Marc Couture — 2019-01-18
One of the issues raised by researchers in the discussions on Plan S is the requirement of disseminating scientific works, in journals as well as on repositories, under an open licence. The original Plan S statement talked of fulfilling “the requirements defined by the Berlin Declaration”, expressing a preference for the CC BY licence.
This was later detailed in Guidance on the Implementation of Plan S. While still referring to the Berlin Declaration, the document offers its own definition of reuse rights. This definition is very close to what is found in the description of the CC licences on the Creative Commons website (Table 1). This appears justified, as the CC licences, as well as our understanding of copyright issues related to OA, have much evolved since the time of the Berlin Declaration (2003, not a year after CC licences were launched).
Reuse rights in Plan S, CC licences and the Berlin Declaration.
|Guidance document (section 8 Licensing and Rights)||CC BY page on CC website||Berlin Declaration|
|… license to share (i.e. copy and redistribute the material in any medium or format)||You are free to:Share — copy and redistribute the material in any medium or format||… a license to copy, use, distribute, transmit and display the work publicly […] in any digital medium|
|and adapt (i.e. remix, transform, and build upon the material) the work||Adapt — remix, transform, and build upon the material||and to make and distribute derivative works [in any digital medium]|
|for any purpose, including commercially,||for any purpose, even commercially.||for any responsible purpose|
|provided proper attribution is given to the author.||You must give appropriate credit, …||subject to proper attribution of authorship|
The Guidance document then goes further by:
- reiterating that CC licences are recommended for all scholarly publications;
- requiring CC BY (and accepting CC BY-SA and CC0) for scholarly articles;
- forbidding the NC (NonCommercial) restriction (again referring to the Berlin Declaration).
The text is less clear though on the ND restriction, which applies notably to translations. It is allowed for works other than articles, the most salient instance being monographs, which form a substantial portion of research output in social sciences, and even more so in the humanities.
The text states only that this restriction “should not be necessary”, arguing that authors’ “legal and moral rights” are protected by the “rules of good research practices” and the Berne Convention. This may seem justified: the former are often treated under the notion of “responsible conduct of research”, while the latter gives authors the right (part of what is called the moral right of integrity) “to object to any […] derogatory action, in relation to the said work, which would be prejudicial to his honor or reputation”.
However, although most countries are parties to the Berne Convention, one must know that moral rights for literary works don’t exist in the US, and don’t apply to translations in the UK . The protection provided by moral rights is thus not universal. Furthermore, the legal code of all six CC licences (CC0 not being a licence proper) states that the Licensor agrees to waive or not to assert their moral rights, up to a “limited extent”. The exact meaning and possible consequences of this provision (parag. 2b1 of CC BY legal code), especially the exact extent of the waiving/non assertion, are extremely difficult to assess though, due in part to the convoluted way in which the clause is construed .
In any event, except in a few jurisdictions (France and Japan, notably) moral rights give authors at most very limited protection against possibly problematic uses, as compared for instance to the absolute control over adaptations or commercial uses conferred by the ND and NC restrictions.
As to ethical principles and rules, enforced by various policies (mainly institutional), one can imagine them playing a role mostly through the notion of plagiarism. However, the attribution requirement of all CC licences means that any work integrating CC-licenced content must refer clearly to the original and state if changes have been made (though not the nature of these changes).
This raises the possibility of meeting the definition of plagiarism, that is “using the words of others presenting them as one’s own”, assuming the licence conditions are respected. At most, one could argue that a busy, casual reader could gloss over, or misunderstand, the licence information displayed in the new work. The impression that all the content originates from the reuser could also be reinforced by the fact that reused parts wouldn’t be highlighted (by quotes, colour, smaller font and/or wider margins), unlike what ethical rules stipulate for excerpts reproduced verbatim. However, one has to ponder here the respective responsibilities of the reader and the author.
In any event, it would certainly be a good practice for authors reusing CC-licenced content to explain, in a succinct but clear manner, the changes they have made, even if the licence doesn’t require it. Failure to do so could be interpreted as obfuscation, or even misrepresentation, which could be viewed as a form of scholarly misconduct, for instance if the adaptation consists only in making a few minor changes or corrections. From an ethical perspective, this situation bears some similarity with multiple authorship, in which case it is difficult to assess the respective contributions of named authors, in the absence of specific information to that effect.
It’s thus less than clear if this dimension of integrity is protected in the absence of the ND restriction. One must also consider that infringement of copyright can only be fought through legal action, which almost never actually takes place for scholarly works, due mainly to the costs involved .
Nevertheless, the question remains as to the necessity, or relevance of this protection not only for authors, but also for scholarship and its wider audience.
