An increasing number of journals are moving beyond copyright to license their material more openly. One very popular set of licenses was developed by Creative Commons, a non-profit organization focused on making creative works available for discovery and reuse. Initially developed by Harvard Law professor Lawrence Lessig, Creative Commons licenses provide an alternative to standard copyrights, allowing authors to specify ways that their works can be used without having to grant permission for each individual request.
The Creative Commons system is very easy to comprehend on a basic level, but there are many questions about the individual licenses. I first heard about these licenses in a graduate school course in 2008, and I am still appreciating new nuances. Here is some information about what these licenses mean, along with some advantages, disadvantages, and issues involved:
The least restrictive Creative Commons license, the Attribution or CC-BY, allows any user to "distribute, remix, tweak, and build upon your work," provided that they credit the original authors in all cases. This license would allow not just for downloading and copying something, but for textmining and other automated processes.
The Attribution-ShareAlike, or CC-BY-SA, license builds upon the CC-BY by requiring that the user license any new products based on the original under identical terms (in addition to crediting the original author).
The Attribution-NoDerivs, or CC-BY-ND, license likewise requires proper credit for the original authors but also that the material be passed along in its entirety without any alteration.
The Attribution-NonCommercial, or CC-BY-NC, license allows for others to remix or otherwise alter the original material (with proper attribution), provided that they are not using it for any commercial purpose. There is no restriction on how the new material is licensed.
The Attribution-NonCommercial-ShareAlike, or CC-BY-NC-SA, license combines the non-commercial restriction with the requirement to share new material under the same conditions, all with due credit.
Finally, the Attribution-NonCommercial-NoDerivs, or CC-BY-NC-ND, license only permits users to download and share the original work (provided they credit the original source), without any alterations or commercial use. This license is the most restrictive of Creative Commons' offerings.
Authors wishing to place works completely into the public domain can do so with the CC0 mark. In such a case, all rights are surrendered, and the image can be used in any legal way.
But which license is best for scholarly publications? The answer is perhaps clouded by the range of options available. The most common license for open access publishers is the least restrictive, the CC-BY, and articles published by these licenses are more numerous each year, although the data are far from perfect. Other choices may seem appropriate for scholarly articles, but there are some issues. For example, the details of what constitute non-commercial use are fuzzy (e.g., do non-profit organizations still count as commercial enterprises?). The CC-BY-NC license could also keep images from one source from ever being used in another publication, textbook, or even class. Moreover, derivative works are an important part of scholarship -- how many times have you seen "adapted from ref. (x)"? It seems unwise to restrict use of publications in that manner.
However, when asked for their preference, authors submitting to Scientific Reports, Nature Publishing Group's open access megajournal, chose the more restrictive licenses most of the time. Many open access advocates believe that this choice stems more from a lack of information about Creative Commons licenses than from a strong desire to close off how one's work is used. After all, when citations are the primary goal, giving readers a wide array of options for reuse seems to be the best bet. As pointed out in a comment by Richard Van Noorden and in the counterpoint by Sanford Thatcher, perhaps the most important step is to make articles free to read without getting hung up on the details of the license chosen. That choice may lead journals to settle for quasi-openness, though. Making research available is good, but making research truly open for reuse and mining would be better, especially in this era of tremendous data production.
Please feel free to write to us with any questions or comments.