My last post about copyright assignment and different versions of a scholarly article set off a small controversy, some of which can be found in the comments to that ppost and some of which took place on other social media venues. Yesterday Richard Poynder posted to the Lib-License list about this discussion, and I felt compelled to respond, since it seems clear this is not an isolated misunderstanding that will fade away.
Here is part of Richard’s post, which summarizes the discussion:
Last week, the Scholarly Communications Officer at Duke University in the US, Kevin Smith, published a blog post challenging a widely held assumption amongst OA advocates that when scholars transfer copyright in their papers they transfer only the final version of the article.
This is not true, Smith argued.
If correct, this would seem to have important implications for Green OA, not least because it would mean that publishers have greater control over self-archiving than OA advocates assume.
However Charles Oppenheim, a UK-based copyright specialist, believes that OA advocates are correct in thinking that when an author signs a copyright assignment only the rights in the final version of the paper are transferred, and so authors retain the rights to all earlier versions of their work, certainly under UK and EU law. As such, they are free to post earlier versions of their papers on the Web.
And here is the response that I just sent to the LibLicense list, in which I focus on copyright as protection over expressive content, rather than arbitrary distinctions between different versions of that content:
I had really hoped I could ignore this rather muddled controversy, mostly due to a lack of time to address it. But a tweet from Nancy Sims, of the University of Minnesota, made me realize that my original post used slightly careless language that may contribute to the confusion. So I feel I should set that straight, and respond to the whole business.
I wrote that different versions of an article were derivatives of one another. That is probably a defensible position, but Nancy made the point much clearer — the different versions are still the same work, so subject to a single copyright.
Throughout this discussion, the proponents of the position that copyright is transferred only in a final version really do not make any legal arguments as such, just an assertion of what they wish were the situation (I wish it were too). But here is a legal point — the U.S. copyright law makes the difficulty with this position pretty clearly in section 202 when it states the obvious principle that copyright is distinct from any particular material object that embodies the copyrighted work. So it is simply not true to say that version A has a copyright and version B has a different copyright. The copyright is in the expressive content, not in different versions; if all embody substantially the same expression, they are all the one work, for copyright purposes, because the copyright protects that expressive content. Hence Nancy’s perfectly correct remark that the different versions are the same work, from a copyright perspective.
Part of the point I wanted to make in my original post is that this notion of versions is, at least in part, an artificial construction that publishers use to assert control while also giving the appearance of generosity in their licensing back to authors of very limited rights to use earlier versions. The versions are artificially based on steps in the publication permission process (before submission, peer-review, submission, publication), not on anything intrinsic to the work itself that would justify a change in copyright status. If we look at how articles are really composed — usually by composing one file and then editing it repeatedly, it is easy to see how artificial, in the sense of unrelated to content, the distinctions are. How much time must elapse before a revision is a different version? If I do some revisions, then go have a cup of tea before returning to make other revisions, have I created two different “versions” entitled to separate copyright protection? The question is absurd, of course, and shows how unworkable the idea of different copyrights in different versions of the same work would be.
It has been said that no publisher makes the claim I am here suggesting. But if we look at actual copyright transfer agreements it is easy to see that they do. The default policies for Wiley, for example tell authors that they can archive a pre-print and archive a post-print, subject to certain conditions, including rules about the types of repositories that the archiving can take place in and a limitation to non-commercial reuse. If an author transfers rights only in the final version, how can Wiley make restrictions on the use of these earlier versions? The better — indeed the only logical — interpretation is that the copyright that is transferred covers the work as a whole, which is the nature of copyright, and that Wiley then licenses back to authors certain rights to reuse different versions. Those version rights are based on what Wiley wants to allow and to hold on to, not on any legal distinction between the versions. Elsevier’s policies are similar — they allow the preprint to be used on any website, the post-print to be self-archived on a scholarly website ONLY if the institution does not have a mandate and with acknowledgement of the publisher, and do not allow any archiving of the final version. Again, all of this is grounded on a claim that a copyright that is inclusive of the different versions, because they are the same work, has been transferred to Elsevier.
Let’s imagine what would happen if a dispute ever arose over a use of an earlier version of an article after the copyright had been transferred. A court would be asked to determine if the use of the earlier version was an infringement of the copyright held by the assignee. Courts have a standard for making this determination; it is “substantial similarity.” So if the re-used version of the work was substantially similar to the work in which the copyright was assigned — that language is itself bound up in the misunderstanding I am trying to refute — a court would probably find infringement. This has been that case in situations where the works were much more different that two versions of a scholarly article. George Harrison, for example, was found to have infringed the copyright in the song “He’s So Fine” when he wrote “My Sweet Lord,” even though the court acknowledged that it was probably a case of unconscious borrowing (see Bright Tunes Music v. Harrisongs Music, 420 F. Supp. 177, S.D.N.Y. 1976). And the author of a sequel novel to “Catcher in the Rye” was held to have infringed copyright in Salinger’s novel even though they told very different stories, due to similarities in characters and incidents (Salinger v. Colting, 607 F. 3d 68, 2d Cir. 2010). If these very different “versions” of the same work were held to be copyright infringement, how is it possible that two versions of the same scholarly article could have separate and distinct copyrights?
In many ways I wish it were true that each version had a distinct copyright, so that transfer of the rights in one version did not impact reuse of the earlier version. That situation would make academic reuse much easier, and it would conform to a basic sense that most academics have that they still “own” something, even after they assign the copyright. But that position is contrary to the very foundations of copyright law (and not just U.S. law), which vests rights in the content of expression, not in versions that represent artificial points in the process of composition or publication. And much as this mistaken idea may be attractive, it has dangerous consequences; it gives authors a false sense that the consequences of signing a copyright transfer agreement are less draconian than they really are. Instead of plying our faculty with these comforting illusions, we need to help them understand that copyright is a valuable asset that should not be given away without very careful thought, precisely because, once it is given away, all reuse of the expression in the article, regardless of version, is entirely governed by whatever rights, if any, are licensed back to the author in the transfer agreement.