comply. However, to my mind it is the closest we've come to a significant move forward in this space, and I look forward to the roll-out of the UK version of this licence, the UK Scholarly Communication Licence, later this year.

You can find a link to the full study on UK university copyright policies here:

Richard Poynder said...

Thank you for responding Elizabeth. You say:

In our case the licence agreement had no 'upstream' clause and to my mind related only to the author final version.

As I understand it, your argument is that each version of a research paper has a distinct set of rights, and so when authors sign a publishing agreement they assign copyright and/or exclusive rights only in the final version of the paper. As I indicated in my document, this is a contested area. Personally, I doubt your view would prevail in a court of law.

Either way, let me offer two opposing views. First this from a US perspective; second this from a UK perspective.

Danny Kingsley said...

Dear Richard,

First thank you for pulling together some of the many complex tendrils that surround copyright and open access into this discussion piece. I have listed below some comments, questions and possible clarifications.

You mention, on page 9, a scholarly meeting in Cambridge last year where issues about reuse of work were discussed. I am not sure if it was the roundtable held on CC-BY but if it was, the write up from that discussion is available. "Is CCBY really a problem or are we boxing shadows?"

On page 13 there is mention of the issues with the Copyright Licensing Agency and Copyright Clearance Centre where the retention of the copyright determines where money collected through photocopying income should be directed. I find this confusing. If an article is being made available under a CC-BY license then there are no copyright dues to be paid, surely? This money should not be collected by the agencies at all. Or do I have completely the wrong end of the stick?

In your discussion about the confusion about versions (with which I wholeheartedly agree), you possibly inadvertently introduce a level of confusion by referring to the 'pre-print version of the paper' when you are indeed talking about the Author's Accepted Manuscript. In the pre-print/post-print world, the AAM is the equivalent of the post-print - which of course is actually yet to be 'printed' hence the need to move away from these confusing terms!

On page 21 there is a short discussion of 'copy requests' which links out to the blog posts we have published discussing the service at Cambridge. The attached footnote (53) then mentions librarians doing this on behalf of researchers. Can I make it categorically clear that at Cambridge we do not facilitate these requests on behalf of the authors. We will spend some time tracking the author down if they have left the institution to send on the request. If we are unable to locate the researcher we reply to the requestor to say the request is unable to be fulfilled.

Buried right at the end of the whole piece is a footnote that I feel is one of the most important points in this issue. Footnote 90 discusses how commentators have failed to notice how scholarly publishers are changing their focus. This is vitally important. Sally Rumsey, Head of Scholarly Communications and RDM at the Bodleian Libraries and I gave two talks last year on this topic: "Watch out its behind you: Publishers' tactics and the challenge they pose for librarians" This presentation is already out of date, as activity in this space has accelerated exponentially.

Finally just a call out on a couple of sloppy language issues. On page 10 you assert "As we have seen, most researchers prefer more restrictive licences". While this is true within the small set of examples provided in the piece, there is no comprehensive overview or analysis of researcher opinion, so this claim is something of an overstatement. Also the statement on page 33 that funders have 'slavishly followed OA advocates down a primrose path' is very much editorialising (and slightly insulting to funders). This kind of language weakens what is otherwise a substantial and important piece.


Dr Danny Kingsley
Head, Office of Scholarly Communication
Cambridge University Library

Richard Poynder said...


Thank you for your detailed and thoughtful comments Danny. Let me respond to the seven points you raise:

1. The scholarly meeting I attended in Cambridge was OpenCon, which was held in November, but thank you for the link to the March roundtable event.

2. On the question of fees earned from the Copyright Licensing Agency and Copyright Clearance Center: I agree that there should be no fees arising from papers licensed under CC BY. However, as you know there are also non-commercial CC licences. In any case, we might want to wonder why CUP would include this clause in its Gold OA copyright licence form if it did not expect such revenues to arise. There is also the possibility, of course, that money is sometimes mistakenly paid for the copying of CC BY papers. The larger point I was making is that legacy publishers will seek to monetise research papers however they can, even in an OA environment.

3. I think you are right that I could have been more precise in my use of the terms preprint and AAM/postprint, and I could have made the distinction between them clearer. However, I am not sure this is a confusion I have introduced into the discussion. As the SHERPA/RoMEO FAQ explains, “Publishers may use the term pre-print to define all forms of the article prior to print publication. SHERPA follows an academic practise of defining pre-prints as a draft of an academic article or other publication before it has been submitted for peer-review or other quality assurance procedure as part of the publication process. Preprints cover initial and successive drafts of articles, working papers or draft conference papers. Post-print is then the form of the article as it is submitted for printing.”

