Digital publishing has some major advantages. Digital copies of articles can be produced at essentially zero cost. As many copies can be made as needed by readers. Readers can access materials over great physical distances. Digital text has advantages in search and accessibility. Web links make it easy to follow citations. One of the biggest advantages is the lack of degradation. The pages of a digital journal can be turned hundreds of thousands of times without deteriorating.
As I described previously, there can be challenges with digital preservation as storage hardware and file formats continue to evolve. Most of those challenges center the experience of the individual or organization holding the digital assets. But there are also potential implications for those outside the organization who rely on that information. Publishing oligarchs like EBSCO and Elsevier hold the rights to huge numbers of journals and articles that they allow access to through their databases, but for how long?
When a publisher no longer sees a financial benefit to providing access to a journal, it may simply not make it available at all. If a publisher goes out of business, the rights to those journals may go with them. As we’ve recently seen in the United States, governments can also engage in censorship by removing research from public access, often by destroying it. As grass roots efforts to archive some of that material continue to unfold, it highlights the risks of losing scientific knowledge.
Eve (2024) set out to find the scope of the risk. The results were not encouraging, with more than half of the examined academic publishers barely meeting a threshold of 25% of their articles being included in a single archive, and more than a quarter of the articles examined were “seemingly unpreserved.” The study focused on articles with a DOI, a URL resolving service that allows for continued access to an article even if its URL changes. The DOI resolution provided by Crossref is another potential point of failure, because without it many articles will become difficult or impossible to find, functionally equivalent to being removed by a publisher.
In the days before digitization, preservation of journals and other scientific and academic writing was focused on keeping physical copies of print materials accessible and in good condition. But in the modern digital age of journal publication, libraries and archives don’t have physical copies they can maintain if a publisher disappears. Instead, they have licenses and subscriptions to journal databases as a service. The licensing terms generally enforce terms that prevent libraries and archives from making complete backups of the materials, preventing these institutions from filling that traditional role.
As Eve noted, there are some archives of these materials that publishers can choose to contribute research to. Sometimes the organization funding the research will stipulate that the results must be included in various archives like CLOCKSS to protect against these risks. But many articles, especially those published by smaller journals, aren’t included.
In a world of publisher oligarchies, restrictive licensing terms and intellectual property laws, and government censorship, it’s unclear who has the legal rights needed to produce archives of scientific literature. We’ve yet to see a clear assignment of social responsibility for creating and managing these archives in a digital age. The Library of Burnt Books only exists because the works were preserved despite such archives being illegal in Nazi Germany. As the Western world continues to swing in a capitalist authoritarian direction, questions about the ethical responsibility to create similar archives are now being raised. But with access to the materials in question already significantly restricted by publishers, the creation of such archives brings its own challenges.