The default in open access: CC-BY-SA

Posted by Frank Huysmans on 17 maart 2013 | 4 Comments

The contribution of Creative Commons (CC) to a less restrictive copyright regime is, I think, invaluable. The great thing about CC is that it gives makers (that is, creators of all sorts: writers, composers, visual artists, photographers, and professions we cannot even think of yet) an opportunity to share their work more openly. And here’s the beauty in it: It does so within the confines of copyright legislation itself.

Licensing your work under one of the six current CC licenses amounts to saying: okay, I created this work, but of all the rights copyright legislation is bestowing upon me as a maker, I am happy to waive some for it to get a wider dissemination. And depending on the license I choose – and I can opt for a different license for different works – I give others more or less freedom to do with it what they want. If I want a maximum reach for my work, I do not care if others re-use it for profit making, and I will not use the Non Commercial-clause. If I want others to respect the integrity of my work, I’ll add the No Derivatives-clause.

CC-BY as the standard, or rather CC-BY-SA?

It looks as though more and more open access advocates opt for the CC-BY as the standard for open access in (scientific) publishing. In a recent piece Michael W. Carroll indeed argues that CC-BY should be the gold standard for open access. Though I can follow his arguments, I am not sure I can agree. I would like to advocate another position: the default option as an open access standard should be CC-BY-SA.

The point is that in the second stage of re-uses (remixes) CC-BY-licensed content is no longer protected from being locked up again.

Why? Simply because the Share Alike-clause guarantees that all future derivatives (‘remixes’) of the work will be open. Using CC-BY, you are never sure that others who elaborate on your work will eventually publish it under a more restrictive license. They can use your work, add a paragraph, and lock it up under full copyright, or under a CC-BY-NC-ND license. What is nice is that you will still be listed as one of the authors of the work. What is not so nice is that this particular remix of your creation will not be able to spread as far as you wished when you threw it into the world.

Stallman’s legacy: General Public License and Share Alike

The power of the Share Alike-clause struck me when I was reading some of the essays by Richard M. Stallman, who turned 60 yesterday. Stallman is the intellectual father of the free software movement and the licenses under which free software is distributed. In his bundle of essays Free Software, Free Society (second edition), he describes how he and his companions were trying to find a way to keep the code they were writing free – in the democratic sense of that word – ‘libre’ – and not necessarily ‘gratis’ (whether or not software is sold at a price or shared without economic transaction is not relevant here).
As coders, they advocated the right for everyone to use, study, redistribute and alter computer code, or even the code contained in domestic appliances, so that these tools better fit the purposes of the user.

But by default, changing code is not allowed by intellectual property (IP) law. IP law bestows alienable rights to creators and awards them exclusive rights to put their concepts to the market for a limited period of time. The logic behind IP law is that this promotes innovation by economically stimulating creators to come up with innovations – to the benefit of society as a whole. This comes at a price, however: it seriously restricts sharing and working together on innovations, which is more often than not helpful or even necessary in today’s highly complex technological engineering. You simply need people with different specialisms to construct something as complex as a stable operating system.

With a simple trick of ingenious beauty, they succeeded to create the free sphere they wished within the existing intellectual property rights regime. The General Public License (GPL) allows others to change the code for their own purpose under only two conditions: (1) that the original author of the code be mentioned and (2) that alterations of the code be published under the same license. In Free Software, Free Society, Stallman describes how the license effectively forced universities and companies to redistribute alterations of GPLed code by their employees as free software – which otherwise most likely would not have happened.

Free: the ethical imperative

I’m not going into the important discussion here about whether free software is the same as open source software (Stallman vehemently argues that open source is a designer problem, whereas free software is a social movement with an ethical imperative). More important here is tackling the question whether what counts for computer code also counts for text and copyright.

I am inclined to think it does. For all the differences in sorts of use, amount of alterations and redistributions between computer code, research data, metadata produced by libraries, of which I am aware, the effort of keeping free information free is an important ethical imperative. It is certainly true that the CC-BY-license awards the users of your own work more freedoms than the CC-BY-SA-license does. The point is that in the second stage of re-uses (remixes) CC-BY-licensed content is no longer protected from being locked up again.

So, to repeat the argument: Suppose somebody else takes your book, alters some phrases, adds a chapter or two, and publishes it under a more restrictive license – let’s say CC-BY-NC-ND. This remixer is perfectly allowed to do so (provided, of course, that s/he mentioned your name as original author). But that is not what you, being that original author, had in mind when you were sharing your work. You probably wanted others, like this author, to use your work and share improvements under CC-BY.

And suppose you are director of a national library that decides to put its metadata, generated with taxpayer’s money by many librarians over decades, on the web under a CC-BY-license or even CC0 (which is, by the way, not a license, but that is another matter). Then you want others to be able to use that database and possibly improve it, adding new records and fields. What expression will take over your face if after a while a company, or another library for that matter, distributes an improved and expanded version of your data set under a license that effectively prohibits you from re-using it yourself?

