Anytime that academic authors sue each other over a journal article, it is worth attention in this space. A couple of weeks ago, the U.S. District Court in Massachusetts ruled in such a case, and the ruling raises some interesting points to consider for those of us involved in working with scholarly publications.
Note first that this is a very fact-specific case and a decision by a district court, so it is not a control precedent for anyone other than the parties. A decision would have more weight if this ruling were appealed, but, because the motive behind the lawsuit seems to have been more wrath than reason, I think that is unlikely.
The facts of the case are pretty common, even though the end result, a lawsuit, is not. A postdoc named Mallon was hired by a scientist at Brown University and together these two, with a third Brown faculty member, wrote at article in 2011 that was submitted to, and rejected by, the journal Neuron. The article was later sent to PLoS Biology, which required extensive revisions as a condition of publication. During this drawn out process, the relationship between Mallon and the faculty authors “deteriorated,” according to the court. Also, the article was revised extensively, with new experiments, new data and new text. Dr. Mallon disagreed with the inclusion of some of this new material and told the court that he did not have confidence in the integrity of the later versions of the manuscript. Mallon and the faculty authors had no communication for 15 months prior to the ultimate publication of the heavily revised article. The article was finally published in early 2013; Mallon’s contribution to the early draft was mentioned in the acknowledgements, but he was not listed as an author on the published version.
Mallon sued seeking a declaration that he was a joint author, and also asking the court to order that the article, which he claims in its published version is a product of scientific misconduct, be retracted by the journal. The court correctly notes that these two demands are irreconcilable. In order to meet the standard for joint authorship, which is to have created original expression with the intention that it be merged into a unitary or inseparable whole, Dr. Mallon undermines his own claim that the article in its final form should never have been published. Clearly his intentions in regard to the article changed over time. Ultimately, the court ruled that Mallon was not a joint author of the final version, because of this lack of intent, and also, I am glad to say, that court-order retraction would not be appropriate.
For me, the most interesting issue in this case is a controversy that the court did not resolved, because it did not have to. What is the legal relationship, under copyright, between that first draft, on which Mallon was listed as first author, and the final, published article, on which, according to the court, he was not an author at all? I have written before that the copyright in an academic article extends to all versions of that article; this is why publishers who receive a copyright transfer are able to dictate, as they usually do, how the authors of the article are allowed to use earlier versions – so-called pre-prints and post-prints. This case raises a situation that tests the limits of that assertion, where the first version and the published version may differ so much that they would be considered separate works (or, perhaps, the later article could be a derivative work from the first draft), where distinct copyrights would exist. This certainly seems to be implied in a situation where someone is an author of one version and not an author of the other.
The possibility of distinct copyrights really turns on the degree of dissimilarity between the first draft and the final article. The court notes how different the final publication is, with all the new additions and revisions, from the first draft. There is, of course, a point at which a revision becomes a new work. That point will be different in each situation, and in many case — perhaps most cases — there will be no such point. In other words, most final articles are not substantially dissimilar from the first draft; not different enough to be a separate and distinct work. This case exists at the edges, where the differences are much more substantial, apparently, then is usual.
As I have said, the court does not resolve the issue of where, or if, a separate copyright exists. It notes the concern expressed in Garcia v. Google that if every iteration of a work had a separate copyright, this would “make Swiss cheese of copyrights,” but notes that in Mallon v. Marshall, the case it is deciding, there is a logical division between the paper Mallon intended to contribute to and the one he disavowed so strongly. This is enough to decide the issue of joint authorship, so the question of distinct copyrights does not arise. The court specifically says that it will not rule on whether or not Mallon holds a copyright in the first draft (the defendants suggested that it was work made for hire) or whether the final version is a derivative work based on that first draft. These questions would help us define the boundaries of copyright in a work that goes through multiple versions. From this case, however, all we can conclude is that there are extreme cases where an early draft might be so different from the final published article that the copyright transferred to the publisher would not encompass some earlier iterations. Even in those rare cases, however, scholars would need to consider the value of what they have retained, since control over an early draft that is so different from the final article might have very little scholarly utility, precisely because those authors decided to make radical changes.