The individuals and organizations participating in the discussions of Plan S have expressed a number of statements related to reuse rights. This is hardly a new debate though. Anybody who has witnessed the OA movement  since its beginnings (arguably when the Budapest Open Access Initiative was launched, in early 2002) is aware that the issue of reuse rights, even the basic redistribution right common to all CC licences, was regularly discussed. The discussion peaked at certain periods, notably in 2003, when PLOS and BMC adopted the newly forged CC licences, and again in 2012-2014, in the wake of Finch report and the subsequent RCUK policy requiring CC BY for OA publishing.
For some, adaptation rights, or even general reuse rights are unnecessary or unimportant for scholarly texts, or at least scholarly articles. Immediate and perpetual toll-free access are viewed as sufficient, user licences being at most desirable. To others, on the contrary, reuse rights are as important, if not more, as toll-free access, and are an integral and indissociable part of Open Access (see Table 2 for actual statements).
Most statements though are of a less general nature, focussing upon the advantages or downsides of giving – or not – blanket reuse rights. Regarding adaptations (or derivatives), various considerations are put forth (see Table 3 for actual statements).
In support of allowing adaptations (not using the -ND restriction), one finds:
- compatibility issues when combining (unmodified) CC-ND licenced works with others bearing different restrictions;
- difficulty of assessing what constitutes exactly an adaptation (or “derivative”);
- difficulty of assessing the quality of a translation in a language in which one isn’t fluent;
- importance of non-ND-licenced content for certain types of works: encyclopedias, collections, catalogues, educational resources in general, as well as works based on text-mining.
In support of controlling adaptations (using the -ND restriction), one finds:
- risk of “flawed” adaptations, notably translations, “corrupting” works by substantially changing their meaning or introducing errors, thus putting at risk the reputation of authors, scholarship itself, and(or) the public;
- risk of plagiarism (see discussion above);
- ethical concerns in works based upon studies with human participants (including collectives), whose rights could be affected or consent overridden by unwarranted adaptations.
General statements about reuse rights
|“… music and most books are not like software, because they don’t generally need to be debugged or maintained. Without that requirement, […] the rational incentives for some equivalent of open-sourcing therefore nearly vanish.” E. Raymond (2001)|
“There is no question – or need – of republishing [the texts that constitute the “body of research”] or altering them.” S. Harnad (GOAL, 2003)
“many open access supporters […] feel that reuse and redistribution are as, if not more, important than free access” M. Eisen (GOAL, 2003)
“re-use and re-distribution rights are vital to open access” M. Cockerill (GOAL, 2003)
“It is not even clear whether a blanket right to mix-and-match scientific content (even with attribution) would be a good thing. (Figures need not be re-posted, for example; the OA version can be linked; HTML even allows pinpoint linking to the specific figure rather than to the document as a whole).” S. Harnad (GOAL, 2007)
Statements in support or against the ND restriction in GOAL/american-scientist-open-access forum and other venues.
|Arguments in support of the ND restriction||Arguments against the ND restriction|
|“Authors of peer-reviewed papers [want] to be ensured that the text is not altered or corrupted in any way.” S. Harnad (GOALa)||“The NoDerivatives restriction can lead to significant problems with the combination of different content, e.g. in remixing, sampling or joined publications.” J. Kreutzer (2014)|
|“… changing wording could substantially change the meaning of a work […] a flawed derivative (in the eyes of the original author) could have a negative impact on the author’s reputation.” H. Morrison (GOAL 2012)||“To determine which uses are adaptations is much more difficult […] To which extent licensees can republish adapted material, will vary from jurisdiction to jurisdiction.” J. Kreutzer (2014).|
|“[With CC BY, it is] impossible for an author or publisher to exert any control over how translations are prepared and published abroad”, so no protection against an “intentionally bad translation” or a “poor one innocently done.” S. Thatcher (2012)||“… while scholars do not necessarily want to grant blanket rights for all kinds of derivatives, many would be okay with blanket permissions for translations.” H. Morrison (2013)|
|“… since the potential translator would actually need to contact the author to obtain permission to translate the work, such an arrangement could facilitate the formation of the sort of trusting relationship necessary for a good translation.” J. Britt Holbrook (2018)||“… it is very difficult to check a translation unless the original author can read the language into which their work has been translated.” Danny Kingsley (2016)b|
|“[With CC BY] …plagiarism, and republication of an author’s work will be possible.” Editors from 21 UK history journals (2012)||“CC-BY licences do not encourage plagiarism.” Danny Kingsley (2016)|
|“… the work could be altered so the meaning is entirely changed and it would still be attributed to the original author.” D. Kingsley (2016)||“Free-software-like licences are appropriate for textbooks and encyclopaedias and, generally, for educational resources.” B. Lang (GOAL, 2002)|
|“In disciplines in which the content of an argument or analysis depends exclusively or mainly on the words used and the way the sentences are constructed, authors should have the protection from misuse, misquotation or mistranslation.” British Academy (2018)||“Educational resources need to be modified and translated in order to make them useful in other parts of the world or for different target groups.” J. Kreutzer (2014).|
