Abstract: The storage of eye-readable information at a location removed from its parent institution has a history of more than 2,000 years. Despite changes in the kinds of information that are stored and the technologies that enable their storage, the relationship between a reader’s time and the distance of material from the reader is a constant challenge to information providers. Competing visions of service and economics, to which remote storage is one response, are timeless.

This is a two-hundred-fifty-line history of library storage, written at a time when the physical management of research collections relies increasingly on the use of remote facilities to house paper materials. It races over two millennia, hardly pausing to document but pretending to identify a series of trends: the long-lived tension between recorded scholarship and physical space, tradeoffs between physical ownership, access, and physical space, and the changing solutions applied to these problems by many generations of librarians. It begins in antiquity.

It is written, in a book, that Socrates worried about the corrupting influence of books on learning. (Plato 1955, 66-69) In his age-- the sixth century BC-- knowledge traveled by word-of-mouth, and though scrolls and codices existed, Socrates was convinced that scholars would never use them. If only we had listened. The fixing of texts in papyrus, animal skins or paper, embodied ideas-- turned them atomic in Nicholas Negroponte’s post-modern phrase (Negroponte 1995,11-13)-- and began the quest for finite library space, now well into its second millennium.

Three centuries after Socrates’ lament, acquisitions rates at Alexandria, the world’s first comprehensive research library, began to threaten its storage capacity. A document dating from 257 BC shows that the library received 434 papyrus scrolls in 33 days. (Manguel 1996,188) And, yes, Alexandria developed a remote storage facility. Several sources cite a depository of 48,000 duplicate scrolls from the library housed in the Temple of Serapeum, located in the Egyptian quarter of the city. (Millares Carlo 1993,228-29, Brundige,n.d.)The hubris of a quest for universal knowledge has, from Alexandria forward, produced the nemesis of space crisis. At Alexandria, bibliographic overcrowding was relieved not by building but by destruction. The collapse of classic civilization, and the preservation of knowledge in monastic retreat, postponed the next age of monumental knowledge building in the West for a millennium.

•Gutenberg(1450) •Vatican Library(1455) •Ordonnance de Montpellier(1537) •El Escorial(1565) •National Library of France(1595) •Bodlean(1598)

Nicholas V, Pope from 1447-1555, and a liberal patron of the arts, unified several Vatican collections into a single repository during his papacy. This consolidation of resources, and subsequent assembly of comprehensive collections to foster research and statecraft found advocates in Spain, where Philip II armed his Escorial residence with a huge, imperial library, in France, where Henry IV established a Bibliotèque Nationale in Paris and in England, where Mr. Bodley’s library became a feature of Oxford University. (Harris 1995,132) Agustín Millares Carlo points to the Ordonnance de Montpellier, where Frances I dictated a decree intended to gather a copy of all works published in France at the Royal Library, as the first national deposit law. The Ordonnance also legislated an early approval plan by stipulating that a copy of every book imported to France be offered to the Royal Library for purchase. (Millares Carlo 1993,261)

Two centuries of collection development in North America culminated in a debate between two Harvard administrators that students of library history, and of remote storage, cite as a seminal framing of the issues.

"I am not proposing a crematorium for dead books, but only a receiving-tomb. Neither am I proposing that the bibliophile or the antiquarian should be absolutely deprived of his idols, but only that his access to them should made somewhat less convenient and attractive." (Eliot 1902,55)

"The point to be carefully considered is, how will the books thus set aside be treated; how will their segregation affect the interests of scholars; to what degree are they still to be accessible?" (Lane 1903,11)

These statements of opinion, preserved on the pages of Library Journal, set a tone that resonates a century later. The writers’ views on the issue of handling expanding library collections led to very different proposed solutions. Harvard President Eliot’s logic led him to suggest the creation of four storage facilities to serve the entire United States. Librarian Lane’s vision was less sweeping. He proposed a cooperative repository operated by Harvard, the Boston Athenaeum, the Massachusetts State Library and other libraries in New England.(Line 1980,102-103) As a historical footnote, this particular controversy gradually ended with Eliot’s retirement in 1909 and the completion of Widener Library in 1915. However, the basic issues of the debate on storage facilities -- economics versus service, a lack of agreement on what predicts use and on the proper locus of cooperation-- remained and remain unresolved.

