Brian F. Lavoie
Stewards of the scholarly and cultural record have grappled with the difficulties of preservation since the advent of recorded information. Necessity is the mother of invention, however, and preservation objectives have been met in some highly innovative ways:
On the dissolution of the Jesuit Order in 1773, the books from their Brussels house were allotted to the Royal Library. As the Library had no space to accommodate the new accessions, the volumes were left temporarily in the Jesuit church. The building was infested by mice, and the problem of how to protect the books was anxiously debated. The solution was to employ the secretary of the Literary Society to make a selection. “Useful books were to be placed on shelves in the middle of the nave, and the remainder left on the floor. In this way, it was calculated, the mice would satisfy their appetite on the latter, leaving the former unharmed” (Hobson 1970, 15).
New threats, such as “yellow snow,” “vinegar syndrome,” and “bit rot,” have long since succeeded mice as the bête noire of preservationists. Yet even as the enemies menacing the perpetuation of society’s memory change over time, one challenge is as familiar to modern preservationists as it was to the eighteenth-century Belgian Literary Society: the difficulty in marshaling sufficient resources to counter the ravages of mice and their modern equivalents. Unfortunately, the eighteenth-century solution to this problem persists to this day: Hard choices must be made, and all too often, only a portion of the materials at risk-and not always the most valuable-are selected for preservation, leaving the rest to be nibbled away over time.
Technological innovation leaves in its wake a steadily improved capacity to create and disseminate information. Recorded information is the raw material that serves as input to the preservation process; consequently, as we extend our ability to make more of it, there is a corresponding need to expand the scale and scope of the processes aimed at securing its long-term retention. But the capacity to produce information has overtaken, and indeed is accelerating away from, the capacity to preserve it. As a result, preservation efforts are left to cope with the twin challenges of an ever-increasing quantity of at-risk materials, recorded in formats of ever-increasing sophistication.
These challenges can be partially met by nurturing the continued development and improvement of preservation techniques, processes, and workflows: in other words, by building and refining the technical infrastructure needed to support the long-term retention of the scholarly and cultural record. In concert with this, however, must come the development of the associated economic infrastructure: the mechanisms by which resources are allocated and organized to achieve preservation objectives.
This issue is certainly not new. The rapid development and proliferation of digital technologies have only amplified the scale and scope of the problem. But the digital revolution has also exposed weaknesses in traditional strategies for confronting the economic imperatives of preservation and, in doing so, has reinforced the need to revisit the question of how preservation can be shaped into a sustainable economic activity.
Preservation: Past, Present, and Future
The scope and scale of preservation, as well as the form its processes take, are, like many activities, constrained by limited resources. Responding to Nicholson Baker’s charge that librarians, archivists, and other information professionals have overseen the loss of countless information resources in their original form through misguided reformatting initiatives, Richard Cox argues that “preservation is expensive and . . . preservation that assumes the maintenance of all originals is expensive beyond our (or Baker’s) wildest dreams” (Cox 2001).
Partly as a consequence of its significant cost, preservation has frequently been characterized by procrastination. This in turn has led to sporadic bursts of preservation activity and funding, often taking the form of large-scale, Manhattan Project-type programs aimed at retrieving a situation that has already reached a state of crisis. The Brittle Books Program established by the National Endowment for the Humanities in the 1980s in the United States was a response to the belated recognition that materials printed on acidic paper were disintegrating into “yellow snow.” The discovery that motion pictures produced on nitrate cellulose film stock prior to the 1950s were also crumbling led to the establishment of organizations such as the Hollywood-based Film Foundation and the Library of Congress National Film Preservation Board, tasked with arresting the process of decay and loss. Information resources are not the only subjects of this crisis-management approach to preservation: The famed American frigate Constitution was a deteriorating hulk, relegated to service as a floating barracks, when she was rescued and restored in time to celebrate her one-hundredth anniversary in 1897.
