We have become so used to accessing scholarly articles, books, and research materials online that it is easy to assume that once something has been published digitally, it will continue to exist somewhere, especially when it is hosted by a well-established publisher or academic institution. Yet anyone who has worked with digital collections for an extended period knows that websites disappear, publishing platforms change, files become corrupted, and organizations that once seemed financially secure may eventually close or change direction, leaving us with the question of what happens to the content they were responsible for maintaining. These are familiar problems in the digital environment, and their implications become more serious when we consider scholarly publications that are expected to remain part of the permanent academic record, long after the technologies and institutions that originally supported them have changed.
Looking through the CLOCKSS document on threats to digital preservation, which identifies 13 categories of risks that could affect the long-term survival of digital content, what is striking is how many of these risks have relatively little to do with the way we usually think about storing content. The document examines everything from hardware and software failures to natural disasters, cyberattacks, human mistakes, economic pressures, and organizational collapse, and it brings together issues that are generally handled by different people and departments but that, from a preservation perspective, are closely connected. Each can ultimately affect whether information survives. This document also lists concrete measures used to reduce the impact of these risks and provides a useful explanation of why digital preservation requires considerably more planning than maintaining copies of files in several locations.
Hardware failure is one of the risks highlighted, and although we are all familiar with computers breaking down or storage devices needing replacement, the issue becomes much more profound when we consider what happens to a collection of millions of scholarly publications that must survive many generations of hardware and software changes. The problem also goes beyond equipment that simply stops working, since digital files can become corrupted while remaining in storage, sometimes without anyone realizing that something has changed until the information is needed again. This is where the LOCKSS technology used by CLOCKSS comes into the picture, with copies of the same content preserved across different locations and regularly compared to identify inconsistencies, so that when a file becomes damaged or incomplete, the system can use the other copies to repair it. What is especially important about this approach is that it recognizes how ordinary and inevitable these failures are over long periods of time and builds the ability to identify and recover from them into the preservation process itself.
Another highlighted risk has to do with where digital content is actually stored, and this is very interesting because we tend to associate the loss of content with something happening to the technology itself. The circumstances surrounding the institutions that maintain digital content can be just as important. A natural disaster, for example, could affect an entire institution or region, damaging equipment or interrupting operations for an extended period, which is one of the reasons CLOCKSS maintains copies of its preserved content in different geographical locations. But when we think about what might happen over several decades, there are also political developments, changes in national regulations, economic difficulties, and institutional reorganizations to consider, all of which could influence whether an organization is able or willing to continue maintaining all of its digital collections. Some of these changes may happen gradually, while others can occur quite unexpectedly, and the value of distributing preserved content across different institutions and countries becomes clearer when we consider how impossible it would be to anticipate all the circumstances that might affect any one location in the future.
Other important areas are cyberattacks and human error. We often think about the security of digital information in terms of protecting systems from outside attacks, however some of the risks come from the everyday activities of the people responsible for maintaining the content. Someone might make a mistake when updating software, changing a configuration, or managing access to a system, and although these are routine activities in any organization, the consequences can become serious when dealing with scholarly publications that are expected to remain available permanently. There is also the possibility of deliberate interference, whether by someone attempting to gain unauthorized access or by someone who already has access to parts of a system. CLOCKSS addresses these concerns through a combination of distributed storage, careful controls over administrative access, active monitoring and repair, but much of the effectiveness of these measures also depends on the people managing the archive and the procedures we follow. When we consider that preservation systems are expected to operate for decades - during which people and skills will inevitably change, it becomes easier to understand why managing human and organizational risks has to be part of the preservation process itself.
Financial and organizational risks are also crucial for the long-term preservation of and access to digital content. Consider how much scholarly publishing has changed over the years and how many academic institutions, publishers, and publications have experienced changes in management, financial difficulties, or mergers, This is a reminder about how important it is to pay attention to digital content providers, and the arrangements they have in place for continued access should their organization close, particularly where the content they provide remains important beyond the lifetime of the people and institutions making decisions today.
The same questions apply to preservation organizations ourselves, of course, since sustaining an archive over time requires a sustainable organization. CLOCKSS has a succession plan and a Trustee Committee in place, so it is clear who will take decisions about the content entrusted to us for long-term digital preservation if we can no longer look after it ourselves.
The long-term survival of scholarly information depends on a combination of technical resilience, institutional stability, financial planning, and cooperation between organizations whose responsibilities and circumstances may change considerably over the years. We often discuss preservation in terms of infrastructure and technology, which are obviously essential, but the CLOCKSS threat model makes it clear that the decisions surrounding governance, funding, succession planning, and operational responsibility are equally relevant to the survival of the scholarly record. A distributed archive can protect content against many forms of technical disruption, while the international community of libraries and publishers supporting our archive provides the sustaining framework through which digital content can continue to be protected and made accessible.
Based on CLOCKSS: Threats and Mitigations.