How to acknowledge copyright for figure reproduced from arXiv preprint?

Question

I am coauthor of a review article, in which we reuse a figure from another article that is, so far, available only on the arXiv. In such a case, one would usually obtain the copyright from the publisher and acknowledge the copyright holder with a statement similar to

Reproduced with permission [reference]. Copyright [year], [publisher].

What is an appropriate analogous phrase for an article published on the arXiv?

From what I understand of arXiv's help on licenses, the default option only grants arXiv the right to distribute the article, while the copyright remains with the authors. However, including all authors of an article in the copyright statement can make the statement very long. Perhaps one should restrict it to the author that submitted the article to the arXiv?

Reproduced with permission [reference]. Copyright [year], [submitting author].

P.S. In this particular case, the arXiv article was authored within our group as well, i.e. we are free to choose the copyright statement. In general I assume it would be up to the authors to choose how they would like to be acknowledged.

Answer 1

As Dan mentioned, you have both an ethical, academic obligation to cite your source, and a legal duty to comply with whatever licence the image was made available under.

However, in general, the possibilities of that licence are somewhat broader than the CC-BY licence explained in Dan's answer.

In general, a paper found on the arXiv can be licensed in five different ways, which are explained in the arXiv License Information page. Authors can do the following:

• grant arXiv.org a non-exclusive and irrevocable license to distribute the article, and certify that he/she has the right to grant this license;
• certify that the work is available under one of the following Creative Commons licenses and that he/she has the right to assign this license:
  • Creative Commons Attribution license (CC BY 4.0)
  • Creative Commons Attribution-ShareAlike license (CC BY-SA 4.0)
  • Creative Commons Attribution-Noncommercial-ShareAlike license (CC BY-NC-SA 4.0);
• or dedicate the work to the public domain by associating the Creative Commons Public Domain Dedication (CC0 1.0) with the submission.

In addition to this is the question of whether the authors have retained copyright for the material after the arXiv posting. If the paper has since been published, chances are that they have ceded copyright to the publisher and you will need to go through the proper channels to get it. Since the question is specifically about the preprint, I will assume that the copyright of the material remains with the author. (Nevertheless, you still need to check that this is the case.)

The easiest case is if the arXiv posting has been put in the public domain (CC0 licence), in which case you do not need to acknowledge any copyright, much as if you were reproducing, say, a passage of Alice in Wonderland. (You still need to attribute it, though, the same as if you were reproducing Alice in Wonderland.)

If the paper has been shared via any of the Creative Commons licences, then you can simply reproduce it, with the proper attribution and a link to the relevant licence. If the licence is CC BY-SA or CC BY-NC-SA, then you also need to make sure that you abide by the terms of the licence: that you're publishing under a compatible licence, in both cases, and that your use is non-commercial in the latter.

The most common case, however, in many fields, is the first one: the authors granted arXiv.org (not you) a license to distribute the article, and that's that. This means that, barring any fair-use arguments you wish to make, you do not have permission to redistribute the work either in part or in its entirety. In this case, what you need to do is contact the authors and ask for permission to republish the image.

In your specific case, however, your group is itself the owner of the copyright, and you do not need to do anything. It is not even a question of you having the freedom of choosing how the initial work is licensed: you're simply publishing the same material in different venues, which is perfectly fine. Of course, if you want to go further and license that material under a CC licence, then all the better! (Just be aware, of course, that doing so could impair your ability to publish it in some venues down the line.)

Answer 2

The crucial points to remember here are that you have an

• ethical duty to cite your source, and a

• legal duty to comply with the terms of the license under which the article was uploaded to arXiv.

Those are your obligations, nothing less or (in my opinion) more. Considering this, after looking at the arXiv license information you linked to, it seems to me that the following language would be appropriate for at least one of the license options (I didn't look at all the options, the answer may vary slightly for the others):

Figure 1. This is a very interesting figure. (Image source: [ABC15]; use permitted under the Creative Commons Attribution License CC BY 4.0.)

where "Creative Commons Attribution License CC BY 4.0" is hyperlinked to this page, and where [ABC15] refers to the bibliographic entry for the article in the references section of your paper.

Note that (as far as I'm aware) there is no need in this case to refer specifically to copyright. I also disagree with your assertion that "In general I assume it would be up to the authors to choose how they would like to be acknowledged." As I said, you have a duty to acknowledge the source of the work and its authors, and satisfy whatever legal requirements may exist pertaining to copyright or other licensing restrictions, but no other obligations that I'm aware of to use specific language that the authors or anyone else wants you to use. See here for a somewhat related discussion.