Shirley Baker

Vice Chancellor for Scholarly Resources and Dean of Libraries

Putting it All in Perspective

What do librarians do? We collect the record of human knowledge, preserve it, order it for use, and provide access for those who need it. Nothing in these primary roles requires that they be carried out in the current, wildly decentralized manner, which duplicates in many places the efforts of other librarians around the globe. We have historically worked in this wildly decentralized way because it works to provide timely access to resources for our patrons. It is likely and, I believe, desirable, that the library of the future will be a single, virtual, universal collection, built and maintained by librarians.

The library collections committee - a group of specialists from all over the globe - would vet what gets added to the collection. A similar group of catalogers, classifiers, and technical experts would organize the information and provide an appropriate (and yet to be imagined) array of access points. The preservation group would have two wings: those who looked after those physical manifestations that had intrinsic value (illuminated manuscripts, the poet May Swenson's hand-written notes), and those that were responsible for migrating the electronic data through changes in technology. The reference librarian equivalents would be closely allied (and perhaps the same as) the specialists doing the collecting - the old branch library model, which has always been immensely satisfying to our patrons.

Levels of access for patrons could be negotiated by affinity groups. Harvard would, of course, provide the highest level of access, with all the potential frills, for its faculty and students. Residents of small rural communities would get a lesser model, but certainly more than they get currently, paid for by their local government. As always, wealthy individuals could buy whatever they wanted.

We shall fall short of achieving this model for some time. Nonetheless, we might take it as our goal and test our current plans against it at every step. Those who are successful in taking their organizations forward have an ideal future in mind and test all immediate plans against that ideal.

Prelude to the Present

Resource sharing has been of interest to me since I supervised an interlibrary loan unit in the 1970's. In those days interlibrary loan in a research library was an essentially invisible operation. OCLC and RLG had just developed their interlibrary loan subsystems as a replacement for paper forms and TWX. The interlibrary loan staff usually worked in a closet, were visited only by a handful of scholars, and were better known to interlibrary loan staff at other institutions than to their own local library colleagues. Interlibrary loan was certainly not considered a key success factor for the parent library.

Public libraries and state libraries have always been more concerned than academics about sharing resources and have often put into place systems for identifying holdings and delivering materials. Research libraries, however, have historically defined resource sharing as a collections issue - making agreements about who collects what, especially for 'esoteric' areas - to guarantee that copies of critical scholarly materials are made available somewhere. The Farmington Plan and the Research Libraries Group conspectus are evidence of this approach. Providing timely access to these scholarly materials was not as high a priority, and, indeed, was not achievable in the pre-electronic world. RLG libraries did support their distributed collections with a resource sharing program, which, while a key part of any library's participation within RLG, never had quite the status of the collections activities.

In recent years, even research library directors' attitudes toward resource sharing have changed. A decade ago, the attitudes of most of us could be summarized as follows: borrowing is an embarrassment, lending is a nuisance. Our patrons have helped bring about a change in those attitudes. As electronic sources allowed patrons to see what existed outside our collections, patrons pressured our libraries to make those resources easily available. Once it became possible to identify the needed item in seconds, it became less acceptable to wait several weeks for delivery of the item.

Attitudes about coordinated collections have changed over time, also. Five years ago, one major research library director was heard to say, "Cooperative collection development is a crock!", visibly astonishing his colleagues. Another was heard to say, when state-wide coordinated collection development was suggested, "Don't waste time and money on that fruitless exercise; just put the money into interlibrary loan." These are crass overstatements, but they include more than a kernel of truth. In all except the richest libraries, agreements to collect in particular areas are hard to make and impossible to stick to over time. As local demands change, local collections emphases must also.

In an ideal world we librarians might be able to agree to collect in a certain area and stick to that position; for other subjects, our users would accept remote access. Unfortunately, this ideal hasn't often worked in the real world. While trying vainly to develop coordinated collections by policy, we have done quite well in developing collections with large numbers of unique resources by chance. Studies of collection overlap persistently show that often fully half of the resources in a group of libraries are held only by one library. What we could not accomplish by action, we have done quite well by chance.

Others have noted (and it is clearly documented in the Winter 1997 issue of Library Trends ) that the current consortial efforts, which may have made initial stabs at coordinated collection development, have quickly changed their focus. David Kohl, in his Library Trends article1, alludes to a thought-provoking emerging model for collection development in a consortium, what he calls the lumpy oatmeal approach. Many current consortia, instead of spending years of staff time documenting historical collections and hammering out agreements for future collecting, have focused on buying shared electronic information, creating electronic links among local catalogs, and providing swift delivery systems, increasingly directly to the patron.

Three Stages of Library Emphasis The author sees, in this century's library history, three stages of emphasis, each of which catalyzes the next stage. The first stage, of course, was collection building. Even today, without a collection you have no library. The massive expansion of literacy and of publication in the nineteenth century accelerated the rate of collection growth world-wide. It brought us approval plans and other attempts at collections in some way other than the laborious title-by-title process. Collection building continues, with each library buying as much as it can possibly afford. Electronic resources are an add-on.

Massive collections need description. Thus it is no surprise that the next stage of emphasis was bibliographic control. Some aspect of bibliographic control appeared for every library as soon as the owner had more than she could remember without a finding aid. In librarianship, discussions in cataloging and classification dominated the 19th and early 20th centuries; remember Panini, Cutter, Dewey, Ranganathan, and Lubetsky. The real breakthrough in bibliographic control, the one that became an obsessions for the author's generation, occurred with the application of electronic technology to bibliographic control. The use of computers, the development of the MARC format for cataloging, and the appearance of indexes and abstracts in electronic form transformed bibliographic control. Suddenly, it was possible to describe our collections in ways that could be dynamic, where records could be changed or enhanced, without an electric eraser.

