When I first became aware of the lawsuit filed by publishing giant Thomson Reuters against George Mason University to stop the release of the open source citation management program Zotero (hat tip to my colleague Paolo Mangiafico for directing me to this story), I wasn’t sure how it was relevant to issues of copyright and scholarly communications. After all, this is essentially a licensing dispute; Thomson alleges that, in order to develop the newest version of Zotero, software developers at GMU “reverse-engineered” the proprietary .ens file format used by Thomson product Endnote in violation of a licensing agreement. Endnote, of course, is a very popular tool in academia, and it is alleged that GMU is marketing its new version of Zotero with the specific boast that it now allows users to convert EndNote files into its own open source and freely-sharable file format
I cannot comment on the merits of the breach of contract claim, and I have no argument with the right of Thomson Reuters to use a licensing agreement to protect its intellectual property. Nevertheless, the idea of protecting these files, which simply organize data about books, journal articles and web sites into a form that can then be mapped into different citation styles, raises interesting questions about the scope of copyright law and where new and troubling developments might take it.
At least since the Supreme Court decided Feist v. Rural Telephone in 1991, we have known that facts and data are not themselves protected by copyright, and that collections of facts must meet a minimum standard of originality (greater than that found in the phone books that were at issue) in order to be protectable. I do not know if the file format EndNote has created to store citation data is such an original arrangement of data and, apparently, neither do they. Rather than rely on copyright law, they wrote a license agreement to try to prevent what they allege took place at GMU. But two questions still bother me.
First, should universities agree to licenses that prevent reverse engineering? In today’s high-tech environment, reverse engineering is a fundamental way in which innovation proceeds. Our copyright law, in fact, recognizes the importance of such activities, providing specific exceptions to certain prohibitions in the law for cases of reverse engineering that have potential social benefits, such as encryption research or making materials available to handicapped persons. So one could legitimately ask if a court should consider the benefits of the research being done when deciding whether and how strictly to enforce a contractual provision against reverse engineering. In general, open source software is a gift that many universities like George Mason give to the academic community as a whole, and the value of that gift is increased if it is possible for scholars who have been using a costly commercial product to move their research resources from the latter into the former. That increased value (an “externality” in economic jargon) could be weighed against Thomson’s loss (which they allege is around $10 million per year) in reaching a reasonable decision about contract enforcement.
Second, will we see a movement to cover databases under some kind of database protection law, potentially separate from copyright, if corporate database verdors are unsatisfied with even the low bar necessary for copyright protection and with the need to use licensing provisions where that protection is unavailable? It is this kind of extension of intellectual property protection to subject matter that has traditionally not been protected that I mean by the phrase “copyright creep.” Such sui generis protection (not rooted in copyright principles) has been adopted in the European Union, and it is common these days to hear complaints about it from scholars in EU countries. At a minimum, such protection would raise costs for obtaining access to commercial databases and, as is shown by the Zotero lawsuit, could be used to stifle innovation and cooperation. The last attempts to introduce legislation for database protection in the US were several years ago — there is a nice summary of those efforts and the issues they raised here — but it is a topic that keeps coming back and about which higher education needs to be vigilent. In many ways our interests would cut both ways in any database protection debate, so it is a case where careful thought and balance would be needed.