“copyleft” by eflon is licensed under CC BY 2.0.

Contributed by Jonathan Poritz.

It was my great pleasure to participate in planning the 10th anniversary celebration of the Open Education Network‘s Open Textbook Library [OTL] in the spring of 2022. Since I tend to think of the OTL as one of the main, canonical OER repositories, it was fun to make graphs with some of its publicly available data, as they are, to me, an almost canonical snapshot of the OER world as it exists now.

I was particularly interested to make this graph showing how popular different licenses are for the works in the OTL:

One way to read this is as:

• a tiny handful of OTL creators want to give their works to the public without any strings whatsoever
• around a quarter of OTL creators are simply taking the widespread advice to use the most permissive of the Creative Commons [CC] licenses, CC BY
• two groups, both around the size of those who use BY, have specific reasons to use BY-SA or BY-NC
• a large group of OTL creators wants to use the most restrictive possible license that still counts as OER by the UNESCO definition, CC BY-NC-SA
• a last group of less than one twentieth of OTL creators doesn’t even want to use a license allowing remix, which of course means they’re not really making OER at all.

Let’s talk through the specifics of these different license options, and see if we can clarify some of the issues which might help creators choose.  By the end of this piece, I hope to convince you that there are good reasons to prefer licenses containing the ShareAlike (SA) provision for OER.  This is not an uncontested position: see, e.g., David Wiley‘s The SA Fallacy.

Recall that copyright vests automatically in “works of original authorship fixed in any tangible medium of expression,”¹ meaning that works with a scholarly or pedagogical purpose have to fight against the restrictions of copyright in order to reach their possible audiences. In the US (and some other countries), the powers given to copyright holders are supposed to provide an incentive for increased productivity via the expectation of monopoly profits: this is the so-called utilitarian rationale for copyrights².

While this, very arguably, is appropriate in works of popular culture, it seems to me that monopoly control stands in nearly perfect opposition to the values of education and scholarship.  As the American Association of University Professors put it in one of the foundational documents which codified the idea of academic freedom in the US,

“Institutions of higher education are conducted for the common good and not to further the interest of either the individual teacher or the institution as a whole.“

(Presumably the same thing also applies to primary and secondary education — maybe even more so, given the basic role of such education to help create the next generation of citizens in a society.)

It doesn’t seem that private monopoly profits do much for the common good, so I would maintain that in education we should always have a very suspicious approach to the role copyright plays.

Creative Commons licenses instead seem perfectly designed to the norms and realities of the academic world³. First, all CC licenses allow reproduction and public dissemination/performance without individual permission – or financial recompense – to the rightsholder: quite the opposite of those expected monopoly profits! In fact, CC-licensed materials effectively can never be a commodity sold purely as a profit center; or, you can try, but once a single copy is out in the public with its visible CC license, consumers have no motive to pay for any other copies.  In the world today where pretty much everything tends to be viewed in terms of its potential to make a profit in some marketplace⁴, it’s easy to wonder why anyone would ever make CC-licensed materials, to which there are two good answers: one is that there are plenty of areas where self-interest doesn’t dominate, but rather some other motivation – like a pursuit of the common good, as the AAUP recommends for higher education; another is that, like was seen in the open-source software movement, an openly licensed artifact can be given away by a company which sells services around that object, and much good business can be done that way.

Next, all CC licenses contain the Attribution or BY clause, which essentially uses the power of copyright law to require reusers to “give credit where credit is due.” This is of course a fundamental tenet of good scholarly practice, one we teach students to follow, and whose breach we punish with some of the most severe academic penalties.  While it may seem perverse to use the law to enforce subcultural norms, BY certainly is consistent with those norms and the plain CC BY license can be seen as the simplest way to license a work that maximizes its usefulness in scholarship and pedagogy.

For creators who want to keep copyright law completely out of the picture for their academic work, there is also the CC0 Public Dedication tool which simply puts a work immediately in the public domain, abjuring all copyright control completely.  CC0 is better than an author simply noting on their work that they have chosen to put it in the public domain because there are jurisdictions on the planet which do not allow creators to do this, while CC0 has various mechanisms that should have the same effect and that work globally.

So much for CC licenses and tools which subvert the basic obstacle that copyrights pose for scholarship and pedagogy, so good for those who want to use the simplest solution to the problem of copyrights.  The remaining CC licenses, however, acknowledge some realities of the way creators feel about the uses which may be made of their works.