As digital technologies for creating and disseminating information proliferated rapidly in the 1990s, a new preservation crisis loomed. The widely cited 1996 Task Force on Archiving of Digital Information report warned that “failure to look for trusted means and methods of digital preservation will certainly exact a stiff, long-term cultural penalty” (Task Force 1996, 3). Anecdotes illustrating the danger of inaction circulated widely: the oft-told story of the unreadable tapes containing a portion of the 1960 U.S. Census is but one example. The task force even recommended a few initiatives on a Manhattan Project-scale to address the crisis, including the creation of a “deep infrastructure” for preserving digital materials, as well as the establishment of a legal right for certified archival repositories to “exercise an aggressive rescue function” on behalf of significant materials perceived to be at risk (Task Force 1996, 8).
Up to this point, the crisis of digital materials has seemed to parallel the crises of embrittled paper, disintegrating motion pictures, and, to some extent, even decaying warships. A significant corpus of material has fallen into a state of disrepair through a failure to anticipate the need to take steps to secure its long-term retention. Programs and initiatives have sprung up to address the problem, supported by an ambitious research agenda focused on the processes needed to perpetuate these materials over the long term. These efforts have largely taken the form of “one-off” activities, funded by grants from public agencies or philanthropic organizations. It would seem, at least from an economic standpoint, that very little has changed in the preservation landscape.
In fact, the digital age has indeed wrought changes on the landscape, and these changes are of sufficient magnitude that the traditional paradigm of preservation through crisis management, or rescue ex post facto, will likely prove inadequate for meeting the preservation requirements of the twenty-first century. A growing proportion of the scholarly and cultural record takes the form of complex, networked digital resources. These resources are characterized by a degree of fragility and technology dependence far exceeding that found in most analog media. Some knowledgeable sources believe it is not an overstatement to say that in the case of digital materials, the preservation process must begin at the time of creation and proceed as a relatively continuous process over time-perhaps not far removed from the day-to-day management of collected materials. The difficulties in achieving the transition from preservation as a discrete event to preservation as an ongoing process will be amplified by the scale of the problem, as well by as the fact that there is little scope for postponing the time when preservation issues are confronted. In short, there is every indication that preservation activities will increasingly become immediate, large-scale, and sustained.
Preservation from an Economic Perspective: Three Key Areas
Much of the preservation research associated with the newest forms of information resources addresses the technical aspects of securing their long-term retention, that is, developing and refining the techniques by which these fragile materials can be maintained over long periods of time. But technical issues are only one aspect of sustainable preservation activities. Ultimately, these technical processes must be coordinated with the economic process of marshaling and organizing sufficient resources to achieve preservation objectives. In this regard, preservation in the twenty-first century will represent a significant departure from traditional practice. From an economic perspective, preservation will be redefined in three areas: responsibilities, incentives, and organization.1
Preservation can be construed as an economic activity, in the sense that decision makers evaluate the associated costs and benefits and, in light of the result, determine the level of resources, if any, they will commit to it. Therefore, in considering the economic implications of preservation in the twenty-first century, a useful starting point is the identity of the decision makers who are likely to bear the responsibility of committing the resources necessary to meet preservation objectives.
Some of these decision makers are quite familiar with and have deep roots in the preservation community; they include collecting institutions such as libraries, museums, and archives, which perceive the perpetuation of the scholarly and cultural record as a fundamental component of their organizational mission. But the newest technologies for recording and communicating information have introduced new stakeholders into the preservation process. These stakeholders also embody a decision-making capacity in regard to preservation, and consequently, the division of labor traditionally governing preservation activity-in terms of the distribution of both responsibility and cost-has become unsettled.
A distinguishing feature of the networked digital age is that culturally significant materials are often outside the custody of the traditional stewards of the scholarly and cultural record. Digital materials that are licensed or subscribed to, such as e-journals, e-books, and online databases, are prominent examples; Web sites are another. The common theme across all these resource types is that the function of providing access to users is often separated from that of maintaining custody of the materials themselves, that is, physical possession of the “bits.” The consequence of this separation of access and custody is that the entity who perceives the value or benefit of taking steps to secure the long-term retention of these materials is often distinct from the entity who owns the materials and therefore controls their long-term disposition.
