Preservation and Digitization - Natural Partners?

IFLA Standing Committee on Preservation and Conservation IFLA, Boston August 23, 2001

Summary of Survey Results

In preparation for the workshop, "Preservation and Digitization - Natural Partners?" taking place at the IFLA meeting, Boston, 2001, the IFLA Standing Committee on Preservation and Conservation sent a survey to about 100 IFLA members who appeared to be digitizing portions of their collections and who might be sending a representative to the conference. Twenty-one institutions completed the survey, with representation from thirteen countries and including nine national libraries, eight academic libraries, two public libraries, and two consortia. Program planners designed the survey to elicit responses about the philosophy, management, selection, organization, and funding of library-based digital programs and how they relate to the library's preservation activities--if they do. The workshop is intended to explore these issues further through an interchange of information and ideas among the respondents and other workshop attendees.

The following is a summary of conclusions drawn from the survey results. The compilers intend this information to be suggestive, rather than conclusive, an indication of the range of programs that have begun and the status of institutional involvement and experience with them.

Survey Respondents

Question 1: WHY does your library produce digital library products? What prompted you to start a digital library program? What are your goals?

Several libraries were prompted to begin digitizing programs initially by a desire to explore the technology, gain experience with it and figure out what it could do; only later did the libraries begin to outline a purpose, strategy, or goal. Two libraries were further stimulated from the outside. In France, for example, President Mitterand began to promote a new national library program that included digitization of materials related to French culture for remote access. At the American University in Beirut, the university administration called for digital contributions to the university's website from all university units. In 1995, the Library of Congress publicized a goal of making available 5 million "historically significant digital items" in five years. After a highly successful fund-raising campaign, the Library met the goal in 2000. Theirs is perhaps the most aggressive program in the world.

Several libraries mentioned preservation as one of several stimulants, but none accorded it first place. Although there is hope that by making digitized materials more accessible, wear and tear due to handling can be lessened, none of the respondents offered any evidence that this has actually happened. The National Library, Oslo perhaps summed up the experience of most libraries whose programs have somewhat matured:

In general, all respondents reported a focus on collections or items that are primary sources, or are rare, fragile, unique, or a special strength. It is not surprising that national libraries in Portugal, France, the Netherlands, New Zealand, and the Library of Congress, as well as some smaller institutions, feel a responsibility to make collections about their nation's history and culture more widely available.

The responses show that libraries uniformly see digitization primarily as a vehicle for expanding their services by being able to present collections to a broader--and distant-- audience. Most focus on primary source materials—rare books, manuscripts, photographic collections, historical materials, archives. Many items are of principal interest to scholars, but all see this activity as a potential contribution to elementary and high school education and lifelong learning.

Another comment from the National Library of Norway in Oslo describes the experience of several respondents:

Question 2: WHAT collection materials is your library digitizing? Describe the types of materials your library has chosen to digitize. Are they single items or collections? Are they organized in any special way?

To summarize briefly the answers to this question would be an injustice to the libraries that are well along in their programs and digitizing a variety of materials. Consequently, an appendix to this survey contains the answers of all respondents to this question, along with the addresses of their web sites. However, some trends can be identified.

Detailed Responses

Nearly all respondents reported a concentration on the rare or special items in their collections, at least as a starting point. The national libraries in Portugal, Norway, France, the Netherlands, Russia, Norway, Korea, the United States and New Zealand quite logically have a special focus on items that display their nation's cultural heritage. The British Library's emphasis has been on its major treasures, such as the Gutenberg Bible, the Lindisfarne Gospels, and the Electronic Beowulf. Other libraries reported similar concentrations: Trinity College, Dublin, on Irish treasurers and manuscripts, Kyoto University Library on the national treasures and important cultural assets of Japan, the Jafet Library at the American University of Beirut on collections related to Lebanon, the University of Toronto Library on Canadiana. The Library of Congress has more than 90 collections in its American Memory Historical Collection, but has also begun to digitize items from its foreign language collections.

