There has been a lot of attention given to the moves by publishing giant Elsevier to enforce its policy regarding authors’ rights to post his or her article on a personal website or in an institutional repository.  Since Elsevier began sending take down notices last fall, first to and then to individual universities, it has become well-known that authors are allowed, under their contracts with Elsevier, to post their “final submitted version,” which refers to the last version of the article, after peer-review, that is sent to the journal, but not the final published version with whatever copy-editing and formatting that the publisher has added.

This is a difficult distinction for faculty authors to understand.  My colleagues and I talk about it all the time with our faculty authors, but they persistently do not see much difference between the two versions, so they sometimes believe that there is little reason to observe the distinction.  Publishers think (or at least say publicly) that they add a lot of value to submitted manuscripts, but a great many authors do not see it that way.

Unfortunately, some of the attention that this new strategy from Elsevier has garnered has made the difficulty of explaining what is going on to faculty authors a little more difficult.  This article from The Economist Called “No Peeking…” is a case in point.  The article correctly suggests that this is going to prove a self-defeating tactic for Elsevier, whose desperation to stem the movement toward open access often leads it into foolish decisions.  The Economist, however,  misstates the copyright law in its article in a way that will unfortunately reenforce a common misconception on campuses.

Here are three sentences from The Economist article that embody the misconception I am concerned about:

Like journalists writing for a newspaper, academics submitting an article to a journal usually sign contracts which transfer copyright to the publisher…. As the University of California, Irvine, which was on the receiving end of some of the takedown notices, points out in advice to its staff, it is usually only the final version of an article, as it appears in a journal, that is covered by publisher’s copyright. There is nothing to stop scientists making earlier versions available.

The problem with the first sentence is that academic authors are really not like journalists.  Many journalists are full-time employees of their newspapers, so that their articles are owned by the newspaper from the start, as works made for hire.  On the other hand, academic authors are not employees of publishers and their writings are not work for hire.  Their rights (as well as those of some free-lance journalist) are entirely governed by the contracts they sign.  The important implication of this is that academic authors have much more control over the rights they surrender and retain than do journalists; faculty members can simply refuse to transfer copyright (because they own it unless and until it is transferred in writing) or they can negotiate the exact terms of publication, transferring or licensing some rights and holding on to others.

The bigger issue in this article, however, is in the second and third sentences quoted above, about how the copyright that is transferred to publishers only “covers” the final version of the article.  This is a common misconception that is both wrong and dangerous.  It is the same misconception that leads some people to believe that if they re-draw an illustration, chart or table from a copyrighted publication, they do not implicate copyright.  But the truth is that a copyright includes any work that is derived from the copyrighted work and is “substantially similar.”  When someone wants to use a figure from a published work, they may well be able to do so under fair use, but redrawing the figure, unless it is redrawn into something quite different (which would undermine the purpose), does not alter the copyright situation.

When we turn to the issue of article versions, the situation is the same.  Each version is a revision of the original, and the copyright is the same for all these derivatives.  When copyright is transferred to a publisher, the rights in the entire set of versions, as derivatives of one another, are included in the transfer.  Authors are not allowed to use their post-prints because the rights in that version are not covered in the transfer; they are allowed to use post-prints only because the right to do so, in specified situations, is licensed back to them as part of the publication agreement.

Once a copyright transfer has been signed, all of the rights that the author may still have are because of specific contractual terms, which are usually contained in the transfer document itself.  In short, these agreements usually give all of the rights under copyright to the publisher and then license back very small, carefully defined slivers of those rights back to the author.  One of those slivers is often, but not always, the right to use a submitted version, or post-print, in carefully limited ways.  For example, many publishers allow posting of the submitted version only on defined websites, usually a personal site or institutional repository.  Often the contracts also allow posting of the submitted version only after some lapse of time.  These restrictions would not make sense or be enforceable IF the author retained some kind of copyright in earlier versions, as The Economist implies.  But they do not; they have only, and exactly, what the contract gives back to them.

One important lesson to be gained from this correction of the language of The Economist article is that publication contracts are extremely important.  They entirely determine what an author can do with his or her own work in the future.  For many academics, signing such agreements is a very bad idea; they should be negotiated, either to make them licenses to publish, which allows the author to retain her copyright, or to be certain that the rights that are licensed back are broad enough and flexible enough to permit the future uses the author wants.  Before the transfer, the author has a good deal of leverage to negotiate these agreements, but afterward she has very little.  So it is vital to pay attention to the agreement itself and not rely on a false sense of security based on a misconception of how copyright works.

Another point to learn from this situation is that the whole idea of article “versions” is artificial.  It has been developed primarily by publishers in order to make a claim that they add substantial value to the final published version, which may or may not be true, depending on the article and the publisher.  Another marketing advantage that publishers get from this fabricated distinction is the ability to claim that they support author rights and reuse of articles to promote better access, while still retaining the ability to slap down authors who use their own articles in ways the publishers have not pre-approved.  As my colleague Will Cross has put it, “This versioning is a creation of publishers to reenforce the sense that they are following the academic “gentleman’s agreement” that Elsevier has been breaking here.”