The likelihood that the preservation process will increasingly locate its starting point in the earliest stages of the information life cycle, at a time when custodianship of culturally significant materials may lie outside the custody of collecting institutions, suggests a need for at least a partial reallocation of the responsibility for preservation away from traditional stewards of society’s memory to entities with no long-standing commitment to, or interest in, preservation of the materials under their control-for example, commercial content providers, software developers, and Web masters. It is imperative that these new decision makers in the preservation process be aware of their responsibilities and take steps to meet them.
There are signs of a growing recognition that the need to address preservation responsibilities extends beyond collecting institutions to encompass other stakeholders. A joint statement in 2002 on “preserving the memory of the world in perpetuity,” issued by the International Federation of Library Associations and Institutions and the International Publishers Association, observed that “Libraries have traditionally taken care of the publications they have acquired, and have saved the physical artifact to safeguard the information contained in it. With digital information the safeguarding of the content becomes more of a shared responsibility between the producer and the collector of the information” (IFLA and IPA 2002).
It is easy to imagine the responsibility for preservation becoming even more diffuse than a two-part division of labor between collecting institutions and content providers. Efforts to collect and perpetuate materials of cultural significance may be undertaken by individuals and organizations motivated by nothing more than a keen interest in, and a willingness to assume the trouble and expense of, preservation. This approach has been successfully adopted in regard to preserving portions of the Web perceived to be of long-term significance. A useful example is September11News.com, a Web site that describes itself as a “permanent Internet archive of the events of September 11, 2001.” Created by an individual named A. D. Williams, the site contains images, news articles, speeches, and other content documenting the tragedies of September 11.
Preservation of the scholarly and cultural record cannot rely solely on ad hoc, informal efforts. The bulk of preservation activity must take place through professionally managed, programmatic efforts undertaken by stakeholders directly associated with the information life cycle of creation, management, and perpetuation. The decision makers among whom preservation responsibilities are apportioned must take steps, either severally or in concert, to achieve preservation objectives. To ensure that this occurs, each decision maker must perceive appropriate incentives to participate in the preservation process.
As mentioned above, it is likely that preservation in the twenty-first century will increasingly take the form of an ongoing process, rather than a one-off activity conducted at irregular intervals. In a sense, this suggests a reinterpretation of what is meant by preservation, that is, preservation as process, rather than event. Transformation of the preservation process into an ongoing, perhaps even day-to-day, activity is likely to affect the allocation of resources to preservation in two ways: first, by increasing, in absolute terms, the amount of resources that must be committed to preservation; and second, by making the availability of these resources a matter of continuous, rather than occasional, concern. The incentives to preserve-the underlying motivations that collectively induce an entity or set of entities to contribute toward the realization of preservation objectives-must reflect these new imperatives.
But the question of incentives runs squarely into the issue discussed in the previous section, namely, the likelihood that preservation responsibilities will extend beyond the customary stewards of the scholarly and cultural record to encompass decision makers not traditionally associated with preservation. For a collecting institution, the incentives to preserve are couched within their mission of collecting, managing, and providing access to information resources. But for other stakeholders, the incentives to preserve are not fundamental to their organizational missions and therefore will be subjected to careful scrutiny prior to any resource commitment.
So, even as the incentives to preserve become more important, in the sense of needing to reflect an expanded commitment to preservation on a relatively continuous basis, they also become less assured. The new forms in which the scholarly and cultural records are manifested-in particular, as networked, digital materials-have driven a wedge between the objectives of preservation and the incentives to carry it out. In the past, collecting institutions typically dealt with physical materials that were purchased outright and transferred into their custody. The incentives for preserving materials of this kind, as well as the legal right to preserve, were vested in the same entity. But as a growing proportion of collections takes the form of materials that institutions provide access to, but do not physically possess, the incentives to preserve, as well as the right to preserve, may be distributed over multiple entities.