Both rare individual items and thematic collections predominate among selections, along with much 19^th century material (books, newspapers, journals). Thematic collections most often are of non-book formats, such as photographs, manuscripts, maps, postcards, newspapers, posters, engineering technical papers, broadsides, archival materials, parliamentary documents. The emphasis is still on paper-based and visual materials; only three libraries mentioned recorded sound, oral histories, or moving pictures. The Smithsonian Institution Libraries is digitizing collections of trade catalogs, primary documents that illustrate the history of technology and manufacturing. The first collection covers sewing machine trade catalogs, the earliest dating from the mid 19^th century, from 228 companies. The British Library reported a new project to create "a collection of learning materials on the theme of a national sense of place" for a website called "In Place," which bears more investigation.

Although the responses were not detailed enough to determine the extent, exhibitions are clearly another stimulus for digitizing projects. Online exhibitions often accompany physical ones. A trip to the appropriate website is required to determine whether the online exhibitions contain complete copies of items displayed, as opposed to single or specific images.

Two libraries mentioned production of books on demand. The University of Toronto is doing this in partnership with a Japanese Book Company (13,000 out-of-print monographs scanned), and the National Library of New Zealand inaugurated its digitizing program with an on-demand project called Timeframes.

It is clear that digitizing projects provide opportunities for collaboration. For example, the Russian State Library and the Library of Congress are conducting a bilingual, multimedia English-Russian digital library called "Meeting of Frontiers," to tell the story of the American exploration and settlement of the West, the parallel exploration and settlement of Siberia and the Russian Far East, and the meeting of the Russian-American frontier in Alaska and the Pacific Northwest. The Bibliotheque Nationale and the New York Public Library organized an exhibition called Utopia, which stimulated digitizing efforts on both sides. The technology has great appeal for consortia as well. The Colorado Digitization Project pulls related collections from more than one institution; in one example, two institutions worked together on documents and photographs about the only World War II Japanese Internment Camp in Colorado. The Research Libraries Group, through its Cultural Materials Initiative, is creating a broadly structured digital repository for collection of documents, images, sound, motion and objects that will become available later this year.

Question 3: HOW are the digitized items selected? Who makes the decisions about what is digitized? How much is the choice dependent on funding?

With few exceptions, most libraries appointed a committee, task force, or cross-representational group to direct the process of selecting the items or collections to be digitized. In some cases, the group sought proposals from library staff, and then evaluated the proposals according to criteria established by the group. Most selections are of items not covered by copyright.

Criteria offered by two respondents offer different perspectives and are quite instructive. The Smithsonian Institution Libraries’ evaluation criteria assumes that there is an item/collection proposed for digitization, then focuses on practical aspects of selection and the relation of the work to the subject area, basically asking the questions, "can it be done physically and technically?" and "in the presentation, could some value be added to the source item?" Doing so creates a new version of the item, a "digital edition."

1. Will the item/collection complement other digital resources?
2. Will the product have more value than the source material because of indexing, links, finding aid, availability, or other features?
3. Will multiple audiences find it useful?
4. Does the original contain unique materials never before generally available?
5. Is content accurate, representative, insightful, and/or useful for research?
6. Is the item/collection relevant to current scholarly concerns?
7. Is there existing secondary or complementary literature relevant to the collection? Can this be made available on the web to enhance the texts?
8. Will the physical condition of the item/collection allow for handling and scanning?
9. Will wear and tear on the original be reduced if the item is available digitally?
10. What type of metadata will this product require?

On the other hand, the Digitisation Principles of the National Library of New Zealand establish a priority area and focus on the philosophical underpinnings of the work.

Few respondents mentioned involving persons other than library staff in the selection process. One exception is the Kyoto University Library, where the Special Committee for a Digital Library, consisting of twenty-five professors, selects the items. At the University of Toronto Library, faculty and subject specialist input were the prime movers in selecting some items. The Biblioteca Nacional in Portugal established a scientific commission, comprising scholars, librarians, and web designers. When its own Development fund is the source for support, the Bodleian Library, University of Oxford, proposes to use "academics who, with our help, propose digitization projects, funding for which will be awarded by an editorial board." After some experience at making internal decisions, the Toronto Public Library is "about to initiate a new round of electronic resource priority-setting, which will be a joint process undertaken with that of our virtual Reference Library and will involve staff, public and stakeholder consultation." The Smithsonian Libraries set up a focus group of natural history scientists and curators to make suggestions on items to digitize that would be useful for research as well as visually attractive.

