Legal Pressures on the Public Domain: Licensing Practices
Susan R. Poulter
Imagine, as suggested in a recent article in the National Law Journal,1 that you enter a bookstore (or a library) and you find some books, unlike the ones you typically find, that have locks or snaps on the cover. There is a notice on the cover that states, “As a condition of opening (or purchasing) this book, you agree to the terms contained inside.” Since your curiosity is engaged, you open the book, only to find additional terms and conditions that state, among other things,
You agree that you will not copy any portion of this work or disseminate any information contained herein without the express written permission of the publisher and that you will not publicly criticize this work or the authors thereof. You further agree that in the event you make any commercial use of this work or develop any commercial enterprise or product from the use of this work, you will negotiate an agreement for such use with the publisher, terms of which may include but are not limited to a reasonable royalty.
“Ridiculous,” you say? In the paper world this would be a startling development. In the world of electronic publishing and databases, however, such terms can already be found.
Licensing practices are indeed a significant pressure on the public domain; they have the potential to become the vehicle for restricting access to and use of public information that trumps other existing and proposed forms of protection, such as copyright and database protection.
The term “licensing” in the context of electronic publishing means contracting for various activities involving the access and use of articles or the information in the articles. It includes licensing an article for publication (which can also occur through the conveyance of copyright, as has been typical in scientific publishing) and licensing the subscriber to access and read (and print) a copy. By way of comparison, paper copies of journals, including the accompanying scientific data, are typically sold outright; the electronic forms of journals more typically are licensed, however.
There are pressures on the public domain at the level of access to articles and the underlying data and on the uses that can be made of scientific and technical information once accessed. Access to articles (and the data they contain) requires subscription, and few would question the need for a mechanism to recoup the costs of electronic publishing. The primary issues are reasonableness of the subscription price and the scope and duration of the license. A more significant issue, in my view, is whether the subscriber (user) has the right to use the data; that is,
the right to extract, reanalyze, repeat, build upon, and extend the work. These rights typically have been available to any reader of paper publications, subject only to the limits of patent law. As other speakers have noted, however, the online environment facilitates technologically enforceable restrictions on access and acquiescence to contract terms that could restrict further use of the data. Moreover, these terms usually will not be negotiated in the usual sense of the word because of the unequal bargaining power of subscriber and publisher.
Will licensing provisions that restrict further use of data be upheld? Will such contracts become commonplace? For the remainder of this discussion, I will talk about the applicable legal principles and then will highlight some developments in scientific publishing that may indicate which way the wind is blowing.
THE LEGAL ENVIRONMENT
The primary legal principle applicable to licensing provisions, or at least the starting point for any analysis, is “freedom of contract.” Freedom of contract is not unlimited, however, and traditionally has been subject to limitations imposed by various other legal doctrines and policies, such as preemption by federal law (including copyright and patent law), competition policy, the doctrine of unconscionability in contracts, and perhaps even First Amendment principles. Lemley, Reichman, and others have written extensively about these possibilities in limiting objectionable terms in electronic licenses.2
With regard to restrictions on the use of scientific data, preemption under copyright law seems the most promising because the U.S. Supreme Court held in Feist Publishing v. Rural Telephone that copyright does not protect facts, and protects compilations of facts only to the extent that the selection and arrangement are original.3 Thus, copyright does not prohibit the extraction of isolated facts from a compilation, even significant portions of them, so long as the selection and arrangement are not copied (assuming it is original enough to be protected in the first place). The Supreme Court has not had occasion to decide, however, whether Feist’s declaration that facts are free for all to use can be trumped by contract. That question has been addressed in a number of lower court decisions considering “shrinkwrap” and “click-wrap” licenses that typically accompany commercial software and other digital or electronic products.
For a time, most courts held, on contract formation grounds, that the shrinkwrap licenses were invalid and that their terms did not bind the purchaser. Because the consumer was not aware of the terms of the license before making the contract (i.e., agreeing to purchase for the stated price), the license was not part of the contract between buyer and seller. (This problem need not be an obstacle in the online publishing environment, in any event, because the subscription to an online journal or database can be conditioned on acceptance of terms made known to the subscriber at the time of the subscription, before the contract is made.)
The trend toward invalidating shrinkwrap licenses began to wane in later cases, however. In ProCD v. Zeidenberg, the Seventh Circuit upheld a shrinkwrap license that accompanied a CD-ROM containing a nationwide telephone directory, overcoming the contract formation issue.4 More significantly for our purposes, Judge Easterbrook held that the license’s prohibition on further copying and use of the data on the CD-ROM was enforceable, rejecting the argument that the Copyright Act preempted terms that were in conflict with the policy of copyright as expressed in Feist.
The ProCD court appeared to be concerned about the impact of the defendant’s complete copying of the database on the plaintiff’s investment of about $10 million and on the incentives for others to invest in electronic databases of public information. The problem with the decision, however, is that the court’s reasoning suggests that other kinds of restrictions, such as a prohibition of extraction of any part of the database or reach-through provisions attempting to capture royalties on downstream uses and developments, would be upheld. The Supreme
Court has not addressed the preemption issue in the context of contractual limitations on copying and use, but in the meantime, the lower courts are tending to uphold online contracts and find that preemption does not apply.
It is unclear whether there is room for a more nuanced preemption analysis of license terms limiting further use of electronic information. The ProCD court applied the express preemption of Section 301 of the Copyright Act. This might be considered a narrow preemption analysis, because Section 301 requires a close comparison of the right asserted and the subject matter protected by the contract, with the subject matter and rights of copyright law. Another court might consider “Supremacy Clause” preemption, which depends on whether the state right in question interferes with the policy of the copyright law. Under this kind of preemption, the court might conclude that, because a strong policy underlying copyright law is that facts should be freely available for all to use, contracts cannot restrict their use once they have been made public (or available to the public). This kind of preemption analysis might also allow a database creator to prevent wholesale copying of a database but prohibit a restriction that would prevent extraction and use of more limited amounts of data.
Several proposed changes to the law will also impact licensing rights, generally strengthening the hand of content providers. Database protection has been considered in Congress since the mid-1990s, at least partly in response to the European Commission Database Directive. A database protection statute will likely strengthen contract claims because a federal law would expressly sanction protection of databases. Moreover, database protection is not the only legislative initiative that will strengthen the hand of electronic database providers and publishers. In 1999 the National Conference of Commissioners on Uniform State Laws proposed a new uniform law, intended for consideration by state legislatures, that would generally validate shrinkwrap or clickwrap licenses. The Uniform Computer Information Transactions Act (UCITA), although instigated by the software industry, is not limited to software licenses, but would cover online publishing.
UCITA contains a number of provisions unfriendly to consumers, but for our purposes perhaps the issue of most concern is that the law appears to endorse the concept that, even where information is marketed to a large group or to the general public (i.e., a “mass-market” license), the publisher can restrict the further dissemination and use of information provided pursuant to the license, whether or not that information qualifies for copyright protection. In other words, the commentary prepared by the drafters endorses the kind of approach endorsed in ProCD.
At present, UCITA has been adopted in only two states, Virginia and Maryland, and three other states have enacted legislation to prohibit the enforcement of UCITA against their citizens. UCITA also has been revised, but not to address the problem identified above. Renewed efforts to enact UCITA in other states seem likely.
ELECTRONIC PUBLICATIONS AND DATABASES IN SCIENCE
The remainder of this discussion will provide a sampling of practices and events in scientific publishing that are essentially anecdotal, although in some instances they are very significant anecdotes. I also will mention some countervailing developments, even trends, although I will leave most of that to the speakers later in the program.
At the level of access, high prices (particularly for commercial journals) are a widespread concern. But more broadly, a typical one-year subscription permits online access only during the year of the subscription. A subscription may provide access to several years of back issues; beyond that period, back issues usually are charged separately. Clearly, making back issues available online is a costly activity—the issues here are primarily ones of how electronic publishing products are packaged. It is worth noting, however, that the limited period of access means that even a long-term subscriber has no access once her subscription ends, rather than being able to retain access to those issues to which she subscribed.
A number of practices also limit the usefulness of online publications, especially for the typical researcher today who does most literature research online. In one instance, a leading scientific journal required an individual subscription for full-text access online, even where the institution had an institutional subscription. (This practice has apparently been discontinued, however, by the journal in question.) Commonly, linking to references within an online article requires subscription to the referenced journal as well. Another worrisome trend is that proprietary databases are the only source of some information developed with public funds (see the Commercial Space Act of 1998).
At the level of restrictions on further use of data, in other fields, online database providers assert the right to restrict the use of data extracted from their databases, even when that information is available elsewhere and is not separately entitled to any form of intellectual property protection. An analogous area involves the use of Material Transfer Agreements (MTAs) for the sharing of unpatented specimens or samples. MTAs commonly prohibit commercial use of the materials, or include “reach-through” provisions, requiring negotiation of licenses for downstream commercial uses.
Traditionally, professional society journals have not imposed limits on use of data. Recently, however, Science magazine contracted with Celera Genomics for the publication of the Venter et al. paper on the sequencing of the human genome, without making the sequence data freely available to all subscribers. Contrary to the usual practice, Celera (rather than Science) maintains the database of sequence data supporting the article, and readers must access the data through the Celera Web site. Access for academic, noncommercial users is free, and such users can download one megabase per week. They may not commercialize the data, but there are no reach-through provisions on developments from searching the database. Access for commercial users requires the execution of a MTA, again prohibiting commercial use, or negotiation of a separate agreement with Celera. Despite the expectation that Celera data would make their way into GenBank, the free public repository of genetic sequence data, an article that appeared as recently as August 2002 indicated that no Celera data had been deposited.
The justification for Science’s arrangement with Celera is that it made important data available that otherwise would have been accessible only through commercial arrangements, if at all. The arrangement is precedent setting, however, and is contrary to scientific norms in most fields in that it permits the authors to “have their cake and eat it too”—garnering scientific recognition while retaining the ability to exploit the underlying data by limiting access.
The influence of the Science-Celera deal remains to be seen. It does, however, provide a precedent for a scientific publication to require less than full disclosure of the underlying data. It further is a precedent for agreements prohibiting commercial use of data, perhaps requiring reach-through provisions, and for restricting extraction and further publication of data. How will journals evaluate other cases where authors want similar arrangements, perhaps even academic authors who want recognition while maintaining a head start much like the situation with x-ray crystallographers in past years that engendered a great deal of criticism? Will journal editors be able to determine when the benefits of such an arrangement outweigh the negatives by making available data that otherwise would not be available? Or will such arrangements allow restrictions on data that otherwise would be published without restriction? Ultimately, will restrictions and reach-through provisions delay and inhibit worthwhile downstream uses of science and technological information?
Members of scientific professional organizations can be expected generally to be opposed to proprietary rights in data or publication with less than full disclosure, although some, especially those who perceive commercial value to the data they generate, may seek to make such arrangements themselves. The structure of scientific publishing may also facilitate such arrangements. Many journals now use supplemental, online repositories for detailed experimental procedures and data, which could easily be subjected to licenses that include use restrictions. These supplemental data repositories are not always published in the paper edition of the journal, although many journals now make them available without an online subscription. Notwithstanding their current availability, online restrictions could completely control access to and use of data.
Not all developments, however, point in the direction of greater restrictions. Authors of some journals have retained the right to use their scientific papers for their own research and educational purposes, including, in some cases, the right to post their articles online. These movements have the potential to facilitate new kinds of highly specialized, interlinked information products. A recent report from the American Association for the Advancement of Science (AAAS) supports these efforts.5 But author-sponsored movements will likely have limited utility if the goal is widespread access. A major function of journals is to vet scientific publications for quality; unless these “entrepreneurial” databases establish similar vetting systems, they are likely to be incomplete.
Scientific authors have also tried to influence the other side of the equation through various efforts to persuade journals to allow free access online after a period of subscription-only access, say 6 months to a year. This movement has precedent—the American Astronomical Society has all journals online, with free access after 3 years. Other initiatives include the Public Library of Science, PubMed Central, and the Budapest Open Archives Initiative. A movement by life scientists to boycott journals unwilling to agree to this concept apparently has fizzled, but the movement to unrestricted access seems to be gaining momentum, with the decision of Science (the leading publication of the AAAS) to allow unrestricted access after 1 year.
In summary, digital lockup is a realistic possibility. Technological and legal tools support the kinds of restrictions that can deplete the public domain. Science today generates many large data sets (e.g., gene sequences) that can be managed only in digital form. Moreover, the legal framework does not currently set limits on licensing provisions needed to preserve public domain. Member interests may restrain professional society publications and databases from imposing more onerous restrictions, although it is unlikely that there will be unanimity on this issue. Directions are even less clear for proprietary and industry-sponsored databases.
Trends are not uniformly pointing toward greater restrictions, however. Authors’ rights movements and the movement toward free online access are welcome developments. The real question is how the system of technological tools and legal rules can be adjusted to strike the right balance—maximizing the data and information in the public domain to “further the progress of science and the useful arts.”