The NoDerivatives [ND] clause, in licenses CC BY-ND and CC BY-NC-ND acknowledge the fact that creators sometimes feel such a strong sense of identification with their work that it would cause them pain if it were modified in any way, even while they are happy to have free public distribution/performance.

While this is a reality, it is perhaps an unfortunate reality, at least in the academic context: to say that our knowledge is perfect at some point and must never be changed seems, at least to me, to be very much contrary to the ideals of academia⁵ Perhaps because of this, the UNESCO definition of OER does include language which clearly excludes the ND clause.

Another CC license clause is NonCommercial [NC], in licenses CC BY-NC and CC BY-NC-ND, which rules out the use of the shared work “for commercial purposes.”⁶ While this clause causes some nervousness in the community, I like to think it actually isn’t too complicated: it excludes uses primarily intended to make money (or profit).  So a campus bookstore can sell a course pack with an NC-licensed work in it at the cost of reproduction, but it may not add a profit margin to that sales price.  Note also that the NC clause applies to the use, not the user, in that a for-profit entity could use an NC-licensed work without violating the license and a non-profit could violate an NC clause, both depending upon the use they make of the work.

In my experience working with OER adopters and creators, they tend to choose the NC clause based on reasoning along the lines of “I did all this work and I’m happy to let the public use it without paying me … but if someone is going to make money off of it, it shouldn’t be someone else!” Working within the academic world, for the public good (right?), it seems to me that this reasoning is making a big deal out of a side issue, but I have no great objection to it.  Since the ability to make a large income from a CC-licensed work is already seriously compromised by the free public sharing/performance right all CC-licenses have, the NC clause is preventing something that is not likely ever to happen in a big way, even without NC.

The last CC license clause to discuss is ShareAlike [SA], in licenses CC BY-SA and CC BY-NC-SA, which requires adapters of the work to apply the same license, or one of a very short list of “compatible licenses”, to their adaptation. This kind of self-perpetuating licensing was called — very derogatorily! — “virality“, when it first appeared in the GNU General Public License [GPL] for software⁷, because it passes like a virus from the original work to all derivative works made from that original. A less terrifying way to think of this licensing feature is as copyleft, a kind of intentional mirror image of copyrights, where the original rightsholder has sole and despotic dominion⁸ over not only the work but all of its future adaptations; under copyleft licenses, through the looking glass, the rightsholder abjures not only control of the original, but also forces that same freedom on all future adaptations, no matter by whom, since they must continue applying exactly the same license to those adaptations.

Note that the SA clause is once again offering a possibility for creators to control some aspect of the uses of their work, as do all the CC license clauses.  SA does not prevent adaptations, as ND does in a way I argued goes against the scholarly ethos.  Nor does it prevent commercial uses of a work⁹ like NC. Instead, the copyleft SA clause locks not only a work but also all of its downstream derivatives, on through the remixing generations, into the open.

Effectively, a creator who uses an SA license is saying: “I did all this work for the public good, and I am happy for it to be shared. I may or may not care if it is used to make a profit (see whether I included the NC clause or not). But what I will not tolerate is for anyone to make an adaptation of my work but place that adaptation under their own all-rights-reserved copyright [ARR (c)], preventing that work from being freely shared and adapted for future public good.”

Because, of course, works released under what is said to be the purest and most open OER license, CC BY, do have the possibility of being adapted and those adaptations kept under ARR (c). Creative Commons has a chart of licenses which can be used on adapted works which leaves out this possibility, perhaps to encourage folks to think something like “once in the open, always in the open.”  But a more complete chart — here is an adaptation license chart I made – shows that this is a distinct possibility.

Unfortunately, SA isn’t a perfect tool to use copyright law to enforce two norms of the academic world (attribution must be given and all future descendents of this open work should also remain open): it can provide an obstacle to future remixing.  There are two licenses with the SA clause, one with the NC clause as well and one without.  Any adaptation of a work or works that uses one of these licenses must be licensed in exactly the same way.  This means that there is no single license which will work on a remix of both a CC BY-SA and a CC BY-NC-SA work: such remixes are not legally possible¹⁰. Potentially, then, there could be two distinct streams of permanently open literature that cannot be remixed across streams¹¹, one under CC BY-SA and one under CC BY-NC-SA.  Many additional details on these remixing scenarios are in these Remix Compatibility Charts.

