Does anyone really understand how Creative Commons licenses work?
It may seem like a flippant question – after all, at a certain level these licenses are admirably clear and easy to understand, and there are lots and lots and lots of convenient explainers out there (even if the details are often quite a bit more complex). It seems as if no one, whether author, publisher, funder, or reader, would really have an excuse for not knowing what they’re getting into when they choose — or are compelled — to adopt an open license for their work.
And yet, here we are.
There are many interesting takeaways from a recent author survey undertaken by the American Association for the Advancement of Science. This survey focuses not on free access to scholarly content, but rather on reuse rights under open licenses – and on the degree to which authors both understand the implications of open licensing and have concerns about how their openly-licensed work might be used by others.
The survey asked authors questions such as these:
- Do you share research results in non-peer-reviewed formats such as on preprint services, blogs, or similar outlets?
- Please rank the criteria you use to select where you publish peer reviewed research results.
- How familiar are you with Creative Commons licenses?
- When choosing a license by which to publish your work, what influences your decision?
- Are there potential commercial uses of published research that [excite/worry] you?
- If you saw misrepresentations or misuses of your published work, how and by whom would you expect them to be fixed?
As can be seen from the published report, the responses AAAS received were fascinating. They include:
- Nearly half – 48% – of respondents rated themselves as less than familiar with how open licenses work.
- 75% share their research results in venues other than formally published peer-reviewed journals.
- 42% of respondents rated “audience reach and audience relevance” as the most important factor when picking a publication venue; “Creating access to work, including for the general public” was the second highest-rated factor, followed by “journal impact factor.” (Unfortunately, both of the two most popular response options conflate two very different criteria: specificity of readership, and breadth of readership.)
- The least important venue-selection criterion was “existing APC discounts or transformative agreements.”
- Authors are quite ambivalent about commercial reuse of their scholarly work, with 36% expressing “excitement” about such use and 63% expressing “concern.” Some respondents reported feeling both. They also expressed both excitement and concern about the use of openly-licensed scholarship by artificial intelligence companies.
- 40% expressed concern about the lack of available recourse in the event of misuse or misrepresentation of their openly-licensed work. (Though what might constitute “misuse” of research published under an attribution-only license is an interesting question.)
- 28% of respondents overall (but 53% of respondents between ages 60 and 70) expressed concern about how their work might be reused under a CC BY license, while 29% of total respondents believe there should be no restrictions on the reuse of published research.
These responses, and the ongoing frustration of those who want to convince authors to be more enthusiastic about adopting open practices, reflect a very difficult balancing act: on the one hand, we may not want authors to be able to gatekeep their research reports and data such that those who interpret them differently in good faith are prevented from advancing their interpretations; on the other hand, we may want authors to be able to prevent the potentially egregious misrepresentation of their work, or its reuse by others whose goals the authors find repugnant. (The CC BY license does allow authors to require those who reuse their work to strip the attribution from their derivative versions; however, this provision is buried in the small print, and not mentioned in the more easily findable summary description of the license. And it addresses only the issue of wishing not to be publicly associated with such downstream work; it does not address the issue of not wanting to support or contribute to it.)
Taking away from authors all control over the use of their work is a founding principle of the open access movement. Should we be surprised that authors are ambivalent about this?
And none of this touches on the trickier and more subtle question of whether authors should be able, for example, to restrict commercial reuse of their work (including, notably, the use of their work to train artificial intelligence large language models), or the creation of derivative versions of their work. Creative Commons licenses exist by which authors can exercise such restrictions, but in order to choose between license options effectively, they have to understand the licenses – and, of course, they have to be permitted to use them. Taking away from authors all control over the use of their work is a founding principle of the open access movement, and continues to be an explicit goal of many of the most influential entities in that community. Should we be surprised that authors are ambivalent about this?
Of course, neither retaining the exclusive prerogatives of copyright nor selecting open licenses that are more restrictive than CC BY can prevent others from misinterpreting or misrepresenting one’s work. But preserving some or all of those prerogatives does give the author or publisher a meaningful degree of control over the reuse of that work, and legal recourse in the event of misappropriation – whereas CC BY is designed specifically to allow any and all reuse. If the author has not reserved any of her exclusive rights under copyright, there is no point in the author worrying about how her work will be reused – she has granted an irrevocable license allowing any and all reuse. That is the express purpose of the license.