|“If presented in the wrong way, altered research outputs could affect not just their research but also participants.” D. Kingsley (2016)||Assembling a collection of articles fragments (like methods), “may in itself be extremely useful contribution to science” M. Cockerill (GOAL, 2003)|
|“[…] the researcher enters in a covenant with [participants in a research in oral history] about how their work can be used. This would not be able to be dealt with ethically under a CC-BY licence. The issue is about subsequent control over reuse of research, with concern about it being co-opted and used in another context.” D. Kingsley (2016)||Example of “practical implications” of “giving the right to redistribute and create derivative works from the articles” : including published species descriptions (and pictures) in a catalogue M. Cockerill (2003)|
|“… a research article in the area of pharmacology. Someone creates a derivative – but makes a mistake on the dose. People die.” H. Morrison (GOAL, 2012)|
|“If the work of a serious scholar is deposited in Wikipedia, it could be “corrected” by someone with far less knowledge. […] it seems highly likely that the result will be incorrect citations of scholar’s works. This would harm rather than benefit scholarship.” H. Morrison (2013)||“There are definitely benefits to including scholarly materials in Wikipedia – this would certainly improve Wikipedia – but potentially downsides for scholarship as well.” H. Morrison (GOAL, 2013)|
|“… case report databases, text-mining enhancements and data visualizations are all examples of how additional value can be created by allowing derivative use.” C. Redhead (2012)c|
|“If you produce a data set based on the [text/data] mining, then that data set might be considered an adaptation of the original work[s].” R. Dernoncourt (2017)c|
a. The american-scientist-open-access forum was renamed GOAL in late 2012.
b. Note that D. Kingsley (2016) doesn’t present Ms Kingsleys’ opinions: it is an overview of opinions expressed by participants in a roundtable at Cambridge University.
c. The last two quotes had first been put in the wrong column (corrected on 2019-01-20).
These issues are certainly worth exploring; indeed, some have been extensively discussed. But what I find most troubling is that these statements express opinions or hypothetical scenarios, never facts or actual situations. For instance, there is widespread concern about bad adaptations harming authors’ reputations, but (to my knowledge) not a single example of such a dire situation.
In the same vein, many arguably valuable reuses are envisioned, but (perhaps with the exception of Wikipedia), one encounters very few, if any, descriptions of adaptations having been made possible of facilitated by non-restrictive user licences. This is somewhat surprising, as CC licences have been used in scholarly publishing for 15 years. In the words of a participant in a recent roundtable on CC licences (Kingsley, 2016), “Are we just repeating myths?”
I think particularly of Open Educational Resources (OER), for which the availability of non-ND-licenced works is crucial, not only because educational content (open or not) must be adapted in many ways (translated, abridged, simplified, locally adapted, etc.) when reused, to accommodate various contexts and target audiences, but by virtue of the very definition of OER, which states that they must be “released under an open license that permits […] adaptation […] with no or limited restrictions.”
Also, it would be interesting to have an idea of the actual number of adaptations that have been made of scholarly CC-licenced works. While I reckon that such a study could be difficult to do, I can imagine that automated analyses could be used to some extent, as the CC licence information is often machine-readable.
For these reasons, personal stances on the relevance of allowing – or not – blanket adaptation rights are based essentially upon beliefs, and involve a quite subjective evaluation of the trade-off between hypothetical foreseeable benefits and potential risks. Having become convinced that the former largely overcome the latter, I ceased a few years ago to use any restriction for the CC licences I put on my works.
- See, respectively, the Copyright Law of the United States of America, §106a and the Copyright, Designs and Patents Act 1988, s. 80(2)(a)(i). ↑
- Waiving of moral rights (or agreement not to assert them) must be limited to what “is necessary to allow [the user] to exercise the licensed rights”. The question is: can we say of a Licensor choosing CC BY, but keeping the possibility of using (retrospectively) moral rights to oppose to certain adaptations, that he or she nevertheless allows users to exercise the right to make an adaptation? If the answer is yes, on the basis that users don’t have to obtain upfront the Licensor’s permission to adapt the work, then it is never necessary for the Licensor to waive their moral rights, meaning that the clause is completely pointless. ↑
- Indeed, as far as I can tell, very few cases involving the moral right of integrity have been brought before the courts, none of them concerning a scholarly work. Following a similar train of thoughts, Kreutzer (2014) concludes (concerning the NC restriction, but it would also apply to ND): “if the licenser is not willing or not able to enforce potential violations of the NC restriction by taking legal action, it hardly makes sense to impose it in the first place.” ↑
- For this piece, I searched exhaustively for mentions of CC licences and of reuse in the archives of the forums I knew such issues were discussed. The main one was GOAL (and its predecessor American-Scientist-Open-Access). I also searched for relevant posts in some others (CNI-copyright, Lib-license, ScholComm, open-science), and read (or re-read) the texts that were referred to in these posts. ↑