The Library of Congress >> To Preserve and Protect

Publications (The Library of Congress)

Innovations in Security and Preservation

21. What Can We Afford to Lose?
Abby Smith

Preservation is deemed an excellent thing by all, yet is funded by few. Why? What prevents institutions and individuals from being willing to "pay their way" in this area as they are willing to do so in many others: cataloging, acquisitions, hardware, and software?

There are a number of factors at work, some of which are social, some psychological, and some of which have to do with traditional library business practices.

First, we must acknowledge that there are powerful social forces that keep preservation from competing successfully for our attention. We are not a culture of ancestor worshipers here in America. On the contrary, our culture places high value on things having immediate reward, no matter how small, over against those having delayed benefits, no matter how great. The national savings rate, which is now calculated to be in negative territory, is but one exemplar of this attitude. The savings rate is below our rate of expenditures either because we choose to ignore warnings about the need to save for the future, or, for those more financially savvy, because the return on investment available on the market makes saving appear to be a waste of time and resources. The so-called new economy is booming precisely because of technologies—information technologies primarily—that maximize immediate return over long-term gains.

Something parallel is occurring in the current sociology of libraries. These same information technologies are making libraries more effective at delivering services to their patrons anytime, anywhere. One of the unintended consequences of these technologies, though, is that they divert libraries' attention from preservation to access. They divert not only our attention, but also our funds. The reasons are not hard to see. We all feel an urgent sense to keep up with the fast pace of technological change, and to do so takes enormous sums of money involving us in a never-ending search for good people, because we cannot seem to retain our best people for long, and obliging us to educate our funding bodies and trustees about the consequences of changes that we do not ourselves fully understand.

Second, preservation has lost its sense of urgency for what might be called professional psychological reasons. With so much to do, why do today what you can put off until tomorrow? This psychological preference for instant gratification over delayed gratification is perfectly understandable. Frankly, delayed gratification too often feels like no gratification, for in the preservation game, the pay-off is indirect, accruing to others. In a way, even preservation people have to admit the intractability of this psychological disadvantage. It is generally easier to recruit bench conservators than preservation managers, because the rewards of handling the materials are so immediate. Repairing damaged items often feels better than preventing damage in the first place.

But as professionals, have we really lost the sense that we are the beneficiaries of the actions of those who came before us? I do not think so. At the end of the day the chief stumbling block to funding preservation is that we have yet to find the right answer to the key question: "What is the value of this stuff, and what benefit as individuals, institutions, and a society do we derive from keeping and sharing it?"

We do not understand how to demonstrate the value of preservation in a meaningful way. People struggle constantly to do so and nearly always fall back on anecdotes, usually involving item-level conservation. We could, of course, put a dollar value on our collections, but that tends to fix the value of individual objects as they would be valued if they were to appear on the market today. Artifactual value may effectively be calculated this way; but what about objects that have high intrinsic research and information value but little artifactual value? In considering artifacts, take the example of the Bible that Abraham Lincoln held when he took the oath of office as president, an item found in the Library of Congress's collections and often trotted out to give potential donors and visiting potentates a case of the shivers. Frankly its research value is close to nil. It is one of an undistinguished print run, and the text is well known, to say the least. If we lost that item, we would lose no information. But to the extent that it has associational value, it clearly is irreplaceable. Because of the charisma that attaches to it through association, that Bible would fetch a handsome price were it to go on the market. And because of its charisma, it is the beneficiary of strict security protocols and responsible preservation care.

But what about those other items in a research collection that have no appreciable market value but are, in their own way equally irreplaceable—sheet music from the nineteenth century, for example? Or the early editions of Huckleberry Finn—not the first edition, but subsequent ones that yield so much information about the reading public of the time. How do we avoid the problem of having to replace items like that? The best approach for securing these institutional assets—for that is what they are—is to identify the factors that put these items at risk as objects with research value, and to mitigate those risks in the most cost-effective way possible with current technologies. This is what I refer to as the risk assessment model. It works for collections that include items of high financial value as well as high research value.

This approach does not ask the question "How much can we afford to spend on preservation?" That answer, as we all know, is "never enough." With an answer like that, it is hard to know where to begin to invest the resources we do have, no matter how inadequate. Rather, I propose that we ask "How much can we afford to lose?"—knowing full well that preservation is about reasonable trade-offs, that technology will offer us solutions in the future that we do not even dream of now, and that planning for failure is the best way to mitigate its effects.