These issues will continue to be difficult and vexing as the scholarly communication ecosystem continues to evolve. This will be the case especially for authors, who are increasingly whipsawed between the constantly changing and sometimes contradictory goals and demands of funders, governments, movement advocates, publishers, and institutions.
Discussion
30 Thoughts on "Confused and Ambivalent: Scholarly Authors and Creative Commons Licenses"
Your statement that the use of CC licenses takes away all control is simply wrong. Even the Budapest document you link to says, and I quote, “ the only role for copyright in this domain, should be to give authors control over the integrity of their work and the right to be properly acknowledged and cited.” Integrity, acknowledgement and citation does not equal loss of all control. Instead, it encourages reuse, as the licenses were designed to do.
Emily, I think you and I are thinking of “control” in different ways. In my view, the requirement to acknowledge the original author does not constitute the author’s retention of any control over the work.
In your view, under a CC BY license, what control does the author retain? Or, to put it another way, under a CC BY license what uses of her work can the author constrain?
To answer your question, attribution. What alternative to CC licenses do you propose in an OA environment?
Saying that you must acknowledge me as the creator of the original work doesn’t constitute any constraint on how you may reuse my work.
Sorry, Emily, I failed to answer your question. I have no objection to CC licenses and therefore won’t propose an alternative to them. I think they’re good and useful. That said, I do think it’s important that we be clear on (and not obfuscate) the implications of using them — especially of using CC BY, given the radical effect that license has on authorial control.
As with OA in general, I have no objection to the use of CC BY or any other CC license. I do have a concern about people with power over authors using that power to compel them to adopt CC licenses (or OA).
Perhaps we could reframe it this way. In many cases (see my comment below), just by publishing something, you open yourself up to having your argument misrepresented or misused. Even if your work is behind a paywall and protected by copyright, it can still be selectively cited and quoted from; its argument can twisted and used for harmful purposes. The absence of a CC license does not protect you from that.
By appending a CC BY license to your work, you enable easier access to it. And by requiring others to attribute the work to you, you ease the ability of readers encountering a misrepresented or misused version to find the original work and engage with your actual argument. This isn’t a panacea, but it is a form of protection through openness.
I also think we need to be careful not to reduce everything to the CC BY license. Creative Commons licenses come in many forms, and are designed to give authors varying levels of control based on their specific needs, wishes and desires.
Josh, you and I agree that retaining the exclusive prerogatives of copyright does not prevent having one’s work misrepresented or misused. (Or, as I put it in the post, “Of course, neither retaining the exclusive prerogatives of copyright nor selecting open licenses that are more restrictive than CC BY can prevent others from misinterpreting or misrepresenting one’s work.”)
We also agree that not all CC licenses are CC BY licenses. (Or, as I put it in the post, “Creative Commons licenses exist by which authors can exercise such restrictions” [i.e. restriction on commercial reuse, on the creation of derivatives, etc.])
I don’t agree that applying a CC BY license to one’s work offers anything that can reasonably be called “protection through openness.” One can make one’s work easily and freely available to the reading public while still retaining some or all of one’s exclusive copyright prerogatives.
Would it then be fair instead to say that “the use of CC licenses takes away almost all control” for authors? As you note, the Budapest document clearly states that rights are to be drastically reduced. Also, as far as I know, there’s nothing in a CC BY license that plays any role in citation (attribution is a different concept and is required under different circumstances).
In practice, I feel like attribution equals citation in most cases with published works. Can you point me to a study where this assumption is proven incorrect? What is your proposed alternative to CC licenses?
My understanding is that the CC licenses require attribution when the original material is re-used, that is, the specific words and images in the specific order used by the original author are re-published or re-distributed. So if I re-publish your paper, I need to attribute it to you. This is an entirely different concept from citation, which reflects use of the article, not re-use. If I read your article and it influences my subsequent work, I should (under conventions of the field, but not under requirement) cite you. But that behavior has nothing to do with a CC license and is not covered in any way by those licenses.