What are the prospects that a set of materials, perceived to be at risk, will in fact be preserved? Answering this question must begin with an examination of the incentives to preserve in the context of all relevant decision makers acting as stakeholders in these materials. Three key decision-making roles are of particular importance: the decision maker constituting the need to preserve, in the sense of recognizing a value in securing the long-term retention of the materials in question; the decision maker constituting a willingness to preserve, in the sense of setting up and managing the necessary preservation processes; and the decision maker embodying the right to preserve, in the sense of being vested with the authority to permit (or not permit) the preservation process to go forward.
A single entity may embody one, two, or even all three of these decision-making roles, leading to a host of possible relationships between the need to preserve, the willingness to preserve, and the right to preserve. The organization of these decision-making roles, in regard to their distribution among one, two, or three distinct entities, can have profound implications for the underlying incentives to preserve. Imagine, for example, if the right to preserve is embodied in a decision maker distinct from the one that identifies a need to preserve? What are the incentives for the former to commit resources toward achieving the preservation objectives of the latter? An economically sustainable preservation activity must be supported by sufficient incentives to preserve, and these incentives are shaped largely by the organization of the decision-making roles underlying the activity.
In circumstances where the incentives to preserve are not sufficient to achieve preservation objectives, policies and other measures may be instituted to correct for any perceived incentive gaps. A variety of instruments may be employed to enhance, or artificially create, preservation incentives. These include, but are not limited to, government subsidies to fund preservation in the absence of private incentives to do so; legislation that makes preservation compulsory in certain circumstances, either by forcing the owner of the materials to undertake it or by requiring the transfer of the right to preserve to another entity willing to undertake preservation; and negotiation among relevant stakeholders to strike a balance between preservation objectives, preservation incentives, and the distribution of responsibilities and costs. The appropriate policy instrument will be a function largely of the circumstances surrounding a particular set of materials.
An appropriate distribution of preservation responsibilities among relevant stakeholders, accompanied by sufficient incentives, or motivations, for preservation to be undertaken, still falls one step short of characterizing the components of the complete economic infrastructure that is needed to support sustained preservation activities. The last piece that must be addressed concerns the organization of these activities: more specifically, strategies for organizing preservation resources to realize preservation objectives in the most efficient way.
A number of alternatives are possible. Historian Robert Darnton proposes the establishment of a national agency, tasked with “maintain[ing] a record of everything printed, painted, sung, acted, and composed within the public sphere.” Darnton goes on to observe that a “collective memory bank of this sort should not be an expansion of the Library of Congress but rather a new entity-public but independent of politics, open to all but closed to lobbying, autonomous but administered in the public interest by a board of trustees” (Darnton 2002).
The creation of a national agency entrusted with such a diverse portfolio of preservation responsibilities is both ambitious and, in all probability, unrealistic. However, Darnton’s suggestion hints at what is likely to be one of the most important themes regarding the organization of preservation activities in the twenty-first century: that preservation will become an increasingly centralized, large-scale activity. Several factors point toward this conclusion.
The media on which information is recorded are becoming more sophisticated and functional, but also more fragile and technology dependent. Consequently, the time horizon beyond which preservation issues must be addressed is shrinking. In the extreme, the preservation process may begin as soon as an information resource is created and proceed on a continuous basis over time. Preservation will almost certainly become a more ubiquitous component of the day-to-day collection-management responsibilities of custodial institutions and will absorb a commensurately larger portion of their operating budgets. In these circumstances, it is reasonable to imagine preservation undergoing a transformation from something akin to a cottage industry, performed by a relatively small number of highly trained practitioners, to something that more closely resembles an assembly-line operation, performed more frequently and on a larger scale than in the past.
If preservation does in fact become an increasingly routine component of collection management, rather than an activity that can be postponed indefinitely or even disregarded altogether, several consequences are likely to follow. First, as discussed above, more institutions, organizations, businesses, and individuals will come to perceive themselves as stakeholders in the preservation process. Second, the increased frequency and scale of preservation activities will encourage the development of a consensus in terms of what “successful preservation” means in regard to particular classes of information resources. Third, the emergence of a consensus of this kind will eliminate much of the idiosyncratic nature that currently characterizes the preservation of complex digital materials, leading to preservation processes that are well understood and standardized across broad communities of stakeholders.
