Which license should be chosen in arXiv for reports and preprints to be published in Elsevier?

Question

First, imagine the following scenario: A "report" is first archived in arXiv as version one (v1). From that report, a "preprint" manuscript (with the same title as the report, but probably shorter than the report, in order to comply with the number of pages required by the journal) is prepared, archived in arXiv as v2 and later sent to Elsevier. After acceptance, the "postprint" (version improved with the comments suggested by the reviewers) is archived in arXiv as v3.

Here we consider only the case of green open-access (the copyrights are transfered from the authors to Elsevier), and not the case of gold open-access (the copyright still belongs to the authors), since in the later case it is clear that any license that the authors wish can be applied to the report, preprint and postprint.

Said that, the more detailed questions are the following:

1. Is the "preprint" considered a derivative work of the "report"? Is the "postprint" considered a derivative work of the "preprint"?

2. Do the authors of the "report" and "preprint" still hold copyright on each of them (despite the copyrights of the postprint have been transfered to Elsevier)? See a sample of the publishing agreement and Elsevier's sharing policy FAQ (where the issue is not completely clear, at least to me)

Are you asserting copyright over preprints?

No, preprints can be used anytime anywhere by authors. We encourage authors seeking to publish in Cell Press, The Lancet, and some society owned titles to check the author section on the journal homepage for additional information.

3. In case of yes, does it mean that ANY creative commons (CC) license (CC:BY/4.0, CC:BY-SA/4.0 or CC:BY-NC-SA/4.0) can be applied to the "report" and the "preprint"? See Elsevier's sharing policy (the issue is again not clear to me)

Preprint

• Authors can share their preprint anywhere at any time.
• If accepted for publication, we encourage authors to link from the preprint to their formal publication via its Digital Object Identifier (DOI). Millions of researchers have access to the formal publications on ScienceDirect, and so links will help your users to find, access, cite, and use the best available version.
• Authors can update their preprints on arXiv or RePEc with their accepted manuscript .

4. In case of no, what open-access license could have been applied to both the "report" and the "preprint"?

It is clear from a previous post that the "postprint" should have a CC:BY-NC-ND license in the .pdf file and uploaded to arXiv under "arXiv's non-exclusive license" since both are compatible and satisfy the requirements of both Elsevier and arXiv.

Also SHERPA/RoMEO doesn't mention what kind of license can be applied to the "preprint".

Thank you for your help

Answer 1

1. Is the "preprint" considered a derivative work of the "report"? Is the "postprint" considered a derivative work of the "preprint"?

This probably needs a lawyer to answer definitively. Needless to say, I am not one and cannot give an authoritative answer to this question.

My experience suggests that academics and publishers often treat the preprint and "postprint" - or, in general, different versions of the manuscript - as effectively the same work for copyright purposes (e.g. copyright transfer agreements make a point of addressing what the authors are allowed to do with preprints), even though, strictly speaking, I don't believe this is supported by US copyright law.

If one is to consider the preprint and postprint and other versions as separate works, it seems much more plausible that the postprint would be considered a derivative work of the preprint, rather than the other way around. The argument that a derivative work could be created before the work it is derived from is questionable, although I suppose some odd things of this sort are legally valid, so I can't say for sure.

2. Do the authors of the "report" and "preprint" still hold copyright on each of them (despite the copyrights of the postprint have been transfered to Elsevier)?

I suppose this depends on the legal interpretation of the copyright transfer agreement.

The passage you quote, which answers the question "Are you asserting copyright over preprints?" with "No....", suggests that, as Elsevier sees it, the authors retain ownership of copyright in the report and preprint.

3. In case of yes, does it means that ANY creative commons (CC) license (CC:BY/4.0, CC:BY-SA/4.0 or CC:BY-NC-SA/4.0) can be applied to the "report" and the "preprint"?

Well, at the time the report and preprint are uploaded to arXiv, the authors are indisputably the copyright owners and there are no other legal conditions that might interfere with applying a CC license to the preprint. So of course the authors can choose to apply a CC license. If they do, in principle, Elsevier could refuse to accept the manuscript for publication; or, if they do accept the manuscript and later find out that a preprint had been uploaded to arXiv, they could argue that the author executed a copyright transfer agreement under false pretenses and the agreement should be voided, or something of that nature. I can't say whether they would actually do this in practice.

Elsevier's sharing policy says

Authors can share their preprint anywhere at any time.

That means that authors can upload their preprints to a server, and one would hope it implies authors can also grant the server operator the rights to further share the preprint with readers. It's not clear whether Elsevier's policy allows the author to grant others the right to prepare derivative works, which would be (legally) required to upload it to a site like arXiv which distributes preprints under a Creative Commons license.