As for my proposed alternative to CC licenses, I don’t have one, but then again, I’m not advocating for their use nor the activities they cover.
I think it’s important to remember that the BOAI statement pre-dates the development of CC licenses. CC BY is closest to the BOAI copyright statement, but not a perfect mapping. (Want to hear more on this? Catch me in the SSP OA 101 workshop next week!) It is also really important to remember that there are use rights (fair use/fair dealing) that are not at all dependent on a CC license (see some of the conversation above), and a CC license can’t restrict that … it can only enable an expansion. Sometimes authors seem to think that the problem is the CC license when, in reality, the root cause is having published it!
I would like to echo Emily’s point and add that part of the purpose for many in the open access movement was to allow authors to *regain* access to/control of their own work. Some scholars and authors joined the movement precisely because they felt it was unfair that publishers held the copyright to their work and could thereby restrict their access to it.
At the same time, it is true that making work available under open licenses means it can be misrepresented or misused. It is an act that, to some extent, involves authors giving up control over their work. But to a lesser extent, the same is true whenever an author publishes something: the very act of printing and distributing a work exposes it to the risk of misrepresentation or misuse. This is precisely why copyright law was introduced—to provide authors with protections that would encourage them to publish, thereby promoting innovation and creativity. In many ways, I see the best aspects of the open access movement as pursuing those same goals through open licensing.
part of the purpose for many in the open access movement was to allow authors to *regain* access to/control of their own work
I respectfully but strongly disagree. There is nothing in the push for mandatory CC BY that can reasonably be construed as an effort to give authors more control over their work. That license leaves the author with no control whatsoever. (Of course, the same is true when authors assign copyright to publishers, which is one reason I believe authors should not be required to do that.)
At the same time, it is true that making work available under open licenses means it can be misrepresented or misused.
As noted above, though, you and I agree that making one’s work publicly available under any rights regime (including the authorial retention of all copyright prerogatives) means the work can be misrepresented or misused. That’s not the issue. The issue is whether and to what degree the author retains any ability to constrain reuse. Under CC BY, he doesn’t.
There is nothing in the push for mandatory CC BY that can reasonably be construed as an effort to give authors more control over their work.
I’d request we refrain from conflating “mandatory CC BY” (your words) with the “open access movement” (my words). The open access movement cannot be reduced to a program of “mandatory CC BY”. Indeed, none of the three major open access declarations even mention creative commons licensing and each has a different understanding of what “open access” is. Certainly some parts of the movement are *at this moment* for mandatory CC BY. But not all of them are.
Can you point me toward any significant examples of OA advocacy organizations that are not working towards a future in which all scholarly publishing is required to be OA with CC BY (or a functionally equivalent license)? I’m not aware of any, but of course I can’t claim to know everything about every OA advocacy group.
There is a massive difference between an author choosing to irrevocably waive certain rights in order to choose an open license and being forced to irrevocably waive those rights by a funder or employer. In the former, it is a choice and constitutes a measure of control and agency for the author, assuming they fully understand the implications of an open license. In the latter, it’s functionally forced copyright waiver (and may or may not be legal in regard to the federal government).
In the US, open licenses are generally unnecessary for readers to read and cite a work, which raises the question of why funder mandated open licenses should be the default or the preference instead of a case-by-case author decision.
It’s important to understand that the OA movement’s goal is not just to make scholarly content free to read, but also to make it free to reuse without restriction. The problem with leaving the decision up to authors is that many of them would then choose not to adopt a CC license, or opt for a more restrictive one than CC BY. So the choice to do so has to be taken away from them. (In the immortal words of cOAlition S founder Robert-Jan Smits: “Why do we need Plan S? Because researchers are irresponsible.”)
