Nature | News
Researchers opt to limit uses of open-access publications
Advocates of open publishing fret that misunderstandings lead scientists to choose restrictive licenses.
Unlocking access to research papers may free them for widespread re-use — or it may not, depending on the publishing license.
Academics are — slowly — adopting the view that publicly funded research should be made freely available. But data released yesterday suggest that, given the choice, even researchers who publish in open-access journals want to place restrictions on how their papers can be re-used — for example, sold by others for commercial profit.
That stance is directly opposed to the views of major funding agencies, such as the seven UK research councils and the Wellcome Trust in London, one of the world's wealthiest biomedical charities. Advocates of open access say this shows that researchers don’t understand how publishing licences affect ‘open’ research papers, and that more work needs to be done to explain why licences matter. But some publishers argue that restrictions are needed.
Open, but not open?
A paper that is free to read on the Internet is not necessarily legally open to other uses — such as ‘mining’ the text with computer software to draw conclusions and mix it with other work, distributing translations of the text or commercially selling republished versions in derivative publications.
In many parts of the world, anyone wishing to re-use papers must get permission from the copyright owner (usually either the publisher or the author). Often, the owner will forbid re-use or demand payment. Supporters of open access argue that free papers should come with licences attached, making it clear what kinds of re-use are allowed.
From 1 April, the Wellcome Trust and the UK research councils will mandate that when they pay the fees for research that they fund to be published under open-access conditions, the work should be available under the most liberal of licences, allowing anyone, even commercial organizations, to re-use it. In the parlance of Creative Commons, a non-profit organization based in Mountain View, California, this is the CC-BY licence (where BY indicates that credit must be given to the author of the original work).
“There has been very little opposition to our plans” from researchers, says Chris Bird, a lawyer at the Wellcome Trust. But he says that is partly because most researchers don’t understand the issue.
Tough choice
One piece of evidence on researchers’ opinions comes from the open-access journal Scientific Reports, which since July 2012 has been offering researchers a choice of three types of licence. One is CC-BY. A more restrictive version, CC-BY-NC-SA, lets others remix, tweak and build on work if they give credit to the original author, but only for non-commercial (NC) purposes, and only if they license what they produce under the same terms (SA, or 'share-alike’). A third licence, CC-BY-NC-ND, is the most restrictive, allowing others to download and share work, but not to change it in any way (ND, ‘no derivative works’), or use it commercially.
The journal’s publisher, Nature Publishing Group (which also publishes Nature) yesterday revealed that of 685 papers accepted between July 2012 and mid-January 2013, authors chose either of the more restrictive licences 95% of the time — and the most restrictive, CC-BY-NC-ND, 68% of the time. In November 2012, the publishers wondered whether the ND licence might be most popular because it was the middle option listed on the copyright form. They changed the order, putting the ND licence first — only to find that even more researchers began to choose this option.
“All it indicates to me is that researchers who publish in Scientific Reports tend to choose conservatively, and perhaps don't fully understand their choice,” says Ross Mounce, a palaeontologist at the University of Bath, UK, who also works for the Open Knowledge Foundation in Cambridge, UK. Mounce, who has published a list of journals classified by the open-access licences that they offer, thinks that providing a choice is a bad idea. “It allows authors to make the mistake of choosing a less open licence,” he says. “Are there really any common circumstances in which they might want a less open, free-to-read licence?”
But such circumstances do exist: some researchers, particularly in the social sciences and humanities, don’t want commercial organizations to be able to re-use their work. In December 2012, for example, the editors of 21 UK history journals released a statement arguing that allowing commercial re-use would be a “serious infringement of intellectual property rights”.
A different path
Many publishers are also arguing against CC-BY, concerned in part about the loss of income if others can resell open-access works. Indeed, the International Association of Scientific, Technical and Medical Publishers, a global trade organization based in Oxford, UK, is working on an alternative open-access licence that does not allow commercial or derivative use in reprints, abstracts or adaptations, but explicitly allows text-mining and translations.