Although I cannot demonstrate it, there must be some sort of converging/diverging relationship between acquisitions rates and construction costs-- similar to the supply and demand curves of classical economics-- that drives the intensity of discourse and action on remote storage. Periods of great prosperity, the 1920s and 1960s in this country, the oil-boom in some regions of Latin America and the Middle East, make monumental construction relatively cheap. Hard times reduce the prospects for both construction and collections, but between prosperity and depression lie long periods when acquisitions outstrip available storage space. This is where the First World has been since the 1970s and where it seems likely to remain for the foreseeable future.

Figure 2: Major Events in the Modern History of Remote Storage

•New England Depository(1942) •CRL(1951) •BLLD(1973) •N.California Regional Facuilty(1980) •Buhr Shelving Facility(1981) •Harvard Depository(1986); :

The second time line fixes six events in the recent history of remote storage. The New England Depository, which opened in the unlikely year of 1942, conformed to Lane’s vision from forty years before, a cooperative, regional facility. A similar approach took shape at the Midwest Inter-Library Center, which became the Center for Research Libraries in 1965, with storage shared by eleven research libraries in Illinois, Indiana, Iowa, Michigan, Minnesota and Ohio. The British Library Lending Division (1973) illustrates another vision, the assembly of a collection in a single location intended as a nation-wide lending resource.

By the 1980s many research library collections had surpassed the capacities built to hold them two decades before. Some of the shortfall was met with new construction, such as that at the University of North Carolina, Boston College, the University of Texas at El Paso and Queens College. However, these facilities proved exceptional. The rule became high density storage units located at some remove from the institutions they served. The Northern California Regional Facility, in which state appropriations purchased existing space and new furnishings for the holdings of several University of California libraries; the Buhr Shelving Facility, a former manufacturing plant which the University of Michigan bought and retrofitted to store its collections; and the Harvard Depository, which pioneered the construction of specifically-designed library storage facilities, illustrate the diverse storage solutions applied during the period in the United States.

The late 1970s and 1980s also produced a boomlet of storage research, published in the library literature. The Association of Research Libraries compiled SPEC kits on the topic in 1977 and 1990 that serve as a good bench marks. And Phyllis O’Connor summarizes much of the periodical and monographic literature in a recent issue of Serials Review. (O’Connor 1994,17-26,44) Given the number of different solutions shown on the time line, it should come as no surprise that these studies do not agree on the most effective solution to bibliographic overcrowding.

Two major positions order the debate. A Metcalf School, reflecting the views of long-time New York Public and Harvard librarian, Keyes Metcalf--and reified in the many editions of his Planning Academic and Research Library Buildings-- espouses the virtues of a regional, cooperative scheme. Its proponents stress that remote storage is best viewed as part of a program that reduces inter-library duplication and fosters cooperative collection development. Donald Swain’s essay, describing the planning of the Northern California facility (Swain 1978) and Michael Buckland’s proposal of a "last copy" scheme with transparent ownership as the purpose of cooperative storage (Buckland 1990), develop facets of Metcalf’s approach.

A Fussler School, pioneered in the work of Herman H. Fussler, which postulates that the cost of housing a large collection will be lower if some of it is in compact storage, follows the economic approach championed by Harvard President Eliot.(Fussler 1969) Although his conclusions and analysis are directly contradicted by Joanne Harrar’s Ph.D. thesis (Harrar 1962), Fussler’s approach holds sway in current library practice. An interesting corollary, developed by Michael Cooper and Michael Gorman, asks where the compact storage should be. Cooper, writing in Library Quarterly, compares storage alternatives and concludes that, with the exception of never-circulating material, greatest savings occur when compact storage is open stack and on-campus. (Cooper 1989) Gorman, writing two years earlier, agrees and stresses that moveable compact storage is optimum, principally because selection will never produce a non-circulating collection. (Gorman, 1987)

Despite the variety of solutions to space shortages proposed in the 1980s literature, 1990s practice clearly favors one approach, that developed in the Harvard Depository. In 1986 Harvard completed a high-density, modular facility built on land sufficient to hold several; there are currently three modules at the Southborough site. Its design features climate control, sodium vapor lighting, high bay adjustable shelving and an inventory tracking system.¹