The overall impression that I came away with after reading your article was that your goal is to instill confusion and doubt about open access publishing. If an author wishes to maximize the control over their writing, their best option is to not publish it at all. Then nobody can misinterpret it, nobody can misrepresent it, and nobody can reuse it. Authors typically want to do the exact opposite – they want to add to the world’s knowledge and allow others to build upon it. The CC-BY-NC license addresses one concern that authors may have about commercial gain from reuse, and the CC-BY-NC-ND addresses further concerns about derivative work. No license will be perfect, and neither is copyright law. None of it will prevent misrepresentation or unforeseen consequences that the author finds repugnant. Do you have a proposal to address your specific concern? Is it just to say that publishers should be more flexible in which license they will accept and not blindly recommend only CC-BY?
I’m not sure the authors who responded to this survey showed themselves wanting to “maximize control” over their work. They did express concern about certain parameters of control — such as commercial reuse.
The other points you’ve made in your comment either don’t really seem to respond to anything I said in my piece, or seem to be based on a failure to read things I actually did say. (For example, I made the same point about the variety of CC licenses that you did.) Is there something I did say that you found objectionable? In your view, did something in my piece create “confusion and doubt about open access publishing”? If you’d like to share what I said that you disagree with, we might be able to have a productive conversation about it.
One thing that immediately annoyed me was that you based your article on an author survey conducted by the American Association for the Advancement of Science. That is immediately suspect as a source to say anything meaningful about open access. They charge an APC of $5450 to publish in the journal Science under a CC-BY or CC-BY-NC license. According to their public 990 IRS form, they received $42M in revenue from subscriptions in 2023, and total revenue of $123M in 2023. They paid their editor H. Holden Thorp a base salary of $654,263 in 2023. It is in their interest to sew distrust and confusion about open access, so it is in their interest to phrase questions that heighten uncertainty and confusion among authors. To quote Upton Sinclair: “It is difficult to get a man to understand something, when his salary depends upon him not understanding it.”
Putting that aside, you made the statement that “…on the one hand, we may not want authors to be able to gatekeep their research reports and data such that those who interpret them differently in good faith are prevented from advancing their interpretations; on the other hand, we may want authors to be able to prevent the potentially egregious misrepresentation of their work, or its reuse by others whose goals the authors find repugnant.” While we seem to agree that the NC and ND clauses make some progress toward preventing some kinds of repugnant reuse, I didn’t see you propose anything new to advance author controls over either misuse or misrepresentation. As was mentioned by Josh, the best protection against misrepresentation is through openness of the original work, so that subsequent readers can form their own opinions about the author intent.
I’m sorry you found it annoying that I chose to discuss a survey of researchers that was conducted by the AAAS. I discussed it because I felt the findings were interesting — and no less so because the AAAS charges APCs. I do find your logic kind of odd here: you point out how much money they make publishing on an OA basis, and conclude from that fact that “it’s in their interest to sew [I think you meant “sow”] distrust and confusion about open access.” Given how much money they make selling OA publishing services to authors, wouldn’t it actually be in their interest to promote trust and enthusiasm about OA publishing among authors?
As for whether I “(proposed) anything new to advance author controls over either misuse or misrepresentation” — of course I didn’t. As I pointed out in the piece and as has been discussed further in the comments section, there’s nothing any author can do to prevent misrepresentation of their published work, though the law does provide redress for some kinds of misrepresentation after the fact. As for misuse, the tools for preventing that already exist, in the exclusive rights provided to authors by copyright law. Of course, those tools are imperfect, and when you apply a CC BY license even they disappear — which is totally fine when the author chooses to apply the license. When those with power over authors compel them to apply the license, I find that more troubling.
it’s actually even much more confused, and even publishers are confused here. increasingly the important outputs of research are code and data. These should be under different licenses for appropriate use and reuse (usually CC0 for data and an open software license, certainly not CC-BY+…). Yet many publishers don’t distinguish these in supplements or papers or worse place them under the paper license, and nor do authors posting in general repositories. Another reason for things to go in the right, curated place…just saying. More resources needed here for everyone (and less elsewhere).
You could just as easily write an article titled “Confused and Ambivalent: Scholarly Authors and Copyright in General.” Most authors have just as little idea about the rights they have to begin with and what they surrender when they sign an agreement with a publisher.
