Condition Survey of the United States History, Local History and Genealogy Collection of the New York Public Library
The New York Public Library (NYPL) has been deeply concerned for a long time with the deterioration of its collections. Several attempts were made over the years to gain some sort of quantitative information on the size and nature of the preservation problems. In 1982, Wesley Boomgaarden surveyed NYPL's World War I and World War II collections. It demonstrated their severely deteriorated condition and assisted the library in obtaining funds to treat that collection. In January 1983, the Conservation Division accepted James Wellvang, a student enrolled in the Columbia University School of Library Science's Preservation Administration Program, to do a field study project during the spring term. Mr. Wellvang and I continued refining a methodology for surveying the collections.
Past surveys both at NYPL and at other libraries were researched. One important survey had not been reported in the literature. That was the recently completed Yale University Libraries' survey under the direction of Gay Walker and Jane Greenfield. They kindly agreed to explain their techniques and methodology to us. That survey, which has since been published, was the most ambitious and thorough condition survey yet attempted. Their three-year effort resulted in the development of procedures and methods which could be adapted for use at NYPL. Because of the meticulous attention to detail and careful documentation that had gone into the Yale survey, the staff of that institution gave us a great deal of assistance in adapting their techniques to the survey we were planning. In addition, Jeffrey Simonoff, one of the statisticians who had worked with Yale on the survey, worked with us on the project.
One important fact learned from the Yale staff was that a test or pilot survey was necessary to assess methods and procedures. John Baker, chief of the conservation division, decided to proceed with a pilot survey. Because of the large number of subject areas concentrated in the United States Local History and Genealogy (USLHG) Division, for which the research libraries of NYPL had accepted primary collecting responsibilities under the RLG Collection Management and Development Program, that division was chosen as the subject of the pilot survey.
The material surveyed included only bound volumes and those items shelved with bound volumes (packages of unbound serials, envelopes containing pamphlets, etc.). Other material held by the division, such as postcards, glass-plate negatives, and local views, was not included. The number of items in the surveyed portion of the collection totaled 143,253.
Though this was only a relatively small proportion of the library's holdings, it was still a very large number of objects to examine individually; fortunately, that was not necessary. If randomly selected, a much smaller number of samples would give an accurate picture of the condition of the entire collection. Obtaining a completely random sample is an important and difficult task. The extremely complex and diverse nature of the Yale collections required the Yale surveyors to use more sophisticated sampling techniques than any that had been used in library condition surveys reported in the literature. In collection condition surveys, the physical volume must be the basic unit surveyed. The shelflist cannot be used because it does not correlate exactly to physical volumes: one catalog entry may represent more than one volume and, conversely, one volume may be represented by more than one card. Consequently, a sampling method which selects samples directly from the shelf must be used to allow each physical volume in the surveyed collection an equal chance of being selected and evaluated. To do this the collections are separated into areas (or, as Yale called them, "strata"). These are simply portions of the collection in which the shelving arrangement is virtually identical. This permits numbers to be assigned to the physical elements such as aisles, sections, shelves, and volumes. Random numbers can then be used to identify any quantity of volumes within that area.
The USLHG collection was located in two distinct areas, the seventh level of the main stacks and the division's reading room. The stacks are divided into quadrants in which the rows of shelves (at NYPL called "presses") were essentially all the same. Each row contained ten three-foot sections of shelves with generally six shelves per section. When surveying in the stacks the surveyors were given sets of five random numbers. The first identified the quadrant; the second, the row; the next, the section; the fourth, the shelf; and the last, the number of volumes to count from the left to select the sample volume. In this manner each volume in the area had an equal chance at being chosen.
Each of the 945 volumes sampled was evaluated by means of a questionnaire. The questionnaire asked the surveyor to identify the place and date of publication, the number of shelf inches occupied by multivolume titles, and the total number of volumes on the shelf from which the sample came. It went on to ask seventeen questions about the physical nature and condition of the volume sampled. The Yale survey had used a number of terms and definitions that were used in this pilot survey. Several questions that were less important or not relevant to NYPL were dropped. Some questions were added that were important for NYPL to ask because of the nature of its collections and its preservation program.