Why Share Alike should be the default

CC-BY-SA, the Creative Commons equivalent of the GNU General Public License, makes sure that this will not happen – or if it does happen, you will have a very strong case in court. That is the reason why Lawrence Lessig cum suis incorporated it in CC (Lessig’s book Free Culture was heavily inspired by Stallman’s work, the author acknowledges). And it is also the reason why it is the default option in Wikipedia. With the more restricted freedoms it offers for first re-uses (remixes) than CC-BY does come additional freedoms for all re-uses that follow. And so to keep free content free (again, to quote from Stallman’s book cited above, “think of ‘free speech’, not ‘free beer’”). Or to make sure openness is viral and passed on to other projects, as Mike Taylor puts it in one of his excellent tutorials on open access.

That is why, it seems to me, CC-BY-SA should be the default option for open access, not CC-BY. It’s a simple question of how best to serve the public good. Until now, I haven’t heard convincing arguments why CC-BY would be more suited for that purpose than CC-BY-SA. That CC-BY comes closest to the Budapest, Bethesda and Berlin Declarations on Open Access does not imply per definitionem that it is the option to be preferred. For first re-uses it is; for further re-uses it is not, I would say. But I will be happy to change my mind if I oversaw some more powerful arguments!

Please use the “comment’ option or share your thoughts under ‘Geef een reactie’ (sorry for Dutch!).

Creative Commons License
The default in open access: CC-BY-SA is licensed under a Creative Commons Attribution 4.0 International License.

Posted in beleid, opinie, WareKennis | Tagged , , , , |

4 reacties op “The default in open access: CC-BY-SA”

  1. Lukas Koster schreef:

    Interesting appeal Frank. I completely understand your desire to force others who adapt your work to apply the same license as the original work is governed by, in order to make sure it remains free/open.
    However, scholarly publications (and that is what we are talking about with Open Access) are not the same as software. Software is basically a means to an end. If others can adapt the software to achieve the intended end in a better way, or other ends, that it’s fine to do that. But scholarly publications are better seen as “an end” in itself, or rather “intermediary ends” at most. With open access there is a difference between accessibility and use on one hand, and modifiability and re-use, as is stated in the Wikipedia article you pointed to yourself ( “… the text of a research article is not intended for modification and re-use. (In contrast, the content of research articles is and always was intended for modification and re-use: that is how research progresses.)”.
    For this reasons, in my view a CC BY-ND (No Derivatives) license, where texts are “passed along unchanged and in whole” is much more appropriate in the case of Open Access for scholarly publications. I assume that ND applies SA.
    Of course it will still be legal to use parts of the publication as quotes, according to normal copyright laws.

    • Frank Huysmans schreef:

      Hi Lukas, thanks for sharing your thoughts and opinions. Naturally I agree that in the case of scholarly publications, use is much more common than re-use (with modifications). Nevertheless, I would not want to rule out the possibility that unsolicited modifications can improve the scientific article. These days, we see unprecedented changes in the scholarly publication landscape. One of these changes is post-publication peer review, which can help authors to improve their analysis, sharpen their conclusions etcetera even after publication. Right now, it is usually the authors themselves who alter their original article and publish a second, improved version. But since scholarship is a collective endeavour, why not open the publication process in such a way as to elaborate on each other’s work? If in, let’s say, psychological research the results of an experiment are published, other researchers who replicated (or failed to replicate) these results can opt for extending the original article, instead of publishing a separate article. And after a series of replications has been conducted by various groups of authors, they or someone else can easily compile the results of the ‘tradition’ and compose a meta-review, with the earlier contributions as appendices. Note that CC-BY and CC-BY-SA only invite others to make re-use of the work. If scholars want to protect the integrity of their work, they can always opt for CC-BY-ND to make sure re-use does not happen.
      More importantly, the point I wanted to make is that in debates and policy over open access, CC-BY is increasingly advocated as the OA standard – thus, for scholarly publications in the first place! My point is that if you opt for maximum openness of scholarly work, CC-BY-SA might work out for the better in the long run to keep open stuff open.
      Your assumption about ND applying SA is not true, as far as I can infer from the licenses on It seems to me that CC-BY-ND-content can be locked up commercially when reprinted in, e.g., a collection of essays (which, to be sure, leaves the original publication still open). But it would be good to hear a legal scholar about this.

      • Frank Huysmans schreef:

        To support the above argument: take a look at this graph, released by the Open Access Scholalry Publishing Association (OASPA), depicting the number of articles published by OASPA members under a CC-BY license since 2000:

      • Lukas Koster schreef:

        Of course it’s true that in a changing scholarly communication world the traditional unit of publication, the journal article or the book as a monolithic product, will no longer exist as such. This was the topic of the BeyondThePDF2 conference in Amsterdam this week ( In that world CC BY-ND will not do obviously.
        However, current practice still is the scholarly publication as unit of production and measurement. In this world the CC BY-ND license would still apply. There is no CC BY-SA-ND license. So what you say is true. But the original work will still be available under its original license.


Deze website gebruikt Akismet om spam te verminderen. Bekijk hoe je reactie-gegevens worden verwerkt.

%d bloggers liken dit: