Science Commons was re-integrated with Creative Commons. This content is no longer maintained and remains only for reference.

Background Briefing

Background:  Policies for open access to scientific knowledge

Providing standard, responsible copyright agreements ensuring the right of scholars to archive their work on the public Internet.

Scholarly communication primarily involves three kinds of information: (1) data generated by research, (2) peer-reviewed journal articles explaining and interpreting the data and (3) metadata that describes or interprets articles or their underlying data. At each of these levels, the Internet and associated digital networks create a range of opportunities and challenges for changing the nature of what information is gathered, stored and communicated, as well as how and when such information is communicated, identified and located.

Unfortunately, the emergence of powerful new technological approaches, such as text mining and semantic indexing, has not resulted in powerful new public resources. Contracts between publishers and universities can explicitly forbid the use of technology on scholarly resources, and the opportunities dfimplicit in the Internet are falling by the wayside. Efforts to further the open access movement have shown real success. Unfortunately, the majority of scholarly research remains unavailable, either for reading or machine-assisted processing.

There are a number of policy developments that show promise for addressing this problem. Major government funding agencies have begun to join foundations like the Wellcome Trust in adopting policies to mandate open access to funded research. Faculty at educational institutions like Harvard University, and key organizations like the Association of Research Libraries (ARL), are promoting and implementing policies to open access to institutional knowledge. Extensive research has been conducted into the technical barriers to providing access to the literature, resulting in both software and “how-to” guides for open archiving.

However, critical barriers remain. Even when research is funded by taxpayer monies, it is not yet flowing routinely into open repositories. Faculty bodies and university administrators are recognizing the growing importance of open access, and have started to implement institutional policies, yet a gap persists between the enactment of policy and the emergence of an infrastucture to make it easy for researchers to comply. Further, the legal issues surrounding sharing research are highly complex, and standards have yet to take hold.

Barriers to complying with policy

The most authoritative resource on journal policy for self-archiving (the RoMEO/SHERPA project) asserts that more than 90% of journals allow some form of archiving. In theory then, most authors in most journals can indeed legally self-archive. In practice, whatever the journals may state as a policy, many scientists are deterred by the legal issues, which many authors do not understand.

Restrictions against archiving arise from the industry practice of requiring that authors assign copyright to the publishers. In return for the publisher absorbing the costs of peer review and publication, the authors alienate their copyrights. These agreements do often allow authors to make archival copies available on the Internet, but policies and terms vary widely with regard to the timing of release, file formats and location of archives. Some journals obscure their archiving policies on the back pages of websites, or don’t publish an archiving policy at all.

Scholars are reluctant to challenge these policies, out of a fear of retribution or rejection by the journals, or because publication is central to to the scientific career, figuring in grant review, tenure decisions and peer esteem. This means that any legal barrier to archiving, perceived or real, makes elusive the goal of universal archiving.

The legal friction has not gone unaddressed. Funders, librarians and universities have devised several addenda to copyright transfer agreements and suggested contract language to make compliance rights explicit. These proposals provide a mechanism and tools to negotiate and modify an existing copyright transfer agreement. Other approaches include replacing the copyright alienation systems with non-exclusive grants of a “license to publish,” wherein the author retains copyright and thus the rights to archive the article.

Early anecdotes suggest some success in terms of publisher acceptance of these new approaches, but none of the existing legal tools rises to the level of a standard. This creates, in turn, the risk of exacerbating the level of confusion, inconsistency and arbitrariness of the present system.

What would a solution look like?

Authors need to have the clear and unambiguous freedom to engage in their normal everyday scholarly activities without contending with complex technology, continuous amendments to contracts or the need for a lawyer.

Although institution, funder and policymaker may craft different legal text to reserve archiving rights for different classes of journal policy, these variations reveal clear patterns of use: the key variations are time to release an archived document, the format of the archived document and the location of the archived document. These patterns can be identified and expressed in standard contract language. With significant community participation, the scientific community can build a new set of shared norms instead of wasting valuable resources through repeated negotiations.

The solution can also draw on lessons from culture and software. For maximal impact, final customized agreements should exist not only in complex legal terminology (“lawyer-readable” form), but also in a “layperson-readable” summary to enhance understanding and delineate choice.

Furthermore, the solution can also draw on new technologies. The final customized agreements should exist in a machine-readable format, which confers an advantage in ease of use: scholars can pose a set of questions, such as, “In what journal am I publishing?” and automatically create a legal document appropriate to their needs. A secondary benefit of machine readability is a trackable set of outcomes that will facilitate further improvements of both forms and policies.