Selecting Materials for Digital Preservation at the National Agricultural Library

November 1, 1995

Prepared by
the Selection Task Force of the Electronic Preservation Committee
National Agricultural Library
Agricultural Research Service, USDA
Beltsville, Maryland

Selection Task Force Members:

Michael Esman
Ellen Pletsch
Jerome Rafats
Judith Zidar (chair)

The members of the Selection Task Force wish to acknowledge the invaluable contribution of previous task force chair Julie Mangin.

The present Selection Task Force was appointed in August 1995 by the Electronic Preservation Committee. The purpose was to complete the development of criteria and guidelines on the selection of materials for digital preservation at the National Agricultural Library (NAL). This is the report of the Selection Task Force.

The task force gave great weight to the recommendations in Nancy E. Gwinn's May 1993 report to the U. S. Agricultural Information Network (USAIN) entitled, A National Preservation Program for Agricultural Literature, herein referred to as "the USAIN Report." The report stressed the importance of a national, coordinated approach to preservation, and recommended that NAL take responsibility for U. S. Department of Agriculture (USDA) and other agriculturally related federal documents, as well as for pre-1862 imprints.

A tour of the stacks revealed the desperate condition of the older USDA materials located on the 12th floor. Unbound materials with yellowed, scattered, torn pages appeared to be on the point of crumbling. Bound materials fared somewhat better, but even some of these had layers of mold or fungi growing on them, and no protection was provided to adjacent materials, which will surely become infested. This collection is so deteriorated, that protective clothing and face masks will have to be required for any staff or contractors working with it! The condition of the newer USDA materials on the 7th floor and the materials in Special Collections on the 13th floor is somewhat better because some conservation work has been undertaken there.

Following these tours, the task force concluded that preservation efforts must concentrate on the USDA materials on the 12th floor, although reserving some part of the work for especially valuable pre-1862 materials. The need to preserve these materials is urgent, and doing so will make a valuable contribution to the national preservation effort, avoid the problems raised by copyright issues, and help to fulfill the mandate to make government documents accessible to the public.

NAL maintains many rare, and even unique, holdings, and the task force is aware that some of these are in need of preservation. With an unlimited budget NAL could aggressively pursue a preservation program that would encompass all its needy holdings. With a limited budget, however, the library must focus its effort in order to have any effectiveness, and that is the strategy recommended in this report.

These selection guidelines follow the preservation program as outlined in the USAIN Report. In that report, the preservation of USDA and federal documents are the charge of the Library:

"NAL is the major repository for the documents produced by the U.S. Department of Agriculture and the library has taken many steps in the past to preserve portions of this literature. NAL will assume responsibility for ensuring preservation of and access to USDA publications, the agriculture-related documents of other federal agencies (such as the Department of the Interior and the U.S. Geological Survey), and important foreign government documents in its collection." (USAIN Report, p.12.)

The USAIN Report also gives NAL responsibility for preserving pre-1862 publications. However, it recommends that these be preserved in their original form whenever possible. Therefore, in the initial stages of our digital preservation program, only USDA documents will be selected. There may, however, be individual instances when pre-1862 publications will be chosen for preservation. These decisions will be based on both the unique historical value and the deteriorated condition of the texts.

Selectors will need to have both a knowledge of the breadth of NAL's holdings and a grasp of agricultural literature in general, in order to make judgments about which materials should be preserved. These guidelines will help them with their decisions, but their subjective judgments will have to be their own. It is also strongly recommended that selectors and other staff wear protective clothing and, in some cases, face masks, to protect them from the dust, fungi, and mold that will be encountered when working with these materials.

Finally, in the interest of better utilization of available shelf space, the Selection Task Force recommends that where there are multiple paper copies of a title, unneeded paper copies be weeded from the stacks after the digital copy is created, in keeping with the Library policy on number of copies for the NAL collection (Collection Development Memorandum 1-94).

Criteria to be used in determining which publications should be preserved are listed below:

1. Select USDA documents for preservation in the following order:

1. Reports of original research
2. Statistical reports
3. Regulatory documents
4. Extension publications
5. Administrative documents

All other factors being approximately equal, USDA materials will be digitized in this order.

2. Select items that are in deteriorated condition

Determine whether the work is bound, unbound or issued in loose sheets. Other things being equal, select unbound titles or those with loose sheets, over bound titles. Check for worn appearance or roughness, faded print, mold, unpleasant odor, brittleness, discoloration, torn pages, and missing pages or pieces of pages.

For more information on evaluating deterioration, see Appendix 1, "The Eindhoven Variant: A method to survey the deterioration of archival collections."

3. Give preference to works of historical value

Assess the value of documents to the understanding of the history of agriculture in the United States. Consider whether they fill an important gap in completing the history of USDA. Determine if they contain rare illustrations worth preserving. When considering these factors, give preference to materials published by a USDA agency with research, extension, or regulatory responsibility.