The problem is that adding restrictions to the re-use of work — even with good intentions — can create complex legal issues, explains Martin Hall, vice-chancellor of the University of Salford in Manchester, UK, and a co-author of the ‘Finch report', an influential study on open access commissioned by the UK government.
For example, a social scientist who used an NC licence to prevent their work being used in textbooks without the payment of royalties might also be keeping it from appearing on any website that carries advertising content. The field is plagued with misunderstandings, Hall adds — one myth being that putting out a paper with a liberal licence puts featured third-party content such as photos or music scores at risk of unauthorised use.
Ultimately, says Hall, “I don’t think we’ve done enough work spelling out the implications of open-access licences to enable the researchers to have a view.” After all, he adds, until the past few years, researchers have been accustomed to signing their rights over to publishers without thinking about what this means.
- Journal name:
- Nature
- DOI:
- doi:10.1038/nature.2013.12384
Comments
Report this comment | #54877
Out of curiosity, how could publishers lose income on their CC-BY works? ie, what income? Author fees wouldn't be affected, and If it's something like ad fees going down with lost traffic – that doesn't seem right either. They'd seem to have a competitive advantage in attracting readers to their free product, over a for-profit re-seller
Report this comment | #54879
FYI I also sent Richard van Noorden data to show that the vast majority of good open access journals (those that are Thompson Reuters JCR listed) use the CC BY licence.
In fact, all these and more use CC BY because it's a good, tried and tested licence for academic research:
American Physical Society journals
BioMed Central journals
European Geosciences Union journals
Frontiers journals
Genetics Society of America
Hindawi journals
MDPI journals
Pensoft journals
PLOS journals
some SAGE journals
Springer journals
Versita journals
Ubiquity Press journals
Wiley Open Access journals
Report this comment | #54880
@Ross Mounce – thanks for adding the comment that OA journals usually use the CC-BY license. I believe Scientific Reports is unusual in offering a choice.
@Alan Regenberg – publishers say they may lose money on reprints (a surprisingly large proportion of income for some esp medical journals).
Richard Van Noorden
Report this comment | #54886
Several NPG journals do offer options for Open Access publication including the CC-BY and more "restrictive" options. Oncogene is one of these and actually charges more for "straight" CC-BY licensing. This may result in authors selecting these other options – although, as stated in the article, some authors may choose the NC option because they don't want others making money off their work (not realizing this may restrict access from sites that need some form of advertising to support them). What the OA licensing issue does bring to the table is the whole post publication world – a world most authors are unaware of.
Report this comment | #54887
A Stroll Along the CC-BYways
I don't claim to have definitive answers, but I would like to raise a few relevant points about Open Access (OA) and CC-BY:
1. All researchers would prefer if their research were freely accessible online to all of its would-be users, not just to journal subscribers, because researchers' careers and funding as well as the progress of their research depend on the uptake, usage, applications and impact of their research.
2. Although some areas of research (like the human genome project and crystallography) would benefit from more than just being accessible online free for all, but also from being text-mined, re-mixed and re-published, it is not evident whether all or even most of scientific and scholarly research would.
3. Even if all areas of research would benefit from being open to text-mining, it is not clear that this is more urgent today than being freely accessible online.
4. Most publishers are still subscription publishers today: Over 60% of them already endorse their authors' making their articles freely accessible online by self-archiving them upon publication (no embargo): This is called "Green OA."
5. It is not clear that these publishers would or could endorse immediate, unembargoed Green OA if it also included CC-BY, because that would grant rival publishers the right to immediately free-ride on their content, offering it at cut-rate without having made any investment in it.
6. Authors welcome their ideas and findings being re-used, applied and built-upon, but many would not welcome their verbatim texts being re-mixed and re-published, even with the source cited and acknowledged.