Preservation concerns of respondents centered on the condition of the items proposed for digitization and whether they could withstand the process safely, rather than selection just for the purpose of preservation.

Almost every institution reported that the availability of external funds plays an important role in the process, although a few implied there were also internal funds available. The fortunate National Library of Korea said, "All funds for digitization are appropriated by the government." The Toronto Public Library provides a fee-based service to the public, from which it likely derives some revenue. After raising $50 million to support its American Memory project, the Library of Congress successfully argued for federal funding to continue its program.

With the exception of the Library of Congress, which has such an enormous mass of material available, most libraries are still searching for the key to selecting the most appropriate and useful items for digitizing. There has been little feedback from users to help guide the choices or involvement from potential users in the selection process. This reflects the still experimental nature of many programs. The survey did not inquire whether the advent of commercial companies into providing digitized content has had an effect on selection decisions, but this would be an interesting question to pursue.

Question 4: HOW is the work organized at your library? Describe the basic structure of the work at your library. How have you organized library staff to handle this new kind of work? Have personnel changes been required, and if so, what were they? Were new positions or a separate department created? Is all the work done in-house? Do you use contracts with outside vendors, and if so, for what purpose?

Survey respondents were divided between those who organized the work within existing structures, such as the National Library of Korea, the National Library, Oslo, and the University of Delaware, and those who significantly restructured to do the work, such as Bibliotheque Nationale de France, Koninklijke Bibliotheek, and Kyoto University Library. The Library of Congress and the Russian State Library set up new departments and hired new staff. However, no library reported a new program on the scale of the Library of Congress, which at its peak had more than 100 professional staff responsible for a range of tasks that included content conversion and production, rights management, web design, repository development, and contract management.

All respondents indicated that the work involved staff and skills from many existing departments whether or not a new unit was created. These include Preservation, Research and Development, Information Technology, Collections Management, and Special Collections. Respondents such as the National Library of Korea rely heavily on contractors, primarily for scanning; others such as Trinity College, Dublin do all the work in-house. The Russian State Library has a contract with a "closed joint stock company" called "Electronic Book" to digitize some of the library’s holdings. The Colorado Digitization Project has set up a centralized unit for training staff and volunteers of smaller institutions in digitization methods.

Finding staff with the appropriate skills and experience or training existing staff is a concern for several respondents. Skills needed for a full digitization program include subject expertise, cataloging, preservation analysis, conservation treatment, information technology/computer expertise, web design, photography, rights management, contract management, and digital archiving. This is quite a demanding "laundry list" for a relatively new library program.

The responses indicate that libraries are forced to reinvent themselves in order to produce and deliver digital content. Existing organization units have to take on new work or new units must be created. Traditional roles are being challenged. Library staff are acquiring new skill sets whether through training or hiring of new people. In reading all the responses to this question together, there is a sense of increased flexibility entering the management structures and practices of libraries. A picture also emerges of small-scale pilot projects growing very quickly into mainline library programs with hard-to-estimate resource demands.

Question 5: HOW is your program funded? What are the sources of funding for your program? How does the digital library program affect the rest of your budget? Do you think your program is on a stable financial base? If so, how have you achieved this? If not, what are your plans for acquiring resources?

For most respondents, funding for digital library projects is not stable. Some received special grant funding. Most respondents receive their funding from national funding agencies rather than from private or corporate donations. A few libraries reallocated money from existing programs to digitizing. For example, the University of Toronto Library reduced its microfilming program and transferred funds to support digitizing. This allows a certain degree of stability. Encouraging exceptions are the Bibliotheque Nationale de France, which reports, "the digitization budget is now as stable as the classical acquisitions one in the whole library budget allocated by the Ministry of Culture," and the Library of Congress, which managed to raise over $45 million in private funding for its program and just received federal money to continue it.