The evaluative questions are divisible into two groups: the first group were questions referring to the outer protection of the volume--the binding and/or any protective enclosures (questions 5-9). The second group of questions concerned the construction, condition, and strength of the inner part of the volume--the textblock, leaves, and leaf attachment (questions 10-16). The final question (17) asked whether there were any obvious physical characteristics of the volume being sampled that would preclude microfilming as a preservation option. The answers to the questions were entered by the surveyors onto machine-readable forms identical to those used in the Yale survey.
Computer analysis of data collected by the survey made it possible to manipulate the information in many different ways. In addition to counting the answers to each question, we were able to count the number of volumes for which there were a given pattern of answers. For instance, by combining the answers to three questions about the condition of the textblock, we were able to identify volumes in four different conditions ranging from "intact" to "extremely deteriorated." By combining the answers to a different set of questions we could derive an estimate of the number of volumes in the collections in need of the three forms of preservation treatment offered by NYPL: rebinding/repair, custom conservation, and microfilming. Data on the place and date of publication gave a profile of the age and source of the collection, information of interest for collection development as well as preservation. By deriving an accurate figure for the average number of volumes present on an average shelf, it was possible to arrive at a very close estimate of the total number of volumes in the collection.
RESULTS AND CONCLUSIONS
The full tabulation of the statistics from the survey is available on request from the Conservation Division Administrative Office, The New York Public Library, Fifth Ave. & 42nd St., New York, NY 10018. The more important results and conclusions that can be drawn from them are discussed:
THE COLLECTION AS A WHOLE IS RELATIVELY NEW
The rate of acquisition has increased sharply over the last two decades. Figure 1 shows the proportion of the collection that was published in each decade since 1840. It shows a high percentage of works published in recent years. More than 40 percent of the collection is less than thirty-five years old and 54 percent is less than fifty-five years old. Sixty-three percent of the collections is composed of single-volume monographs. Less than 3 percent of the collections was published before 1840.
MUCH OF THE COLLECTION IS BRITTLE
Fully half of the paper in the collection will break if folded four times or less. Almost one-third of the collection is composed of paper which will break after only two folds. (See figure 2.) The relative flexibility of paper, as measured by this simple manual test, is a very important gauge of a volume's durability during handling by readers and staff. This large proportion of brittle or weak paper in the collections is a serious preservation problem, especially if these volumes receive even a moderate amount of use.
The quantity and distribution of weak paper (less than ten folds) is very closely correlated with age. Figure 3 shows the percentage of brittle books by decade of publication. The proportion of brittle volumes is very high (91 percent) in the decades between 1840 and 1920, declines in books published between 1920 and 1940 and drops sharply in books printed since then to zero for new books.
An interesting point suggested by the findings is the abruptness with which paper seems to lose strength. Half the volumes had very weak paper (four or less folds) and a large proportion (41 percent) had strong paper (more than fifteen folds). Comparatively few (5.1 percent) fell between these extremes. (See figure 2.) This seems to indicate that the transition from strong to very weak happens quickly. The strength/age correlation (figure 3) suggests that this occurs, at least for paper stored under the conditions in which this collection was stored, after forty-five to sixty-five years.
PHYSICAL CONDITION IS CLOSELY CORRELATED WITH AGE AND PAPER STRENGTH
Analysis of survey data allowed us to distinguish between those volumes that were intact and those that were slightly, moderately, and extremely deteriorated. "Deterioration" as used here refers only to the physical state of the paper and textblock. A deteriorated volume was defined as one in which the leaves showed some damage, the leaf attachment was broken or incomplete, or both. The condition of the cover of the binding was considered separately. The proportion of volumes in which the internal structure shows any degree of deterioration varies with age. Figure 4 shows the percentage of deteriorated volumes by decade of publication.
Of the volumes that were not brittle, very few (2 percent) were at all deteriorated and those only slightly. Since almost all the deteriorated volumes were also brittle, combining figures 3 and 4 shows vividly that there are many more volumes with brittle paper in this collection than there are deteriorated ones. (See figure 5.)