7. The minority of publishers that are OA publishers ("Gold OA"), and are being paid for it, would not have any objection to CC-BY — but would or should all their authors be equally amenable to it?
8. Well over 60% of researchers today don't yet even make their research freely accessible online despite the fact that 60% of journals already endorse it.
9. The exceptions are researchers whose institutions or funders mandate Green OA self-archiving: Over 70% of those mandated to do it, do it (though mandates vary in their strengths and hence their effectiveness).
10. But fewer than 5% of universities and funders worldwide as yet mandate Green OA self-archiving at all (although the number is growing).
11. It is not clear that this is the time to demand that researchers should provide and their institutions and funders should mandate that they provide more than free online access, while most publishers — and possibly some or even many authors — are still opposed to it.
12. It would seem better to grasp first what is already fully within researchers' reach — mandatory free online — rather than to over-reach for CC-BY, and risk doing without either for yet another decade.
Report this comment | #54888
How to design "open access" for research results supported with public money (taxes) so that it get a wide use? Results should be "commons". However recalling the classical fate of commons, namely having got exploited for private benefit, rises justified concerns. Open Access has to be but who is guarding the Commons?
Report this comment | #54901
Thanks for this, Richard, but as I mentioned in our Twitter exchange (@kaythaney if you want to read through for yourself), there's a difference between a license in line with "Open Access" and something that may or may not be slightly more open. I worry that by listing NC/ND/SA flavors as other "choices" is just adding to the confusion, fear and excuse-fodder for researchers, rather than allowing them to make an educated choice. Those listed, beyond CC-BY (a legal implementation of Open Access), are not Open Access. Slightly more open? It's arguable. OA, most definitely not.
Your comments online about this being confusing is well taken, but a number of us are to blame for that. It's not just that there isn't enough written or said about this. I and others have been speaking extensively about this for years. It's a hard fight, but we're seeing some gains, and mandates are making what before an oft-ignored-point of contention an uncomfortable, but necessary and unavoidable conversation.
Ideally, a platform touting "openness" or "Open Access" would provide far fewer license choices than you listed instead of adding to the murkiness – and the licenses provided would be in line with the definition. Take figshare for example (http://figshare.com). Content is under CC-BY if copyright applies and CC0 (the public domain waiver) otherwise. Full stop. No ShareAlike version as that can pollute databases, queries, follow on work, or NonCommercial (which we still don't really know what it means or what the implications are, but people choose it out of uncertainty). Just CC-BY and CC0. You'd be astonished at the amount of people who balk at that when it's clear (minimal, if any at all). It's about create zones of certainty and minimising the faff / strain on researchers, institutions and funders.
Also, looking to the definition of Open Access (accessible here: http://www.opensocietyfoundations.org/openaccess)
"By open access to the literature, we mean its free availability on the public internet, permitting users to read, download, copy, distribute, print, search, or link to the full texts of the articles, crawl them for indexing, pass them to data as software, or use them for any other lawful purpose, without financial, legal of technical barrios other than those inseparable from gaining access to the internet itself."
The "only constraint on reproduction and distribution", it continues, is the right of the author to be "acknowledged and cited" ... in other words, attribution. All of the other CC licenses break this in one way or another.
More publishers, repositories and tools need to take this stance to clear up what is, inarguably, a bit of a mess. OA zealots need to understand the implications of licenses not only for content, but for data, as well, and do their homework before taking the pulpit. As I said on Twitter, this isn't just a battle over semantics, it matters. And we can't afford to continue misleading the public by making them believe that restrictive licensing is in line with OA.
(Note: I formerly managed the science wing at Creative Commons – formerly known as "Science Commons" where we worked extensively on these issues. I'm now at Digital Science, supporter of figshare, though that's entirely irrelevant for the points made above. Comments here are made by me as an individual, not on behalf of my employer.)
Report this comment | #54905
I like the article and think the data at hand is highly interesting.