These consequences suggest that in the twenty-first century, stakeholders in the long-term retention of culturally significant materials may realize considerable benefits by organizing preservation as a collaborative enterprise, in which broad communities band together to achieve shared preservation objectives and, in the process, leverage common infrastructure, lower costs, and eliminate redundancies. There are multiple strategies for organizing preservation efforts to realize the advantages of centralized, large-scale economic activities. One approach is the development of a sustainable market for preservation services. Such a market would consist of, on the demand side, groups of stakeholders who have a common interest in the preservation of a particular set of at-risk materials, and, on the supply-side, trusted entities specializing in the provision of preservation services. A market of this type could yield a number of benefits, chief among them being that specialization in preservation services creates the opportunity to achieve lower costs and higher production efficiencies through economies of scale. In addition, a growing demand for preservation services, combined with a centralized, large-scale approach to organizing the provision of these services, might serve to expand the menu of standardized preservation services that could be offered in an economically sustainable way, compared with what might be possible if preservation were organized as a highly dispersed, small-scale activity conducted by institutions for which preservation was but one of many responsibilities.
Key to the development of a sustainable market for preservation services is the cultivation of trust between those who would supply preservation services and those who potentially would avail themselves of these services. Even with the efficiencies realized through economies of scale, preservation will remain an expensive proposition. Preservation is an investment-a cost incurred up front in order to obtain benefits that might not be realized for decades. The risks associated with failure are potentially enormous, not only in monetary terms but also in terms of the incalculable loss that would attend the destruction of some portion of the scholarly or cultural record. Institutions that are willing to invest in preservation services will need assurance that these investments will be protected.
From an economic perspective, many of the obstacles to preserving the scholarly and cultural record in the twenty-first century seem quite familiar. In the broadest sense, they can be distilled into the venerable economic problem of matching scarce means to competing ends. But the scale and scope of the current preservation challenge suggest the need to reexamine the mechanisms by which resources are channeled to preservation activities. Fundamental to the development of a new economic infrastructure for preservation is the recognition of an increasingly diverse set of decision makers associated with the preservation process; an understanding of the complex relationships that might arise between the need to preserve, the willingness to preserve, and the right to preserve; and a reevaluation of how to organize preservation resources to meet preservation objectives in economical ways. By engaging all stakeholders in the preservation process, ensuring that appropriate incentives to preserve emerge, and organizing preservation activities in ways that leverage resources and maximize efficiency, significant progress will be made toward preventing the twenty-first century equivalents of mice from eating around the edges of society’s memory.
Cox, Richard. 2001. Don’t Fold Up: Responding to Nicholson Baker’s Double Fold. Available at: http://www.archivists.org/news/doublefold.asp.
Darnton, Robert. 2002. “The ‘Great Book Massacre’: An Exchange.” The New York Review of Books 49(4).
Hobson, Anthony. 1970. Great Libraries. New York: Putnam.
IFLA and IPA. 2002. Preserving the Memory of the World in Perpetuity: A Joint Statement on the Archiving and Preserving of Digital Information. Announcement dated August 12, 2002. Available at http://www.ifla.org/V/press/ifla-ipa02.htm.
Lavoie, Brian. 2003. The Incentives to Preserve Digital Materials: Roles, Scenarios, and Economic Decision-Making. OCLC Office of Research. Available at http://www.oclc.org/research/projects/digipres/incentives-dp.pdf.
Task Force on Archiving of Digital Information. 1996. Report of the Task Force on Archiving of Digital Information. Washington, D.C. and Mountain View, Calif.: Commission on Preservation and Access and Research Libraries Group, Inc. Available at http://www.rlg.org/ArchTF/.
1The following discussion is based in part on Lavoie 2003.