The Abbey Newsletter

Volume 15, Number 4
Jul 1991

Preservation Needs in Science

All text under each discipline heading, except that enclosed in brackets, is quoted selectively from the pages of the RLG publication described in the second paragraph, and reprinted here by permission of the Research Libraries Group.

The Research Libraries Group (RLG), a consortium of over 100 major universities and research institutions in the U.S., was formed in 1974 to share cataloging and other information through MIN, the Research Libraries Information Network. It has greatly expanded its function since then. One of its four principal programs is Preservation, which in 1990 included management of four microfilming projects (sponsored in large part by NEH), and development of a collection-based preservation needs assessment tool.

To help plan where to focus future efforts, RLG recently surveyed information needs in all academic disciplines(the humanities, social sciences and sciences), and published reports in 1988, 1989 and 1991. The report on the sciences is Information Needs in the Sciences: An Assessment, by Constance C. Gould and Karla Pearce (79 pages, acid-free paper, first copy available without charge from RLG, 1200 Villa St., Mountain View, CA 94041-1100, 415/691-2236). Preservation needs (or absence of needs) in the various science disciplines and the importance of access to the older literature are discussed on 20 of its 79 pages.

The science report was prepared as part of RLG's Program for Research Information Management (PRIMA), which is identifying the types of information researchers need better access to, and encouraging development of new data resources. Over 130 teachers, researchers and others closely connected with research were consulted or interviewed in its preparation. Although the report shows that an amazing amount of hard work was done to make it as complete and accurate as possible, the authors emphasize that it is only a preliminary needs assessment.

The reality of preservation needs in science is not often acknowledged, even by scientists (AN, April 1990, p. 30), much less described in print. So when a report like this comes along, it deserves to be made known in the preservation community, especially among those who will be choosing which materials to preserve, and by which method. Permission has been obtained to reprint the extracts from the report that relate to preservation. Sometime adjacent text has been included to provide a context. For the present purpose, preservation is broadly defined: it includes everything from going out and getting the material (saving it from the dump), to ranking a copy of it, to conserving the original.

Since the report is organized first by scientific discipline, then by major and minor headings, perhaps an overview, would be helpful here, although the headings will not be transcribed into the text. Disciplines covered are:

Computer Science

Only computer science is described as having no preservation needs and making no use of the older literature.

Major headings under the disciplines are:
General Characteristics
Use of Information [in the discipline]
Primary Literature
Major Indexing & Abstracting Services
Current Awareness
Data Collections
Handbooks & Compilations
Other Electronic Resources
Future Needs & Directions

Not all major headings were applicable to all disciplines. Preservation was mentioned under every heading but "Current Awareness" and "Handbooks & Compilations."


An important segment of monographic literature--books published after the mid-19th century is deteriorating rapidly because it is printed on high acid paper. Several research libraries have undertaken to microfilm volumes in the history of science, including physics. However, because of the widespread misconception that research scientists do not need older materials, preservation microfilming of these books in general receives a law priority....

Increased attention should be given to the preservation of deteriorating physics monographs.


Chemical information is distinctive in that it deals with atomic and molecular species, which are precisely and unambiguously defined by their molecular structure. Since chemical and physical properties do not change over tine, older literature is as essential as current literature.... Fast access to current literature certainly ranks first among chemists' information needs, but access to retrospective literature is probably a close second.... It is important to ensure access to older chemistry journals that are deteriorating. Generally, preservation of the intellectual content rather than the physical document will suffice.


Like research in other science disciplines, research in biology is impossible without reference to previous research....

[Six electronic indexes are briefly described and evaluated.] All of the above are valuable tools; one drawback noted by researchers, which reflects their dependence on older literature as well as current, is the lack of retrospective coverage in the online indexes .... Formerly, articles consisted of text accompanied by line drawings or photographs.... Now, a smaller quantity of text is accompanied by a large quantity of formulae, charts, tables, color photographs, and complex illustrations.... Research reports published before the mid-1960s will continue to be important....


Geoscience has a strongly historical orientation. Old observational data are still useful even if in many cases the methods of observation or analysis have changed.... The sheer number of titles alone required for a research collection poses a challenge. Further, by comparison with other disciplines, geoscience materials are very diverse with respect to date and place of publication, subject, the variety of formats, and types of information sources....

Preservation is also a concern: many [maps] are printed on acid paper, which deteriorates over time. In addition, they often suffer wear and tear when taken into the field. Because maps are generally not bound, they tend to be pilfered from library collections. Finally, computer-generated maps may be lost when they are not saved in hard copy form or when they are saved on magnetic media that may become obsolete.... Geologists especially value guidebooks, theses, and dissertations for their detailed descriptions and analyses of outcrops and features-in effect, the original data sets of geoscience research.... older serials, maps, reports, guidebooks, and theses contain valuable observations. Even if the earlier analyses are no longer valid, the facts these works contain retain their value. Egyptian flood control reports from the 1930s, for example, are used to provide data for dam construction; Soviet geologic surveys dating from the 1930s, because they describe natural seismic activity in the years preceding the development of the atomic bomb), can be used by scientists to derive a baseline for the purpose of nuclear test monitoring. But these materials are seldom catalogued in a bibliographic utility, thus limiting the likelihood they will be used by anyone other than local researchers. Also, many older materials have seriously deteriorated. In addition to better bibliographic control, preservation of older materials, preferably through some means of full-size facsimile reproduction rather than microform, needs attention....