The talk above of “forcing” and “locking” may make some open practitioners nervous.  But remember that we are trying to build a better world, and large, monied interests are trying to prevent us from doing that.  Just as copyleft licenses were a big part of how free/libre/open-source software [FLOSS] became dominant in the basic running of the internet¹², they can help open educational resources become the future infrastructure of education.

Where does this leave us, as OER adapters or creators who want to choose a license? Perhaps the tl;dr is:

• ND isn’t OER.
• NC is fine, but probably doesn’t deter profit-seeking entities any more than all CC licenses already do – remember, they all prevent commoditization of CC-licensed works already.
• Plain BY is simple and flexible, but it leaves open the possibility that bad actors (from the point of view of the public good) will make adaptations which will not be open at all.
• Copyleft SA prevents those BY→ARR(c) bad actors, but may cause difficulties for remixing between the two different SA licenses.

Practitioners must balance how concerned they are that individuals or corporations will want to lock up the intellectual children (adaptations) of their OER with how much they think that picking one license with the SA clause will close off valuable future remixes with works sporting the other license with SA.

Personally, whenever I can choose, I release my works under CC BY-SA, since I fear openwashing and other bad actions by profit-seekers, which copyleft will prevent, and I do not particularly see the added value of the NC clause, as mentioned above.

The numbers in the chart at the beginning of this post show that my view is not the majority one.  How do you balance the various risks and what license do you end up choosing?

Footnotes

 ¹This language is in US copyright law, although it’s quite similar in all 179 Berne Convention signatory countries, except that the tangible fixation requirement is not imposed in a majority of Berne jurisdictions.

 ²The “Copyright Clause” is one of very few provisions in the US Constitution which gives a justification: “to promote the Progress of Science and useful Arts.”  Another self-justifying clause is the Second Amendment’s right to keep and bear arms for those militias we all depend upon… not sure that’s working out so well, these days….

 ³Perhaps unsurprisingly, as they were designed by a bunch of professors!

 ⁴The word social scientists use for this markets-only worldview is “neoliberalism.”

 ⁵There is a complication here, though, because copyright does not cover ideas, but merely expressions (of ideas!). Nevertheless, I think that even freezing expressions, as ND requires, does not match the academic project, philosophically.

 ⁶Always, of course, without specific individual authorization from the rightsholder!  That is, a work licensed CC BY-NC can be used for commercial purposes, and a work licensed BY-ND can be adapted, so long as the rightsholder gives specific permission to the reuser/adapter.

⁷In fact, the GPLv1 was sometimes called the “GNU Public Virus” by haters: seehttps://en.wikipedia.org/wiki/Viral_license#History.

 ⁸In Sir William Blackstone’s evocative definition of property, from his Commentary on the Laws of England.

 ⁹Or its derivatives, which NC already does: NC is viral, too!

 ¹⁰Without direct permission from the rightsholder(s), of course.

 ¹¹Who said “Don’t cross the streams!”?

 ¹²It is a weirdly unknown fact that much of the structure of our modern life, despite neoliberalism, relies on FLOSS: most of the basic infrastructure of the internet relies on FLOSS, FLOSS operating systems run on more silicon CPUs than commercial ones do, etc.

Jonathan Poritz was for years one of those tattered minstrel mathematicians one sees wandering dusty backroads with nothing but chalk on his coat and a theorem on his lips: in total, he studied and taught mathematics and computer science at around a dozen universities in five timezones on two continents, most recently in a fifteen-year stint at Colorado State University Pueblo.  He has also worked in various IT firms, from start-ups to multi-nationals, doing things like AI, cryptography, and consulting on public policy towards technology. Building on a lifetime of using free/libre/open-source software, Jonathan has for the last decade been involved in open education movement, writing three OER textbooks and serving on the Colorado OER Council.  In his spare time, he teaches the Creative Commons Certificate course and writes snarky articles about the over-hyping of blockchains.  Much more information about him, including (almost) everything he’s written, (almost) all of it released under an open license, can be found at https://poritz.net/jonathan.

This post is by Jonathan Poritz and is released under a Creative Commons Attribution 4.0 International license, except where otherwise indicated. Please reference OER and Beyond and use this URL when attributing this work; for more information on licensing, see our Open Access Policy.