What I think we both agree on is that there is far too much confusion about versions, and some of the definitions really need to be retired in the online world.

4. Thank you for clarifying the situation concerning copy requests at Cambridge. I think the larger point here is whether the copy request button is legally doubtful. Some universities apparently have concerns about this, and so have not implemented the button. I understand, for instance, that Harvard’s repository DASH does not have the button installed, and I suspect that that is because there are concerns about its legal status. In my response to Peter Suber above I was hoping he might clarify why there is no button installed on DASH.

In 2015 I asked HEFCE about the legality of the button. HEFCE replied, “We have not sought legal advice on the status of the ‘copy request’ button … It is for institutions, and for those that operate subject repositories, to decide whether to implement the ‘copy request’ button in their systems.” Presumably, Cambridge University believes that use of the button raises no legal questions.

Richard Poynder said...


5. Thank you for highlighting Footnote 90. I too believe publishers’ moves to reinvent themselves to be an important issue, but one I felt to be outside the topic I was covering. As I also noted, however, it seems to me that publishers’ success with things like workflow systems will be greatly enhanced if they are able to continue controlling the underlying content.

6. On my assertion that most researchers prefer more restrictive licences: I am not an academic and so not entirely clear as to when a piece of research can be considered significant. As I understand it, the Taylor & Francis survey attracted nearly 15,000 respondents, around 13,000 of whom answered the questions on licensing, and T&F described it as “the largest single survey conducted by our Research & Businesses Intelligence Department to date”. In answering the survey the majority of respondents said they prefer more restrictive licences. However, it is true I cited only two surveys, if only because (which I think is your point) there has not been a great deal of work done in this area. This might seem to support my point that it may not be wise for OA advocates and funders to be seeking to mandate universal CC BY when there is a dearth of knowledge about the possible implications and unintended consequences of doing so, let alone sufficient knowledge about researchers’ receptiveness to it.

Either way, I would expect a more comprehensive survey to support the findings of the two surveys I cited, and confirm that most researchers prefer more restrictive licences to CC BY. I say this not just because (as I noted in the document) I believe most researchers have a strong sense of ownership of their work, and so feel they ought to have control over whether, how, when and where it is reused, but because in demanding that they use CC BY OA advocates, funders and research institutions are asking them to do something that conflicts with all the other messages and incentives they are subjected to. As I understand it, researchers are being told that they should embrace CC BY because it is a co-operative and non-selfish thing to do, and that it will benefit their colleagues, their subject, and society at large, at their expense. At the same time, they are being told that if they want to obtain the funding, jobs, tenure etc. needed in order to survive in academia today they must embrace the competitive, selfish and aggressive behaviour characteristic of what Sheila Slaughter and Gary Rhoades have dubbed “Academic Capitalism”. The latter is the dominant culture of academia today. So asking researchers to simultaneously embrace the sharing, caring ethos of CC BY simply does not compute.

7. Finally, you say that I weakened my argument by editorialising, notably by suggesting that funders have slavishly followed OA advocates down a primrose path. Is that the case? I don’t know. Certainly, it is possible that my view could be UK-centric, for I do believe it to be an accurate description of the way in which the RCUK and HEFCE OA policies were introduced.

Thank you again for your response.

Richard Poynder said...

Sorry about the numbering of those last two comments. Clearly, they should have been headed 1/2 and 2/2!

Unknown said...

Great piece of work & interesting comments. My question still is, how to deal with this? Educate researchers? How? When the problem is not perceived as a fundamental issue by them and when a 'top-down diktat' is not the most attractive solution?

Richard Poynder said...

Thanks for the comment Hannelore. You are of course right to point out that the question remains: what is to be done?

My view is that this is ultimately a question for researchers, and I was suggesting that if they don’t wise up they can expect to see a further erosion of their independence and freedom, as funders and research institutions seek to acquire copyright in their works.

And I suspect many would argue that if they did this funders and research institutions would only be acting responsibly, since it would be a practical way of preventing publishers from continuing to acquire exclusive rights in scholarly works and then holding the research community (and ultimately the taxpayer) to ransom.