Much of the currently generated information is now produced in machine-readable form. But the bewildering array of data being collected presents problems of its own. Faced with a torrent of information, researchers are equipped with few means to determine if the information they need exists and, if it does, where it can be located. For instance, researchers now know that the hole in the Antarctic ozone layer discovered in 1986 was in fact recorded ten years earlier by a satellite, but the tapes had lain mused in a vast NASA archives

The problem promises to become more acute. According to recent estimates, spacecraft will produce between 1990 and 1995 as much information as they did in the previous 20 years. As the quantity of information grows, questions and difficulties surrounding the management of it also mount....

Future needs and directions: ... Cooperative preservation of maps and selected older geoscience publications printed on deteriorating paper....


Astronomy data, largely graphic and numeric rather than textual, retain their value indefinitely. Observers report phenomena with the highest possible degree of accuracy, with the knowledge that other researchers will base theories on them. In fact, many observations recorded in the Middle Ages are as useful as those made yesterday.... Since some predictions [astronomers] make are based on old information, the historical literature is essential for in-depth research....

Older astronomy literature contains information that is valuable and used today. Many collections of older literature exist and are most often owned by observatories established in the 19th century or earlier. For example, the library at the Naval Observatory in Washington, D.C., holds about 75,000 volumes covering the histories of astronomy, astrometry, and navigation and reports from other great observatories worldwide. Unfortunately, some of the older astronomy literature exists in unique copies printed on high-acid paper, which is in danger of disintegrating....

To understand and identify new phenomenon theorists must have access to seemingly limitless amounts of observational data, regardless of when they are recorded. Often, later astronomers examine and explain earlier observations and reinterpret them. For instance, radio frequency radiation from space was first recorded in 1932 but not understood. It was not until 1945 that a radio telescope was built, allowing study of the radio waves and their sources and explanation of the phenomenon.... Star catalogs have long been one of the staples of astronomy research. Old bright star catalogs are still as useful as when they were first produced, since, as one astronomer pointed out, "the bright stars are still in the saw place."...

Photographic plates are an important resource, especially for observations of single-occurrence events, but are quite old and deteriorating rapidly. They are expensive to reproduce, but fortunately it is now possible to put them in electronic format, which will preserve the intellectual content if not the plate itself. In fact, the photographic plates of several big surveys such as the Palamar Sky Survey have recently been digitized and will be available to astronomers in the near future. Although the magnitude is still too "coarse" for comparing any but the brighter stars, the electronic format allows for correlating them with objects that emit radio waves It has been estimated that NASA has created more than 1.2 million magnetic tapes since the agency's creation in 1958; many of these contain information of great potential value to astronomers. Yet much of the information gathered has been so poorly labeled and stored that it is all but-or even entirely-inaccessible. The problem are familiar: code books describing how to use the tapes have been lost, tapes have been physically , and the hardware necessary to read older tapes no longer functions. Most significant, no standard method for describing and cataloging the tapes exists.... Projects to preserve deteriorating photographic p and print collections, essential resources for astronomy research, are needed.


Older data are often useful. When petroleum became more expensive in the mid-1970s, for example, research in other fossil fuels going as far back as the 1920s was heavily consulted.... Engineers use virtually all forms of published material. Manuals and handbooks for project design, books for background material, journals and symposia for both current and older information, computerized data and indexes-all are widely used and valuable sources of information .... A technical report may contain the germ of an idea 30 years old that can be explained with a new theory. It may describe a manufacturing process that has not changed substantially in 20 years. It may establish scientific precedence. Consequently, older technical reports are often called for but require more ingenuity to track dawn than recent ones.


Knowledge in mathematics is cumulative, and the different areas-e.g., algebra, geometry, calculus-are closely intertwined. Mathematical literature retains its value over a long period of tire, and mathematicians frequently make use of the core literature.... Nearly all significant published information is in print format.... Serials and monographs, new and old, are absolutely vital to the research of most mathematicians. Active use of old serials, some dating from the early 19th century, is common. Use of older monographs is also frequent.... Librarians point out that one or two locations for a monograph do not adequately support research use and that the addition of information about other important mathematics collections would enhance research access.

In the case of older materials, preservation issues affect access. At sow institutions, monographic series and serials have been microfilmed for preservation purposes. Users, however, object to the inconvenience of microfilm, especially for monographs. It is more suitable for serials, since a single article can be printed without great expense.... A corollary of the collection situation is that, increasingly, mathematicians will be required to go beyond the department library and to rely on libraries at other institutions for the information they need. An efficient and affordable document transmission capability will be necessary to support their needs....

Hard copy reformatting (through photocopy) of older monographs is the preferred way to provide access in many libraries. Since photocopying is not coordinated among libraries, this practice has led to duplication of effort. This is generally not the case with preservation microform, which are produced only after a careful search for records of existing microforms is made. Cooperative preservation projects, in conjunction with retrospective conversion of card catalog records of important older collections, would improve access to older materials.

Because the discipline and the technology on which it depends are changing so rapidly, computer scientists use retrospective literature infrequently. According to one researcher, most computer science literature is obsolete after five years. Thus, computer scientists are on the whole less dependent on access to the historical archives of the research library than other scientists.


Unpublished research: Just as journal literature has grown, so has the volume of unpublished information such as preprints, technical reports, and conference papers, important to certain science disciplines and subdisciplines. These materials present problem of their own: knowledge of them tends to be confined to informal networks; they are difficult to identify, collect, and manage; and access is as haphazard as the management of them....

Primary data: Scientists must cope with scientific data that existed in a much smaller quantity or did not exist at all 20 years ago. Computers and increasingly sophisticated instruments such as linear accelerators and remote-sensing devices produce and store enormous quantities of data. Much of this information, although often produced at great expense and useful to a broad audience, is poorly managed and consequently under-utilized….

Areas of opportunity: … Preserving important older materials that have deteriorated.

 [Contents]  [Search]  [Abbey]

[Search all CoOL documents]