But all the text written above could have been summarized as a single pie chart. And it would have been more informative too...
Report this comment | #54906
@Florian Gysin – I love pie charts! But surely you'd only really comprehend the chart (and the possible limitations of its data) if you already understood the issues about copyright, licensing, and so on. The point of the text written above is to explain those issues, and open up a discussion. Many researchers have never heard of them or don't understand them...
Report this comment | #54913
The share-alike option pretty much renders the non-commercial option redundant, while allowing most legitimate commercial use. What is more, share-alike has the affect of opening up publications that would otherwise be closed. Allowing 'remixing' (ie. derivative use of content) renders the publication even more potent in opening up otherwise closed entities.
In the world of free software, this share-alike is used in this way to great affect. Software is produced by commercial entities that – by virtue of their use of 'copylefted' components – has to be copylefted in turn, and the gain to the community is far superior to what it would have been, had commercial use been excluded in the first place.
Opportunity for some renumeration is retained, as the copyright holder retains the option of dual-licensing, where derivative works are not themselves required to be copylefted, usually in exchange of a suitable fee, should they wish to exercise that option.
Report this comment | #54918
@Timothy Wesson
The Wellcome Trust and RCUK (for gold open access) do not want share-alike because it will limit commercial re-use. Share-alike may be the option that many researchers want (and a cleverer option than non-commercial). We still have an interesting clash of ideologies between those who want CC-BY vs those who want CC-BY-SA... (and I'm not sure that anyone offers a CC-BY-SA license?)
Report this comment | #54928
"[licenses] listed, beyond CC-BY (a legal implementation of Open Access), are not Open Access. Slightly more open? It's arguable. OA, most definitely not."
it isn't really yes/no. Open Access advocates have recognized for years that there may be variant models for different circumstances.
For example, the Open Access Spectrum (OAS), aka How Open Is It initiative was created by PLOS, SPARC and OASPA, "to move the conversation beyond the deceptively simple question of, 'Is It Open Access?' toward a more productive evaluation of 'HowOpenIsIt?'".
Peter Suber's often-cited Open Access Overview discusses common definitions. and notes, "there is some flexibility about which permission barriers to remove. For example, some OA providers permit commercial re-use and some do not. Some permit derivative works and some do not."
Or as Stevan Harnad notes in the comments, the perfect may be the enemy of the good: it's not obvious that all or most scientific and scholarly research would benefit from being text-mined, re-mixed and re-published; or even if so, if this is more urgent today than being freely accessible online.
Demanding CC-BY may undermine efforts to get publishers to permit Green OA archiving, and dissuade many scholars from making their work available, particularly in humanities & social sciences.
I've recently co-launched a new Open Access project for the humanities & social sciences, *Open Library of Humanities, * in consultation with Peter Suber, PLOS, and many other OA advocates. While CC-BY licensing is favored, I also consider that non-STEM fields have been relatively peripheral in Open Access discussions of recent years, and we need to build a case that engages with the real concerns of scholars in the field.
At a minimum, we can observe there is widespread perception that HSS fields have different practices and expectations than STEM fields, so it may not work to just tell people to get with the program, or demand that "a platform touting 'openness' or 'Open Access' ... provide far fewer license choices...instead of adding to the murkiness."
To me, it is a cautionary flag when the only explanation found for opposing views is misunderstanding. Possibly, that is the reason for them, but possibly too, the alternate positions have not been sympathetically considered, or there is a different conceptual or contextual frame to be understood.
I'd also suggest that exploring this frontier is a productive engagement for scientific OA itself: to better understand, for example, alternate ideas/psychologies of authorship, "moral rights," dissemination of work to more popular & non-expert audiences, and different economic structures of non-science fields. Other fields may have different notions of "research" and its communication than science typically considers. One value of the humanities is in helping science better understand itself.
.
