Restrictions in the Age of Open

King Davis’ team is creating a digital archive of historical records from Central Lunatic Asylum for Colored Insane. The platform will facilitate access for researchers and families while protecting former patients’ privacy. Photo courtesy of UT Austin.

As open access cannot apply to all research, data, and publications, former Senior Program Officer Donald J. Waters offers insight about the Mellon Foundation’s support for a range of increased access to scholarly content.

Google, as part of a larger information industry complex that includes Yahoo!, Facebook, and Twitter, and other search and social media companies, has played a huge role in the last two decades in creating a general cultural expectation that the digital networked environment can operate efficiently and effectively only if content on the web is open and freely available.  Openness is a construction, not an inherent feature of the Internet.  However, the general benefits of openness are so consonant with the goals of scholarly research and teaching, that openness in contemporary scholarly communications has become a dominant theme: open journals, open monographs, open data, open access.

Policy change based on this construction has swept through institutions of higher education encouraging faculty to make their research publications and data openly available.  Publishers have realigned their business models, or created new ones to support open access.  The National Institutes of Health has articulated an open access policy for the research results funded since 2008.  An omnibus spending bill signed into law in January 2014 expanded the number of US federal funding agencies with open access mandates for research publications and data produced with federal funds.  And, since last summer, several prominent private foundations—including Hewlett, Gates, and Ford—have followed suit and issued similar mandates of their own, requiring open access in the form of Creative Commons licenses.

Of course, however desirable it is, open access does not—and cannot—apply in all circumstances.  In every announcement of policy change designed to promulgate open access, the fine print contains important exceptions.  For example, deep in the announcement by Ford, one reads:  "In certain situations, such as when the expected product contains sensitive or confidential material, the Creative Commons license will not be a requirement."  As Nicholas Lemann pointed out in New Yorker article (December 1, 2014), even "Google's 'default to open' ethos does not extend to its [own] invaluable search and ad algorithms," which are the source of its ability to generate revenue.

Indeed, at the heart of many scholarly pursuits are materials that are essential to the understanding of various dimensions of the human condition, but may be restricted for use for all sorts of very good reasons, including the need for financial sustainability of the resources, respect for intellectual property, the appreciation of privacy, security, and confidentiality, and sensitivity to cultural significance.  In many cases, these barriers can be neither wished nor mandated away.  Nor can restricted materials be ignored without severely biasing research and teaching toward the large and growing, but still limited and partial, body of openly available material. 

Because of its complexity, the online information environment demands a more robust intellectual property policy than simply emphasizing open access.  The Andrew W. Mellon Foundation is among those private philanthropies that have not made open access the main or defining quality of its policy.  Instead, beginning in 1999, the Foundation articulated its general interest in ensuring "that products developed with the assistance of Foundation grants will be used for the greatest possible educational, social, and charitable benefit."  The statement went on:  "The Foundation's grantmaking philosophy is to encourage proposals and practices that promise to yield products for broad public use and to discourage those that involve or promote proprietary interests except to the extent that a charitable end may also be served."  Crafted in this way in terms of broadest possible access, the Foundation's policy allows it not only to support open access when appropriate, but also (and equally) to encourage other worthy projects such as those that develop access to materials that need to be restricted for a variety of legitimate and important reasons.  See, for example, the grants to support the Open Library of the Humanities and to provide access to the records of the Central Lunatic Asylum for Colored Insane.

But even more is needed from funders than robust policies that respond to a complex information environment.  It is incumbent upon them to help educational institutions, libraries, publishers, and other information agencies give greater visibility to the wealth of restricted materials in a digital world that privileges the open and free.  They must help develop and perfect infrastructure, tools, and methods that advance understanding while balancing the accessibility of information against growing public policy concerns about the ways in which the digital medium routinely violates privacy, security, and related matters, often to the strategic advantage and financial benefit of a few very large companies.

A rudimentary basis does already exist for pursuing this important agenda.  It consists of a variety of elements, including in the legal realm collective licenses, imaginative use of the first-sale doctrine, and the application of copyright exceptions such as fair use.  In technology, advanced software emulation and the controls of machine virtualization techniques make it possible to imagine virtual reading rooms to which sensitive materials can be streamed over the network instead of requiring researchers to come to physical reading rooms.  Widespread adoption of improved mechanisms for authentication and authorization, such as InCommon and related tools, could greatly increase the formation, scale, and complexity of work with sensitive materials or products.  And corpus linguistic tools can be applied to documents to identify names, identifiers, accounts, and other sensitive information for automatic redaction.  Finally, focused experiments with these kinds of building blocks are needed, with critical attention to the creation of norms of use and governing policies.

Because so many modern and contemporary materials are subject to copyright and other kinds of restrictions, what is at stake in the development of these tools and practices is nothing less than future understanding of the 20th and 21st centuries.  For funders, policymakers, libraries, publishers, and other agencies to make a difference in this future understanding, they need to demonstrate much more nuance and imagination than simple and frequent calls for open access.

Donald J. Waters was Senior Program Officer for Scholarly Communications at The Andrew W. Mellon Foundation from 1999-2019.