4. Select materials that are in a core area of NAL's collecting responsibility

NAL uses the terms of the Research Libraries Group Conspectus to describe which subjects are most important in the collection and thus which are collected most intensively. Material being considered for preservation should be given priority based partly on the importance of the subject matter. The preservation selector should refer to two important NAL policies when considering preservation treatments.

First, and most important, are the subjects that NAL collects at the National level. These are listed in the March 10, 1986 Policy on Processing Priorities for Materials to be Added to the NAL Collection. National level collecting for any given subject involves collecting the subject comprehensively, processing material in the subject on a priority basis, providing services nationwide for other research libraries, and preserving the material dealing with the subject indefinitely. The subjects listed in the 1986 Policy on Processing Priorities as being collected at National level by NAL are:

• Animal husbandry -- production, nutrition and care of cattle, horses, hogs, poultry and other livestock
  
  
• Aquaculture -- such as economic development of aquatic resources, fish farming, fisheries technology and production, algae culture, but not whaling
  
  
• Biotechnology -- including but not limited to genetic engineering, biological control, integrated systems (biological) management, but not
  
  
• Nutrition -- foods and food supply including composition, adulteration, chemistry, preservation, storage, industry, processing, technology, packaging, inspection, dietary research, and food values. Not cookery, cookbooks or materials on dietary programs or treatment
  
  
• Plant and seed trade catalogs
  
  
• Textile and fiber plants -- plant culture
  
  
• Veterinary science -- all aspects except primatology, the human/animal bond, comparative medicine, and experimental surgery

Conspectus levels below National level are, in order of importance to a collection, Comprehensive, Research, Study, Basic, and Minimal. These levels are listed and defined further on pages 7 and 8 of the Collection Development Policy of the National Agricultural Library. This 1988 policy outlines all the subjects collected at NAL and shows the level at which each is collected. Material considered for preservation should be selected according to the importance of its subject, and NAL's Collection Development Policy should be followed closely by preservation selectors. Because of limited funding, it is recommended that only materials collected at the National, Comprehensive, and Research levels be considered for digital preservation.

5. Consider the projected amount this material will be used

Speculate whether this material is likely to be requested by a patron within the next 10 years. One use every ten years should be a starting point for projecting usefulness to patrons. Materials used less frequently than every 10 years cannot be considered "useful" when evaluating this factor, although this does not necessarily mean the materials are not valuable. This is not something the selectors should spend a great deal of time wrestling with. Rather, it is meant to be a "best guess" on their part, and is recognized as being only that. If in doubt, the selectors should not checkmark this item on the Worksheet.

6. Give preference to materials unique to NAL or uniquely preserved by NAL

Determine if the work is owned by another institution. Determine if the work has been preserved by another institution either in electronic or microform format. This will be difficult information to obtain, but a reasonable effort should be made, so as to avoid duplicate preservation of a text. Check OCLC, RLIN, the National Register of Microform Masters, and the Internet. Web sites and Listservs may identify preservation programs. It may be useful to search Cornell's on-line catalog. Be sure to check also to see if NAL has already digitized or microfilmed the work.

There may be instances where NAL will choose to preserve a document that is already preserved by other institutions. These may include items such as USDA core publications for which NAL has special responsibility, high use materials in deteriorating condition, or items of particular value to NAL. Decisions on these items should be made on a case-by-case basis.

Electronic Preservation. The purpose of these Guidelines is the selection of materials for electronic preservation, but a consideration of microfilming is still in order. A FLICC symposium entitled, "The Great Preservation Debate: To Digitize or to Microfilm", was presented at the National Archives on September 20, 1995. Although no full consensus was reached, there was general agreement that (1) microfilming is cheaper than digitizing, in the long run; (2) digitizing provides better access and cheaper distribution; (3) it is easier and cheaper to scan older materials from microfilm than from the paper source; (4) digital information is the wave of the future.

The approach recommended at the symposium was to microfilm low-demand materials, and digitize high-demand materials. This would preserve the low-demand materials, but at a lower cost than it would take to digitize them. If the demand for them changes in the future, they could be digitized from the microfilm. High-demand materials can be digitized directly, and the lower cost to access and distribute them would actually save money in the long run. NAL does not keep statistics on the demand-rate of its materials, so this approach could be hard to implement here. Furthermore, older materials that are likely to need preservation cannot be tracked for usage at all, since most of them are not now in ISIS.

The Selection Guidelines in Section II recommend a concentration on USDA documents. An effort is underway throughout the Government to make such materials readily accessible to the public, and accessibility is best and most economically realized through digital means. This is why we are recommending that these particular materials be preserved electronically, even though many of them may not be in the high-demand category. Preservation funds are extremely limited. We therefore do not recommend that materials already preserved through microfilm be electronically preserved as well, even if they fall within this category.