The survey and its results are pretty disingenuous. Many of the concerns highlighted have nothing to do with whether something is openly-licensed or not but that is implied by the framing of the whole survey being about open licences. Are you worried about misuse of your work, or commercial reuse, or whether there is nothing much you can do about these things? Sure, I am — but whether its openly-licensed has little to do with this because such actions are, typically, (a) AI companies relying on fair use not licences, (b) the actual publishers to whom you have surrendered commercial rights in your work (possibly even squirrelled away in the fine print of a contract you signed where you had chosen a NC licence!) or (c) bad actors who will do such things and certainly aren’t saying “oh, this is openly-licensed — great, that means we can do misinformation with it!”
Also, it’s worth noting that Rick’s position in parts of this article are likely due to a US perspective on moral rights. Moral rights — the right to object to false claims of authorship or “derogatory treatment” of your work — are much stronger in other parts of the word, especially in Europe but also common law countries like UK, Australia, NZ, … In a US context moral rights aren’t protected explicitly through copyright law but rather through other instruments like privacy or publicity rights or defamation.
The CC licences all preserve moral rights (along with others):
“The license may not give you all of the permissions necessary for your intended use. For example, other rights such as publicity, privacy, or moral rights may limit how you use the material.”
Richard, when you object to “the framing of the survey (being) all about open licenses,” are you saying that AAAS should not have done the survey? Because the issue here isn’t “framing”; the survey is explicitly and specifically about open licenses. The questions are all designed, very clearly and openly, to assess respondents’ understanding of and feelings about open licenses. It could hardly have been more above-board; there’s absolutely nothing “disingenuous” about the survey or (even less) its results.
What does strike me as disingenuous is the assertion that whether a copyright holder has any control over reuse (especially commercial reuse) has “little to do with… whether it’s open licensed.” If that were the case, we wouldn’t have open licenses, the entire purpose of which is to eliminate existing restrictions on such reuse.
That said, you’re correct that my analysis comes from an American perspective. In some other parts of the world moral rights are written into copyright law and can’t be licensed away. (In those legal contexts, however, it’s not that CC licenses “preserve” those rights — it’s that they’re inalienable, and therefore CC licenses acknowledge their powerlessness to waive them.) But it’s also worth noting that the issues about which authors expressed the most concern in their responses to this survey were not primarily moral-rights issues.
Well, it *could* be more above board if this was independent research, reviwed and approved by an ethics committee, submitted to a peer-reviewed academic journal and reviewed and approved before publication. You might also suggest that if you were doing such research you’d be required to declare conflicts of interest, which a commercial publisher with a strong interest in controlling the economic rights associated with works would have to do.
The survey has some interesting results (not particularly *surprising* results) but is only based on 224 responses and they don’t indicate how those resposnes were obtained (i.e. what skew is on the answers based on who was asked to respond and how motivated then to do so).
The report says:
“One of the most striking findings from our survey — shared by nearly 40% of the scientists who responded — was that when misuses or misrepresentations of work do happen, nothing can be done. The reality is that for work that’s made available with a publisher license instead of an open license, publishers can and do intervene when authors have reuse concerns.”
In my view, the underlying message of the report is there in the second sentence: sign up to a publisher licence and you’ll be safer, not those confusing open things that no-one understands. The report can legitimately report the 40% bit (because that’s what respondents reported) but the second bit (“The reality is…”) is just an assertion that has nothing to do with the results of the survey.
The reality is that insitutions are more likely to succeed in addressing miuse of their work than a publisher. We can and do intervence when authors have reuse concerns. And, I might go further, with some success. …Now, I have no empirical evidence to support these assertions, which is based on my years of experience as my institution’s copyright officer. But you see how easy it is to start a sentence with “The reality is…”
Well, it *could* be more above board if this was independent research, reviwed and approved by an ethics committee, submitted to a peer-reviewed academic journal and reviewed and approved before publication.
I think you’re confusing “above board” (a term that means honest and open) with “scientific.” No one has claimed that this survey constituted a scientific, peer-reviewed study. It’s a survey, with all the strengths and weaknesses of a survey.
You might also suggest that if you were doing such research you’d be required to declare conflicts of interest, which a commercial publisher with a strong interest in controlling the economic rights associated with works would have to do.