This means that there are a number of volumes which, although fragile, are still intact, because they are less frequently used. Experience and common sense as well as the strong statistical correlation between strength and condition lead to the conclusion that these volumes are at risk.
PROJECTIONS FOR THE FUTURE INDICATE A GROWTH OF PRESERVATION PROBLEMS
Because age is such an important factor in determining paper strength and because such a large proportion of this collection is so new, it is possible to predict there will be a very great increase in the volume of preservation problems in this collection in the near future. If all factors remain constant and collecting levels stay the same as they were for the decade 1970-79, the number of brittle books will double in the next fifty years. It is more difficult to make projections regarding physical deterioration, because this is a function of both age (brittleness) and use. It is likely that the number of deteriorated volumes will increase just as explosively. Mitigating factors that cannot be accurately calculated are: increased use of acid-free paper in U.S. publications; temperature and humidity controls implemented in the storage area of this collection just after the survey was done; and the possibility of a future mass deacidification program.
TREATMENT NEEDS EXCEED OUR RESOURCES
NYPL has a large and comprehensive preservation program of which it is justly proud. It can provide virtually any type of treatment the volumes surveyed might need. The survey revealed the extent to which even this program falls short of meeting the amount of treatment the collection needs.
Of the volumes surveyed, 18.5 percent were in urgent need of treatment. Extrapolated to the entire USLHG collection, this represents approximately 26,500 volumes. Of these, about 4,900 need rebinding or repair and are not too brittle to benefit from it; 15,300 need microfilming urgently; and 6,300 need extensive conservation.
Given the present rate of treatment it would take twenty years to do all the rebinding and repair currently and urgently needed. Ominous though this sounds, it is not inconceivable that the library's treatment ability might be increased by a factor of five, which would reduce the time to four years, a more manageable and realistic time frame. If rebinding and repair are estimated at approximately $10 per volume, the cost of treatment is roughly $50,000 or $12,500 per year for four years, an achievable amount.
A much graver situation exists in regards to microfilming. At present quota levels, it would take 66 years to film all that needs to be done immediately, i.e., all those volumes that are both brittle and deteriorated. Extending consideration to those volumes that should be microfilmed because they are brittle but not deteriorated adds another 217 years onto the 66. Dollar amounts are even more staggering. At $100 per volume, it would cost $1,530,000 to film those items that are both brittle and deteriorated. Another $50 million would be needed for all the other brittle books in this collection alone.
The need for extensive conservation treatment including deacidification, leaf repair and strengthening, fine binding, and custom box making is more difficult to quantify in terms of treatment time per volume, but it is obvious that the needs of this collection far exceed any reasonable allocation of resources that the library could make at current staff levels.
Protective enclosures for deteriorated items, often called phased treatment, is a useful and accepted practice for items that cannot be treated immediately. The survey indicated that 20,000 volumes (13.8 percent) would benefit from protective enclosures. At $5 per volume, this represents a cost of $100,000, a figure that is not completely dismaying, but one that must be added to other treatment costs since it is only an intermediate not an alternative to treatment.
It cannot be presumed that this collection is completely representative of the rest of the holdings of NYPL, but it is a safe assumption that the preservation needs of the collection as a whole exceed the library's resources to treat them. This should not be seen as a shortcoming on the part of the library, but as a clear and strong indication that the preservation problem is too large for any one institution to solve. Preservation must be a cooperative endeavor if it is to succeed.
FUTURE SURVEYS COULD BENEFIT PRESERVATION PLANNING
This survey and others like it have demonstrated sufficiently the size and importance of the preservation crisis. No more surveys are needed to make that point. The true value of condition surveys in the future will be to give preservation administrators information to guide the formation of institutional preservation efforts and, perhaps more importantly, to give shape and direction to cooperative efforts. [Figure 1 to 5 Omitted]
REFERENCES Gay Walker and others. "The Yale Survey: A Large-Scale Study of Book Deterioration in the Yale University Library," College & Research Libraries 46:111-31 (Mar. 1985).
Robert DeCandido is Head, Shelf and Binding Preparation Office, Conservation Division, New York Public Library.…