The documentation for these new facilities and the thinking that produced them is no longer-- or at least not yet-- in the library literature, but rather on the World Wide Web. Searches on Internet engines that combine terms such as "remote storage," "library materials," and "high density," summon a broad array of library documents, institutional trustee minutes and press releases, some of which appear as an appendix to this paper. These descriptions include a number of interesting factoids. Remote storage is ubiquitous: Williams College has six science libraries and three depositories holding back runs of science journals. It appears in unexpected places: Rice has a depository under its football stadium, which it shares with the notorious Marching Owl Band. And it holds an ever-larger percentage of library collections: Colorado State reports 500,000 of its 1.5 million volumes in storage. But in addition to minutiae, these texts show an important, public facet of current off-site storage-- how remote storage facilities are described by those who design and manage them.

The rationale for storage facilities is uniformly economic necessity. Given current costs and budgetary realities, off-site, high-density storage seems the only viable alternative to mass deaccessioning. But this is clearly a painful choice and one unhappily made. In an interview in The Chronicle of Higher Education, Paul Mosher reflects on the difficulties that remote storage caused him as a Medieval Studies doctoral student. Now a library administrator at Penn, Mosher sees off-campus storage as a necessity: "All our libraries are full. Something had to be done," he says for the record. (Young 1998,A27)

A reluctance to undertake remote storage in the first place finds reflection in public justification. Some of us cite the addition of public space, "seats" in the argot, as a virtuous byproduct of removing materials; gone is the argument once made that a smaller collection is an easier-to-use collection. The importance of environmental upgrades finds expression from the details of "very-flat" construction and temperature and relative humidity statistics to less technical assertions that the facilities will simply prolong the life of books. And, of course, off-site storage is much cheaper; Yale calculates one-tenth as expensive as traditional, on-campus, open-stacks facilities. But these assertions lack the enthusiasm so evident in the description of other contemporary initiatives, such as networked electronic resources.

Conceptual and political problems, inherent in remote storage, also emerge from the public documents. Variations on the theme of "if you can take them off campus, why keep them at all," haunt our writing. Ohio addresses this dilemma with the statement that its depositories hold "permanently-held but little used library materials." Texas is more forceful, describing its off-campus storage as "... a facility for planned remote storage of permanent, important (my italics), but little used library materials."

What will go into storage vexes us, as well. The mantra, "little-used materials," is both incomplete and misleading. The Fussler School, detailed above, argues that "little" should be as close to "non" as possible, but philosophy and internal politics often lead to caveats promising "flexibility in returning material to campus," quoting the Yale documents. Library users, especially faculty members, raise the crucial issue of how removal of material will effect research. Some of the public documents mention an active faculty involvement in the actual selection of materials, before they are transferred. However, the imperatives of a massive move, Cornell is currently transferring 2,000 volumes a day, limits faculty involvement to consultation at the planning stage.

If librarians have not resolved all the issues that off-site storage raises, they have come to share a series of basic assumptions on how the new facilities will operate. One prerequisite for including an item in high-density storage is to have it represented in a library’s bibliographic database. Not to do so is to consign it to oblivion, but providing a digital surrogate enables readers to "virtually" browse holdings at a computer screen. It also offers the potential of enhancing access through electronic wizardry such as linking bibliographic records for these materials to electronic representations of their tables of contents. Librarians also agree on a rapid delivery of materials from storage as essential for establishing remote facilities. Twenty-four hours to two or three days appear in the new facilities descriptions, and several documents mention transmission of articles by FAX or Ariel.

Ironically, these developments potentially alter the relationship between distance and access and threaten to overturn the intent of placing little-used materials off site. By representing every item in remote storage in automated databases, at a time when significant segments of centrally-housed collections are not, and offering delivery services which place materials in readers’ hands more quickly than the current system of open stacks and self service, libraries may provide better access to materials considered of low utility. That is, we may actually make materials stored remotely more accessible and more used, a danger signaled in Gorman’s work. (Gorman 1987)

Surprisingly, current research and position papers gloss over two major issues. The first is a lack of agreement on a best set of practices for off-site storage. Research from the 1980s is highly contradictory, and the on-again, off-again storage consortium in the New York metropolitan area produced widely-divergent analysis and recommendations for action. (Final report of the working group 1996, Young 1999,A26) Neither does extant literature offer meaningful guidance on the sticky issue of selection of materials for storage, although the use of automated circulation data is often at least part of the mix. A second unfulfilled expectation is analysis of the role of new technologies and how they would effect the facilities of today. In public documents that describe remote storage, digital technology as a strategy is conspicuous only by its absence.², For now, library planning implies that paper collections will continue to grow, substantially, and that more of them will reside at remote locations.

