Decision Making in Preservation
The preservation issue raises profound questions about which research resources are and will be most important to scholars today and in the future. Research libraries lack the funds required to preserve all the deteriorating and endangered materials in their collections. Therefore, choices about what to preserve and how to preserve it must be made. But it is far easier to achieve consensus on the general need for preservation than on the specific strategies to ensure continued access.
Although preservation needs reach far beyond the brittle books problem and the emphasis on saving content through reformatting, it is in this arena that diverse views have been most evident. Scholars and librarians agree on the goal of "rescuing our cultural heritage before it [is] lost," as Nancy Gwinn writes.22 But assessments differ on the methods. Ultimately the defining issue is selection: Which resources should be preserved and in what format?
The library community has pursued several distinct though overlapping methods to select texts for preservation microfilming. These differ in their approaches, values, and prioritizations of the tasks to be accomplished. One method deals with the preservation problem in a framework of individual institutions and the integrity of local collections. Its philosophy is that libraries do not have sufficient resources to save everything of potential value and that priorities should be assigned, according to demand, for the reformatting of deteriorated materials. In this campus-based, use-driven approach, individual items are selected because of their poor physical condition and local needs. The primary aim is to save materials most at risk within a given institution.
A second approach shares the concern that libraries do not have the resources to save all the endangered texts in their collections but holds that selection must transcend the needs of local users, taking place within a context larger than one institution. It proposes "that priority for preservation be given to unusually strong (great) collections because those collections are likely to provide the best presentation of the literature in a given field."23 The aim in this subject- or collection-based approach is to select according to a collection's comprehensiveness and strengths. A central characteristic of this approach is close collaboration among several institutions to ensure that related materials are included and that various projects do not duplicate efforts. This model has been the dominant one in the national Brittle Books Program.
A third approach does not deny the crucial role of librarians in preservation selection, but emphasizes the need to follow an intellectual blueprint based on a discipline's total literature and shaped by input from scholars in the field. This position holds that each discipline encompasses a unique universe of research resources and relies on different research patterns. The aim of this discipline-based approach is to incorporate the collective needs of the scholarly communities in the decision-making process. A number of projects in fields ranging from agriculture to classics to theology have pursued this strategy.
While these three approaches differ in their selection principles, they all reproduce the at-risk text, creating a surrogate on film. Just as views differ on selection criteria of embrittled texts, there are different perspectives on the choice of format. The decision to preserve content through microfilming has major implications for scholars. Not surprisingly, concerns have been raised about the long-term impact of preservation microfilming on scholarship. While many scholars acknowledge the need to reformat brittle print resources, they also stress the enduring value of these resources as artifacts. Selection decisions, they emphasize, need to be made in a context that recognizes the importance of retaining the original object. This position is cogently presented in the Modern Language Association (MLA) "Statement on the Significance of Primary Records,"24 and the accompanying essays by several humanities scholars in the MLA's Profession 95 articulate the important issues concerning preservation of texts whose original formats have scholarly value. The "Statement" arose from the concern that reformatting was endangering the continued availability of the physical artifacts. Central to the policy recommendations of the MLA are the vital importance of the physical features of printed texts and the consequent need to retain primary records for humanistic scholarship.
A related concern stems from the loss of local shelf access. Individual research library collections have been shaped over time by careful selection of those items of value to present and future scholars. Just as libraries are products of "a culture of print,"25 the printed scholarly literature has been indispensable to scholarship in the humanities. Reformatting can have an adverse impact on the physical integrity of collections and can also hinder browsing. Browsability is of particular importance to scholars in the humanities. Dan Hazen notes, "Research characterized by a rather fluid meander through a broad range of materials plays a major role in some humanities fields."26 In Widener Library: Voices from the Stacks, several Harvard University humanities scholars describe their research in the Widener stacks; their vivid examples highlight the crucial role of browsing, which permits a patient search from source to source in an ever-widening circle that leads to new discoveries and connections.27 Taken together, these and other descriptions of the research process of humanists have put into sharp focus the need to keep in mind, when choosing preservation strategies, the ways scholars use library materials.28
Future technological advances in digitization and network technologies will reshape the debate about decision making in preservation reformatting. We are witnessing a fundamental shift to issues in the selection of manuscripts, early printed books, monographs, and visual resources for digitization. The use of digitization technology presents many opportunities for scholarship in the humanities. For example, medievalists already have benefited significantly from the transference of widely dispersed medieval texts to electronic form. One persuasive value of digital surrogates is that they provide access to materials that are frequently impossible for most users to see because the materials are dispersed among distant libraries and archives or because they are fragile. As a recent article in the New York Times announced, "Digitized Artifacts Are Making Knowledge Available to All, On Line."29 The project highlighted in this article is the scholar-led papyrus project, which aims to build a virtual archive of 50,000 papyri from the collections of seven major research libraries. This virtual collection will allow worldwide access and enhance scholarly use of these unique materials. An even larger scale effort is the digitization of the 8,600,000 pages of documents relating to the Spanish Conquest from the collections of the Archivo General de Indias in Seville.
Remarkable as these and other projects are, digitization is very expensive; choices of content and format remain critical. As the authors of Selecting Research Collections for Digitization underscore, "We will be able to convert to electronic form only a small percentage of existing scholarly materials, and to do even that will require substantial investments."30 Not only are there complex choices of what types of material should be digitized, but there are also highly technical choices of what digitization techniques should be used. Clifford Lynch captures the complexity of digitizing special collections: "The challenge is how to convert such collections to digital format in a way that facilitates reuse and enhancement by the broad scholarly community over time—that weaves primary content into a web of commentary, criticism, scholarship, and instruction, and links it to other related content without regard to institutional or geographic boundaries, while preserving the integrity of the digitized representations."31 Only through collaboration will this vision be realized.