Tim McCormick
tmccormick
Open Library of Humanities
Palo Alto & London
Report this comment | #54930
The systems biology community might be more prone to adopt an open licence. Molecular Systems Biology (www.nature.com/msb ), which is co-owned by EMBO and NPG, offers since many years the choice between CC BY-NC-SA and CC BY-NC-ND.
In 2012, 78% of our authors chose the more 'open' licence (CC BY-NC-SA) vs 22% for the no-derivative version (CC BY-NC-ND). The pie for those who like pie ;-)
Choice for CC BY is going to be introduced soon, together with CC0 for source data and datasets.
Report this comment | #54942
It is widely accepted that for an article to be called open access it should be “digital, online, free of charge, and free of most copyright and licensing restrictions†(see http://legacy.earlham.edu/~peters/fos/overview.htm). And a common way articles are made open access is that they are licensed under a worldwide, irrevocable public copyright license that allows the public to freely reuse, distribute, and create derivative works based upon the work for any purpose. The BOAI10 Expert Group on Open Access recommends CC BY as the preferred license for open access publication (see http://www.opensocietyfoundations.org/openaccess/boai-10-recommendations).
Some arguments against full open access (CC BY) note that authors are in the best position to make judgments about how their work may be reused, and by extension they should be able to choose the best license to accomplish thisâ€â€such as a license more restrictive than CC BY. At the same time--and with regard to the RCUK open access policy mentioned in the article--it is clear that a funding body is entitled to set commercial terms as a condition of funding just as publishers do as a condition of publishing. A grant recipient may choose or decline to accept those terms when funding is offered. The choice is theirs to make. They can accept the funding and the accompanying terms, or decline of their own free will if they disagree with the terms offered. The government should be granted a wide berth in setting the terms for the copyright of the materials its funding generates. It is reasonable for funding bodies to require a specific license if such a license will be the best way to fulfill the mission and goals of the funding program.
Report this comment | #54963
Whenever I read something like "It is widely accepted that..." alarm bells ring. Accepted by who? By the research community as a whole? By the researchers themselves? And do they have a fully understanding of the many different interpretations of terms such as "Open Access" and their implications? I don't think so.
The closest experience that most researchers have with the concept of Open Access is indeed Open Source Software, and the most common forms of OSS licences with Copyleft. This intuitively works for many researchers – the idea that they are happy to provide the output of their research for use to anyone as long as those people do likewise. On the other hand, the idea that we should provide our research output free for corporations to redistribute and profit from is something that a huge number, possibly a majority, of researchers would be thoroughly against. There are possibly cultural and national differences involved here, but if I were to take a straw poll of academics, postdocs and postgraduate students in my department, I'm pretty sure I'd have at least 80% who wished to make their work freely available under a Copyleft-type licence but not under a licence that allowed commercial, for-profit redistribution or exploitation.
If the RCUK opinion is different, then I would guess that it is way out of touch with the majority of the UK research community on this point. But that's just my guess – what we need is the scientific community to be properly informed of different OA flavours and their implications, and then polled on which flavour(s) are acceptable to them.
Report this comment | #54965
@Jim Woodgett and @Thomas Lemberger:
Thanks for noting that some other journals do offer choices of open licenses. It'll be very interesting to see how things change at EMBO when CC-BY is offered. I still think it's the case that most journals offer one open license to their authors (and, as Ross Mounce has pointed out, usually this is CC-BY).
@Tom Johnstone:
Couldn't agree more: 'what we need is the scientific community to be properly informed of different OA flavours and their implications, and then polled on which flavour(s) are acceptable to them'.
Report this comment | #54989
An update: Creative Commons have released a comprehensive explanation of CC-BY and its application to scholarly research, countering some common myths. You can find that on this wiki: http://wiki.creativecommons.org/BIS_committee_UK_OA_Policy
You need to be registered with Nature and agree to our Community Guidelines to leave a comment. Please log in or register as a new user. You will be re-directed back to this page.