[Hat tip to Will Cross and to Lisa Macklin of Emory, who discussed the implications of this particular mistake with me by e-mail and provided some ideas incorporated herein. Will qualified his statement quoted above by acknowledging that pre-prints, especially, have a longer history, but the use of these distinctions as contractual dividing lines is related to recent pressures on publishing.]


12 Responses to Setting the record straight about Elsevier

  1. Jo Koster says:

    An extremely valuable article, and as usual, clear and well-presented for laypeople like me. I’m sharing this with my colleagues–many, many thanks.

  2. Philip says:

    I have been wondering about licensing for green OA so thanks for the clarification. I have been putting CC-BY licenses on the manuscripts I contribute to our repository, which is clearly not correct if I transferred copyright to the journal.

  3. Jenny says:

    Wouldn’t it be great if both the article and the agreement were submitted into institutional repositories together, so that future repository managers would know what rights existed where for each article? This would perhaps also help to support authors to keep track of these things.

    As things are, even repositories which check rights on each article can only apply the latest policy of the publisher, if they don’t know what was actually signed: articles from a few years ago may have had different conditions but what can you do without a copy of the agreement?

    Also, some publisher policies even refer to versions that never existed in the writing process, since many journal articles are developed inside publishers’ software and a plain draft has to be fabricated after the fact, for repository deposit.

  4. Mike Taylor says:

    Thanks for this important clarification. This is why I recommend licencing a work under CC By, or placing it in the public domain, before signing it over to the publisher.

  5. Charles Oppenheim says:

    “When we turn to the issue of article versions, the situation is the same. Each version is a revision of the original, and the copyright is the same for all these derivatives. When copyright is transferred to a publisher, the rights in the entire set of versions, as derivatives of one another, are included in the transfer.”

    That may be true in US law (I really do not know), but is not true in UK law. The author retains the rights to earlier versions of the article; only the rights in the final version get transferred to the publisher if the author (foolishly) signs the copyright assignment..

  6. Charles Oppenheim says:

    All a copyright assignment does is give the publisher the rights to the final version, and to stop anyone copying, amending, etc., that final version without permission. I could only assign copyright in earlier versions if the publisher asked me to in writing. But an author would be foolish to agree to this. For example, what if my article is based on an earlier report to my funding agency. How can the publisher ask me to assign copyright in that? Common sense, as well as the law, says if I assign copyright in the final version of an article, that is all I pass to the publisher. Reproducing an earlier draft on an OA platform, the so-called “Oppenheim-Harnad” solution, might be considered unethical by many, but it is not illegal.

    • Kevin Smith, J.D. says:


      I would really like to see the legal basis for this argument. I find it difficult to accept that UK law is so completely different from the law in the U.S. Most authors do sign a written copyright transfer, and those agreements cover all works that are derivative of each other. Only when we realize that do policies like those of Elsevier and Wiley, which proscribe specific permitted uses of all of the different versions, make sense. They are licenses back to the author after copyright in the whole work has been assigned. I agree it is foolish, but I believe it is still the norm.

  7. Julie Evener says:

    I’ve found that students are very reluctant to use “author’s manuscript” copies (so called in PubMed Central) for their schoolwork. They are afraid it is not a valid version of the article and that their instructors will reject it. No matter how I explain that the author’s manuscript copy is post peer-review, they usually hold out for the final, published version.

  8. Charles Oppenheim says:

    Kevin, I explain my thinking at I have seen many publisher copyright assignment forms, and not one has ever mentioned “derivative works”, but even if one did, that would apply to adaptions made later to the final manuscript, not to precursors of the final manuscript. The usual wording is simply (see Clark’s Publishing Agreements, 9th edition (2014), Precedent 8B, “the Author as beneficial owner hereby assigns to publisher the copyright in the Article” – no mention of precursors there! But what if a publisher did demand rights to precursors of the final accepted article? To take a typical situation, a doctoral student completes their thesis, and the supervisor decides to work it up into a joint article. They adapt the thesis in various ways to make it appropriate for the journal in question. The thesis is a clear precursor of the final article. So such an assignment would also assign copyright in the thesis to the publisher. Clearly that can’t be right. The same logic applies to the draft of the article before the manuscript that is finally accepted.

    As my piece explains, the fact that the publisher does not own rights to the precursor is pretty irrelevant because taking advantage of that fact – the so-called Harnad-Oppenheim solution – is almost never used, for good reasons. But that doesn’t mean I have no concerns about Kevin’s remarks. Why should we give the impression that publishers have greater rights than they actually have? It is important to make clear just what publishers own, and do not own following an assignment.

    I would be most interested if Kevin can supply wording in a publisher assignment that refers to precursors (as opposed to later derivatives) of the accepted manuscript.

  9. The writer for the economist clearly understands the law; if he were under the misconception discussed here, he would have said “always” in place of “usually”. In my field (Mathematics), the author “usually” retains even the right to post the publisher’s final version on an institutional website.

  10. […] Smith, J.D., over at Scholarly Communications @ Duke is “Setting the record straight about Elsevier” with regards to Elsevier’s “policy regarding authors’ rights to post his […]