Current Microfilming Efforts. NAL's Document Delivery Services Branch has an ongoing microfilming program. Their selection criteria are:

1. A patron has requested the item; and
2. The item is seriously deteriorated, possibly in a "last use" condition; and
3. The item is considered to add value to the NAL collection.

When these criteria are met, the item is selected for microfilming. A master and two working copies are produced, and one of those working copies goes to the patron. The master microfilm is stored in the vault, and the working copy is used for reproducing copies whenever the item is requested again. If the selected item is part of a series or set, the entire run might be microfilmed, rather than just the one item. This is a very practical, though ad hoc, approach to preservation, which is based on the concept that we need to preserve what our customers need. It also assumes that the item will be requested again, justifying the cost of microfilming.

By designing selection criteria for choosing materials, we hope to take a more strategic approach to preservation, selecting materials to meet the needs and responsibilities of the Library as a member of the agricultural community. We also expect this more strategic approach to enable us to free up additional shelf space, by digitizing long runs of related documents, and discarding one or more paper copies of the items. We are not recommending that DDSB discontinue its microfilming effort. With limited funding, we do not foresee any other way to fill the need it serves. We do recommend a cooperative effort between DDSB and the EPC to ensure there is no duplication of effort and that both teams are kept informed of each other's activities. If DDSB selects an item for microfilming that falls within the selection criteria for electronic preservation, we recommend that the item be digitized if possible. However, the staff may want to make this decision on a case-by-case basis.

Artifactual Value. Items identified as have artifactual value, such as historical papers, rare color prints, first edition books, or works of art, should be preserved in their original form if possible. However, these items may still be selected for digital preservation if they fall within the selection guidelines. Digitization will make them more accessible, as well as decreasing the likelihood that the documents themselves will have to be handled in the future.

One of the key missing ingredients we see in the plans for electronic preservation is an archiving system that is secure and accessible, and can incorporate a long-range refreshment and migration strategy. The system will need a team of experts to run and maintain it, in the same way that such a team runs and maintains ISIS. This system will have to become integral to the function of the library, for it will hold, in some cases, the only copy of important materials.

The design of the system is likewise critical. A system that is integrated with the data capture system will enable NAL to make the best use of its limited preservation funds and staff. Such a system would accept documents as they are scanned in, without further conversion; accept many different types of data, including full text, images, sound, and video; permit in house design of the data tracking portion of the system so that it best serves our retrieval needs; is available over a network and can be accessible via the internet; and is easy and intuitive to access, even for novice users, or can be linked to ISIS for direct access to materials from the cataloging and indexing records.

An Archiving Committee has been formed to study this critical component and recommend a solution. This is not an easy task, but we believe such a system is possible and affordable.

A further component of this system is a national reporting mechanism that can assist in tracking digitizing programs and the materials digitized. NAL should work cooperatively with other institutions to implement a "National Register of Digitized Materials", that would work much like the National Register of Microform Masters does for microfilming programs. This will avoid duplication of effort, and keep track of who is preserving what.

Attached to this report is a one-page summary of the selection guidelines in the form of a Selection Criteria Worksheet. The worksheet is designed as a check-off form, so that staff who perform the selection of materials can use it to describe and prioritize those materials.

On the worksheet, the description portion at the beginning of the worksheet, along with Sections 1 and 2, should be completed in the stacks with the item in hand. If the item is a USDA publication, the publication type should be noted, and this will be used for priority ranking of the materials. Materials with a higher rank will be digitized first, provided the point totals in Sections 1, 2 and 3 are approximately equal (within 5 points). After describing the document or series as indicated in "Description," the number of copies should be noted in the space provided. The subjects listed in Section 2 of the worksheet, under "Comprehensive" and "Research" levels of collection, are intended as a guide, and are not all-inclusive. For a complete treatment of subject coverage at NAL, consult the Policy on Processing Priorities and the Collection Development Policy. Section 3, Uniqueness, is intended to be filled in later, after checking OCLC, RLIN, the Register of Microform Masters, and other tracking tools.

After all three sections have been completed, the TOTAL can be computed. Count each checkmark in Sections 1, 2, and 3 as one point, except for the Subject values, where NATIONAL counts as 5 points and COMPREHENSIVE counts as 3 points. If more that one subject category is checked, only the highest value should be counted toward the total.

At the end of the form is an entry for "Has Artifactual Value". Artifactual value should be determined in the stacks with the item in hand, and this line should be checkmarked only if the item should be preserved in its original form. A check here does not count toward the total points, but rather, is an independent parameter. If the item should and can be preserved in its original form, we may choose to do that rather than digitize it, unless its physical condition is desperate.