One might wonder whether you have any conflicts of interest in publicly criticizing the AAAS survey. Of course, we at least know who AAAS are and what their business is, because they’re honest and open about that — in your case, we don’t even know who you are or what your business is. (See “above board,” above.)
In my view, the underlying message of the report is there in the second sentence: sign up to a publisher licence and you’ll be safer, not those confusing open things that no-one understands.
I can’t speak for the AAAS, but I do think critiquing “underlying messages” is a bit of a cheap shot. When you’re asserting that someone’s message is something other than what they’ve said, it gives you a lot of leeway for mind-reading and straw man argumentation. This comment strikes me as a perfect example of that strategy.
Thank you Rick for posting this interesting article. The information that you mention about the survey is interesting and so is the discussion that followed here. It contains some points that I have not thought about before. As you know, we are publishing a diamond open access journal at our non-profit association, and I am a scientist and author myself, so the copyrigh issue is an important ongoing issue of discussion for us.
Regarding the use of content in generative machine learning algorithms, a lot of webmasters are starting to use a much more restrictive robots.txt file that prevents AI engines from crawling their content. It has no legal basis, and it isn’t enforceable, but internet companies have generally abided by these restrictions for the last 30 years. Google has grudgingly agreed to use a separate user-agent for their Bard engine. See https://blog.google/technology/ai/ai-web-publisher-controls-sign-up/. Cloudflare seems to be taking the lead in tracking AI agents who crawl. See https://blog.cloudflare.com/declaring-your-aindependence-block-ai-bots-scrapers-and-crawlers-with-a-single-click/ I am not a lawyer, but a lot of the companies pursuing generative machine learning are trying to rely upon copyright fair use to justify their mining. The creative commons organization seems to have come down on the side of the tech companies. See https://creativecommons.org/2023/08/18/understanding-cc-licenses-and-generative-ai/ The software world has a much wider range of licenses than the content world, so perhaps we should expect a new generation of content licenses to address this problem. Given the difficulties that a lot of publishers are having with bogus content being submitted that was trained on their previous content, I expect this to become more important in the future.
Agree with many of the commenters here that the survey report & your commentary seem skewed to highlight negative aspects of OA, & conflates individual researcher motivations, the ‘OA movement’ and funder mandates.
I was struck how the final pie graph in the report is about the qn ‘Are there risks of making published work available for reuse under open licenses like the CC BY license?’
The graph is labellled ‘41% believe that there are risks connected with the use of CC BY licenses.’ They don’t label the other half of the graph, which could obviously be labelled ‘59% do not believe there are risks connected with the use of CC BY licences.’
They then include two quotes under the heading ‘What are the risks?’
‘Fear of misrepresentation and the inability to take any legal action to challenge misrepresentation.’
‘Concerns about misuse of data for political gain.’
Reasonable concerns but they are true of anythign not just CC BY.
Agree with many of the commenters here that the survey report & your commentary seem skewed to highlight negative aspects of OA, & conflates individual researcher motivations, the ‘OA movement’ and funder mandates.
When facts are inconvenient for advocacy positions, it can indeed be very easy to see the discussion of those facts as an attack on the advocacy position. What these survey results suggest is that for scholarly authors, there exist both confusion about how open licenses work and ambivalence about whether they’re a good idea. If these responses are inconvenient, one possible strategy is to attack the messenger(s). But I would suggest that it’s worth considering whether advocacy could actually be made more effective by taking into account and engaging with inconvenient facts.
The graph is labellled ‘41% believe that there are risks connected with the use of CC BY licenses.’ They don’t label the other half of the graph, which could obviously be labelled ‘59% do not believe there are risks connected with the use of CC BY licences.’
Yes, it is indeed self-evident that this graph is telling us that 59% of the survey’s respondents do not believe there are risks associated with CC BY licenses. It’s also self-evident, from other graphs, that for 58% of respondents funder policy does not strongly influence their choice of license, that 48% of respondents do not rate themselves as “knowledgeable about CC BY licenses,” etc. I’ve seen a lot of survey reports over the course of my career, and there’s nothing unusual about explicitly labeling some but not all segments of a chart or graph.