Library preservation differs from museum preservation in that libraries are always looking to the item's use and fitness for purpose. The risk assessment model I propose here is focused entirely on fitness for purpose: how is that object going to be used? Let us take an ordinary library object—a book. What threatens a book, makes it useless? It could be misplaced, inadvertently misshelved. It could be incorrectly cataloged and hence unretrievable, or it could be languishing uncataloged in a backlog somewhere on a book truck or cataloger's desk. It could be embrittled and crumble when you turn pages. Or it could be physically damaged through vandalism—the illustrations razored out—or just plain stolen. In the language of risk assessment, these risks pertain to:

(1) inventory control: where is it?

(2) bibliographical control: what is it?

(3) preservation control: is the information intact and the item usable? and

(4) security control: is the item unduly at risk of theft or mutilation?

What is useful about this approach, in my view, is that it describes the day-to-day business of libraries, only in a language that is more accessible to financial officers, presidents, and CEOs than the terms we use among ourselves. Libraries having significant collections are increasingly directed by or responsible to men and women who are not trained librarians. As vitally interested in the health and well-being of their institutions as they are, they do not share the same assumptions, skills, and expertise as catalogers, preservation specialists, and curators. As individuals having fiduciary and financial responsibilities for the institutions they oversee and their assets, they make or are responsible for difficult choices in a time of increasing demands on essentially flat budgets. One of the advantages of a risk assessment model for library collections is that it defines those collections as primary institutional assets, an inventory built up over decades and centuries that is critical to the ability of a library to fulfill its mission: to serve its patrons the information and cultural resources they need. It defines the collections not as sunk costs, but as primary investments that need additional funds to keep them productive.

In partnership with KPMG Peat Marwick, the Library of Congress developed and has implemented a risk assessment model for the management of its collections, known in the federal accounting trade as "heritage assets"—a bewitching term—a term that I understand means that the value of this asset can never be used up. The risk assessment, conducted every year during the institutional audit, works from established benchmarks and provides a rational basis for developing long-range plans and the budgets to implement them. In other words, it is based on evidence—objectively and systematically gathered data about the state of the holdings and their vulnerability to various risks. It provides a flexible and common framework for determining the needs of collection items as various as baseball cards, videotapes, incunabla, and microforms. The value of each type of item in the collections is defined by its purpose, and the well-being of that item by its fitness for purpose, which makes this approach dynamic and focused on the use or potential use of that item.

This model is described in great detail in a report published by the Council on Library and Information Resources in cooperation with the Library of Congress, Managing Cultural Assets from a Business Perspective. [1] The report begins from the premise that, because these are the primary assets of the institution, the question is not "How much can we afford to pour into these collections?" but rather, "How much risk do we take if we fail to invest in our asset base?" It guides managers in identifying specific risks in their libraries and deciding what level of risk is acceptable versus unacceptable. It provides a step-by-step description of a process of risk evaluation that involves everyone in the institution who is responsible for the collections. This means not only those who work directly with collections, but also those responsible for security, buildings and grounds, and, most important, the information technology infrastructure. After all, inventory and bibliographical controls are absolutely essential to all aspects of security, preservation, and service. So, whoever maintains the Online Public Access Catalog (OPAC), the integrated library system, and keeps it up and running, is as critical to the good stewardship of library collections as a cataloger or a rare book conservator.

This model works just as well with digital assets as with rare books or the treasures in the nitrate film vaults in Ohio. With respect to digital assets, it seems clear to many thoughtful people that the growing availability of information online raises the essential question, "Do we really need all this stuff in the first place?" Are we not best off putting our scarce preservation resources into items that will be selected for an exhibition or digitized for Web distribution—something, in other words, with a probable—a calculable—demand for access? As library materials become increasingly available on the Web, do we really need to keep a lot of the nondigital resources that we have now?

The answer, of course, is that many if not most of the items research libraries acquire have been collected for their research value. Because of the numerous constraints on doing research on the Web, many of these materials will never be digitized—not just because they are intrinsically low-use, but because they are valuable chiefly in the original or may, owing to copyright issues, be used only that way.