In conclusion, I turn readers’ attentions to the work of Jorge Luis Borges, who knew a thing or two about libraries, and much more about speculation. Writing of an infinite "Library of Babel," Borges describes two types of intruders. The first are inquisitors, always on the alert for material that offends orthodox sensibilities. "Other men, inversely, thought that the primary task was to eliminate useless works. They would invade the hexagons [Borges’ library shelves], exhibiting credentials which were not always false, skim through a volume with annoyance, and then condemn entire bookshelves to destruction." (Borges 1962,84-85)

In the year of Borges’ centennial, librarians increasingly see space as a resource as precious as capital and staff. But as this paper demonstrates, the commodification of space is not a development of the Information Age. In fact, the inability of facilities to keep pace with the simultaneous acquisition and preservation of information has challenged our professional ancestors for centuries, and remote storage has been used to house collections for more than 2,000 years. Where and how materials are stored–the essence of the Eliot-Lane debate–will continue in contention as long as eye-readable media remain important information carriers.

What is new to the issue, although only incipient in the debate, is the role of digital technology in information storage and retrieval. How long buildings will remain the principal repository of information is no trivial question. Librarians wish neither to fall behind the technological curve nor to be cast in the role of Borges’ second class of intruders, or worse, as poor stewards of the public’s property. If history offers any guidance here, it is that economy and service exist in competition. Low-cost real estate and high-density shelving will be economical only to the degree that they enable the delivery of information to those who use it.

Appendix: Web Sites Describing Offsite Storage Facilities in US Libraries

Texas: Library Storage Facility http://www.lib.utexas.edu/Exhibits/preserve/lsf.htm Accessed: October 19, 1999.

Yale: Yale University Library Off-Campus High-Efficiency Shelving Facilityhttp://www.library.yale.edu/Administration/Shelving/ocs.html Accessed: October 19, 1999.

References

Buckland, Michael K. 1990. Little-used duplicates, cooperative collection development and storage. Collection management 13,no.4:39-52.

Cooper, Michael. 1989. A cost comparison of alternative book storage strategies. Library quarterly 59, no.3:239-60.

Eliot, Charles William. 1902. The division of a library into Books in use, and books not in use, with different storage methods for the two classes of books. Library journal 27,no.7:51-6.

Gorman, Michael. 1987. Moveable compact shelving: the current answer. Library Hi Tech 5, no.4:23-6.

Fussler, Herman H. 1969. The economics of book housing. In Patterns in the use of books in large research libraries 133-40. Chicago, University of Chicago Press.

Harrar, Helen Joanne. 1962. Comparative storage warehouses. Ph.D. dissertation, Rutgers University..

Harris, Michael. 1995. History of libraries in the western world. Metuchen, NJ, The Scarecrow Press.

Lane, William Coolidge. 1903. The treatment of books according to the amount of their use. Library journal 28, no.7:9-16.

Lerner, Fred. 1998. The story of libraries. New York: Continuum.

Line, Maurice. 1980. Storage and deposit libraries.In Encyclopedia of library and information science 29:101-133. New York and Basel: Marcel Dekker.

Negroponte, Nicholas. 1995. Being digital. New York: Knopf.

Manguel, Alberto. 1996. A history of reading. New York: Viking Penguin.

Millares Carlo, Agustín. 1993. Introducción a la historia del libro y de bibliotecas. Mexico: Fondo de Cultura Economica.

O’Connor, Phyllis. 1994. Remote storage facilities: an annotated bibliography. Serials review summer 1994:17-26,44.

Plato. 1955. Phaedrus. Plato. Ithaca, NY: Cornell University Press.

Swain, Donald C. 1978. Regional library planning for northern campuses of the University of California. In Running out of space- what are the alternatives?, ed. Gloria Novak. Chicago: American Library Association.

Young, Jeffrey R. 1998. In the new mode of the research library, unused books are out, computers are in. The chronicle of higher education October 17:A27.

___ . 1999. 3 top research libraries plan vast, new facility to store little-used books. The chronicle higher education April 30: A26.

.