We’ve all monkeyed around trying to sort out the ownership of published content. In the scientific community, copyright and its (mis)application in publishing has authors, publishers, and readers grappling with questions of what is legally possible, what is desirable, and what is “allowable” by any particular party.
The most recent example of these challenges can be found in a PLOS ONE article published yesterday by Stirling et al., which focuses on a re-analysis of images used by other research groups as evidence for the creation of “striped nanoparticles”. The scientific controversy is fascinating and has been covered on blogs and in social media, but the copyright issues that cropped up during the paper’s publication process are also noteworthy.
In the study, the authors re-analyzed key results from previous work written up in the literature by other researchers and wanted to share their findings by publishing them, formally adding to the scientific literature on nanoparticles. To provide context and more effectively discuss the data, the authors of the re-analysis included figures (images) from previous studies in their own paper. However, the original figures were published in journals that owned the copyright of all the written content. And here is where we ran into a problem, and one that was far from simple.
Clearing rights for republication
These journals generally hold a copyright status of “all rights reserved”. With this restriction, there are three possible ways that these figures could be incorporated into a newly published article.
1) The authors could declare that the reproduction is “fair use” or “fair dealing”:
Traditionally, this hasn’t been done for reproduction of single images from papers. The problem from a publisher’s (and author’s) perspective with this approach is that it isn’t totally clear whether it applies in this case. Publishers are pretty risk averse about this kind of claim (as fair use is a defense against infringement rather than a license itself), so we usually prefer to ask permission.
2) The copyright holder could authorize a specific use under a limited license:
This is the traditional approach—the author would be responsible for negotiating a single use license from the rights holder. This will be a familiar experience to many authors who have written reviews that incorporate figures from other papers. It’s common enough that there are online tools to speed up this kind of transaction.
3) The rights holder could authorize the image to be republished under an open license that allows extensive reuse, with the authors as the copyright holder:
PLOS uses a Creative Commons Attribution License (CC BY 4.0) specifically to cover this issue. With the Stirling, et al. paper, we worked with the authors and the publishers on taking this approach, and we want to say thank you to The Royal Society of Chemistry, Nature Publishing Group (NPG), Wiley Blackwell, and the American Chemical Society for working with us to enable the republication of the material. NPG has a standing policy that they will authorize re-publication of images under the same license as the rest of the article, something they’ve done for us in the past. The other publishers worked hard to make it possible on this occasion, giving approval a few weeks ago. We are very grateful for their efforts.
Why the fuss?
So why did PLOS push for option three when the second was more traditional? As Kevin Smith at Duke has pointed out, it’s not a legal question or a problem with the Creative Commons licenses. It is perfectly possible to mark part of an article as being made available under a different copyright license. But as a question of policy, we make every effort to avoid mixing different licenses within articles.
We do this because, at the heart of PLOS’s work on Open Access, is enabling re-use. Every day people grab newly published articles for various kinds of processing. Some are mining text for information, some are collecting images for use in other projects, and some are working on new forms of display and article presentation. Each of these depends on the downstream user having the rights to copy, store, and transmit the articles. All the articles. And all the pieces of the articles.
On any given day, PLOS might publish a few hundred articles. It is certainly possible in principle to add a copyright statement to a single image. But downstream users would need to check every figure legend of every paper by hand to identify single images they aren’t allowed to use. By insisting on a uniform CC BY license throughout, we give the downstream users clarity and confidence in their rights.
In a perfect world, there would be a clear and consistent machine-readable way of designating sections of an article and differing copyrights within them. But such a standard doesn’t yet exist, and even if it did, there is still a clear advantage to the downstream user to have a uniform policy that grants largely unrestricted reuse.
This case has been both interesting and challenging for us and the authors (who we thank for their patience while we have addressed the licensing issues!). It has highlighted the need for publishers to be willing to authorize onward publication rights with reasonable flexibility as to the licensing arrangements. As the industry moves away from copyright assignment to “license to publish” arrangements, it is important that we get the details of those licenses right. There could be immense value in shared model “license to publish” agreements that would help consistency and support smaller publishers. Most importantly, we have learned the need for clear presentation of our legal and policy obligations when it comes to copyright and reuse.