But what is research value? This is a question given too little examination, in my view. It seems to be a lot like pornography—we cannot define it per se, but we all know it when we see it.

But of course, the problem is that we often do not know it when we see it. That is certainly the basis for Nicholson Baker's criticism of libraries' treatment of original newspapers. How many decisions have we made in the past—not only about deaccessioning and pulping, but even about rebinding—that we now regret, even if we try to avoid talking about it in public? How do we measure research value—what are its attributes? Do rarity, association, beauty—all the things that we recognize in the value of the artifact—have any meaning here? If so, how do we recognize these qualities? Can we develop objective criteria that allow us to discriminate between objects that must be selected and retained in a collection and those that need not be, or at least need not be retained in the original in all cases?

If we are to argue for the resources we need in order to keep collections fit and accessible, we must recognize that this is not a question that librarians and archivists can answer alone. It is all members of the research community, however you define that, who need to articulate the new role that collections are playing in the production of knowledge. The Council on Library and Information Resources (CLIR) has been working for a year now with the Task Force on the Artifact in Library Collections, composed of scholars, librarians, and academic officers, to investigate the meaning and role of the artifact in research collections in the context of current information and preservation technologies. Collections can be viewed as assets, not liabilities, only if they are vital for the institutional mission (which of course includes holding such objects of intrinsic cultural value as Lincoln's Inaugural Bible). Risk assessment is premised on the notion that one must keep these collections ready for anticipated use in order to be productive.

What does productivity mean in this environment? The productivity of that Bible is not in dispute: it is valuable for display and for fund-raising. But can we measure the relationship between, say, scholarly output and use of collections? This area is quite problematic, though important, and we must follow closely the changing research habits and strategies of our primary patrons.

In the meantime, we can draw some conclusions from our years of experience as custodians of heritage assets. Let us take a closer look at those fourth, fifth, tenth editions of Huckleberry Finn. An item that has research value is usually part of a larger whole that provides context for its interpretation. Even rare items are often made more valuable by existing within a collection of like and comparable things: an incunable is made more valuable by being part of a number of similar imprints that, through study and comparison, give the first additional value. Neither does Huckleberry Finn exist in the vacuum of an exhibition case. First published in 1885 and issued in hundreds of editions since, it illustrates the point that the research value of any given item or series of items is dynamic and largely unpredictable. It also demonstrates that the research value and the preservation strategy to serve that value can be dramatically affected by new technologies. There were probably few libraries that attempted to collect and preserve all or even most editions of this book, in large part because the text was well known and easy to acquire. Few people thought that the history of the publication and dissemination of the text over time was an important topic for research. Until recently that is.

A group of enterprising individuals at the University of Virginia with an interest in Twain's text, the history of the publication and reception of the text, the changing ways that various characters were represented in illustration—Jim, for example—and any number of other topics that devolve from this book gathered a variety of editions and made them available on their Web site. They have used this digital technology to create a virtual collection of the editions of this book that, with the right mark-up, allow new avenues of inquiry into the phenomenon of Huckleberry Finn. The technology not only allows better use, but it also renders redundant so many of the fragmentary collections that abound. But let us remember that no one asked questions about reader reception forty years ago. Chances are, reception theory so fashionable now, will not be forty years hence.

I will close with one prognostication—not a particularly daring one. I believe that within twenty-five years, many if not most information resources will be created and distributed in digital form, and that, as a consequence, there will be a number of libraries that have amassed large collections of objects—books, maps, videos—that will find these collections, as information sources, inventory that is not worth saving. And yet, libraries that have amassed collections that are valuable as cultural objects, broadly construed, and not simply as information resources, will find their collections just as valuable and useful in the future as they do now, perhaps even more so. Libraries will come to have a higher profile as cultural institutions than as information depots. Therefore, our successors will judge us and the decisions that we make today on the basis of our discrimination between cultural value and informational value. We as stewards of library collections have much to learn from our colleagues in the museum community, and just as much from our colleagues in the information technology world, about how material objects and immaterial digits create and convey meaning.

<<< Previous <<< Contents>>> Next >>>

  The Library of Congress >> To Preserve and Protect
   September 15, 2008
Contact Us