The Koninklijke Bibliotheek reports, "it’s hard to convince potential funding agencies that large investments are needed to broaden access to our cultural heritage. They seem to be more prepared to invest in ‘innovation’ or ‘enhancement of economic infrastructures’ then in ‘basic’ digitisation of collections." We suspect this problem is common to many libraries as "basic" digitization moves from being "cutting edge" and experimental to mainstream, routine, and part of the baseline operation expected of a modern library.

Linda Hall Library and the Smithsonian Institution Libraries are exploring cost recovery efforts through spin-off products from digital images, although there is no expectation that this will cover the full costs of digital production.

In general, digital library programs encourage and enforce a type of institutional nimbleness in their host libraries. Old funding models are revised or called into question. Libraries try new sources and strategies as they respond to rapidly changing technological capabilities, new skills required of staff, and volatile economic environments. Digital library programs also require rethinking and, in many cases, redefining what are and what are not "core" library services.

Question 6: WHO is using your digitized materials? Can you identify the users of your digitized products? If so, how? Do you have statistics on use or other kinds of feedback? Did you identify a target audience for your products, and if so, what was it, and has that audience been reached?

Almost all respondents use standard web logging tools to count numbers of hits and sessions. However, these tools are not suitable for determining "who" is using digitized materials beyond very broad categories such as country of origin. Respondents rely on emails, requests, faxes, subsequent requests for related materials, testimonials, and organized review groups. Some libraries such as the Bibliothèque nationale de France have sent out questionnaires or conducted surveys to get a better understanding of their users. The National Library, Oslo, has sufficient evidence for some of its products to determine that it is reaching a selected audience. In general, national libraries or libraries with a national scope have an intended broad audience of national users. Academic libraries look to satisfy their campus audiences.

The Library of Congress has programs inviting educators to the Library to facilitate and encourage the use of the digitized content in curricula of schools in America. Its Learning Page "is designed for teachers, school librarians, students, and life-long learners. It has tools to help users navigate the American Memory Historical Collections, activities, lesson ideas, teacher-created lesson plans, and other information to help guide educators and their students in using the American Memory Historical Collections' primary sources in the classroom."

Many respondents express a desire for more accurate, more targeted usage data in order to guide selection and design for new products and to better justify requests for funding. Much work and analysis needs to be done in this area.

Question 7: Do you have a preservation program at your library? If so, are your preservation and digitizing activities related? Briefly describe your preservation program and how it relates to your digitizing activities. Is digitizing considered a preservation medium for your library? Do preservation staff have a role, and if so, please describe.

No respondent views current digitizing as a preservation medium. All respondents have some form of a preservation program. Most respondents involve preservation staff in their digital library program though the extent of their participation varies significantly. In a few libraries, such as the University of Delaware and the Smithsonian Institution, the in-house digitizing facilities are staffed by preservation staff. Other libraries follow the Toronto Public Library pattern of using preservation staff in document preparation, collation, and review. Many also use them to advise on handling issues in the selection process. Most respondents view digitizing projects as a way to decrease handling of rare or delicate originals and increase access to users, though there is little direct evidence of this.

Almost all respondents viewed preservation as a core library program. Activities in the respondents’ preservation programs include microfilming, collection assessment, research (e.g., effects of copying on paper), environmental assessment, and conservation treatment, including but not limited to deacidification, training, binding, and repair.

Many respondents are exploring preservation of digital surrogates or natively digital materials and express a need for more research in this area. The U.S. Congress has requested the Library of Congress to lead the development of a National Digital Information Infrastructure and Preservation Program to ensure the long-term availability of "born digital" materials. The Library plans to consult with federal partners to assess joint planning considerations for shared responsibilities and to seek participation from the non-federal sector. Its overall strategy is to work "in cooperation with the library, creative, publishing, technology and copyright communities in this country and abroad."

Any form of reformatting library collections raises preservation issues. Digitizing provides no simple solutions here but requires significant input and guidance from preservation staff to be done properly. While the digital promise of near-perfect copying and near-effortless media transfer seem to offer a future where traditional preservation concerns will disappear, the reality is quite different. How can we authenticate what is in digital format? What hard evidence for accuracy of reformatted materials do we have? What are the long-term environmental effects on existing media? Even in the highly technical areas of preserving digital surrogates and natively digital materials, preservation staff bring needed questions and experience that can illuminate the issues.