OCLC's Fred Kilgour quickly realized the possibility of the technology: transforming cataloging from a cottage industry to a distributed electronic service. Within a very short period of time libraries around the country were eliminating their typing pools, achieving dramatic economies and moving toward unprecedented levels of standardization. Soon libraries ceased using the utilities as card-producing services and begin using them as sources of electronic records for the local online catalog. By the 1990's, libraries were experimenting with outsourcing their bibliographic control. Today's challenge is the description of and location identification for electronic information.

In the 19th century, some libraries put journal analytics - records for the articles in the journals they received - in their card catalogs. With the rapid expansion of journal publication, this was stopped. Now, we are putting those records back, this time in electronic form.

The advances in electronic bibliographic control not only made our collections more visible to our users; they made other libraries' collections visible also. Thus, the third stage of emphasis is access. Previously, only the sophisticated researcher knew to search the National Union Catalog and the Union List of Serials. Now everyone and her cousin are stumbling across electronic indexes and electronic union catalogs, and discovering items they want to read. The interlibrary loan subsystems developed by the bibliographic utilities, while apparently developed as an afterthought, did emerge to serve resource sharing needs.

Where are we now?

We have entered into a new period of resource sharing, driven by our users, our governing agencies, and even by ourselves. Governing agencies, whether university administrations or state or local governments, have become intensely interested in fostering cooperation and resource sharing. One always has the suspicion that their motives are to reduce our budgets, but many librarians have been successful in convincing their governing agencies that consortia will dramatically enhance rather than simply replace our library and information services.

Resource sharing is moving forward on two fronts; one is the enhancement of traditional interlibrary loan and the other is the explosion of consortial arrangements for direct patron borrowing. These distinctions, often blurred, should be clarified before we examine the two developments in more detail.

Consortial arrangements for direct patron borrowing are not interlibrary loan, although librarians and system vendors often refer to them as that. The critical difference is that direct patron borrowing is based on an agreement among individual libraries to treat each other's patrons as eligible borrowers and lend directly to them. Thus, direct patron borrowing, as it currently exists, is what we used to call reciprocal borrowing, a type of circulation rather than a type of interlibrary loan. True interlibrary loan is an agreement among all libraries, based on national and international codes, to lend to each other, library to library, for the benefit of the library's patrons, rather than library to patron.

Many of our colleagues argue that direct patron borrowing makes interlibrary loan irrelevant and efforts to improve it are short-sighted, even retrograde. The author would argue that growth of direct patron borrowing is exciting and life-saving for librarians (and a noble step toward the virtual, universal library). Traditional interlibrary loan, even with dramatic improvements, cannot possibly meet the unstoppable rise in patron demand. Both avenues - direct patron borrowing and interlibrary loan - require our attention if we a librarians are to fill our patrons needs in the coming years. In print in 19952 and again in 19973, library technology guru cliff Lynch details the technical reasons why these two methodologies will coexist for the foreseeable future, desirable as it might be that one should disappear. Studies of patterns of demand, primarily in but not limited to research libraries, show that large portions of demand can be met by direct patron borrowing through consortia, but a surprisingly high percentage of needed materials (as large as 50%) must be obtained from outside sources. Until the world becomes a single consortia, some library to library arrangements must persist.

Won't Electronic Publishing Make This Discussion Irrelevant?

Anyone who has been a librarian for more than a short time has been asked repeatedly, "Won't (fill in the latest technology) eliminate the need for libraries?" This question was probably asked when Gutenberg introduced moveable type. Since books could be printed in large numbers and bought by anyone, why not maintain libraries only for manuscripts, as warehouses of dead information? My answer to this recurring question is, "Nothing goes away in libraries. Everything is an add-on." Books did not replace manuscripts or papyrus or clay tablets. Microforms and videos sit on our shelves beside books. Even if we wish it to happen, electronic publishing will not, in our lifetimes, completely replace all the other forms we keep and will not eliminate scholars' desires to use earlier forms of information.

For the purposes of this discussion of resource sharing we assume the following. The amount of information available electronically will increase, at a rate impossible to predict. Indeed, the Mellon-funded JSTOR project seems to indicate that it may even be economic to replace some of our older materials with electronic versions. Consortia will assist libraries in negotiating for, sharing access to, and preserving electronic information. The availability of electronic information may alleviate the inevitable inter-institutional friction in distance education. Pervasive electronic publishing will transform traditional publishing practices in ways impossible to predict. But, the availability of electronic information will not materially alter the key issues in the following discussion.

Developments and Challenges for Direct Patron Borrowing

Today's developments in direct patron borrowing are based on the current technology or technology in the making. Paper union lists, or even the de facto union lists created by bibliographic utilities, are being replaced by linked local online catalog systems, with patron files, holdings, and circulation status available. Physical delivery is enhanced through telefacsimile, electronic transmission of images, and contracted surface delivery systems. Electronic information is contracted for consortially and shared with participating institutions.

The most common implementation of such a consortium is based on sharing a common technology for the local library systems, with union catalog capability and messaging overlaid. Illinet and OhioLink are examples of this. Creating these systems requires the impetus of an offer that can't be refused. It is the author's opinion that few individual libraries are capable of sublimating local self-interest sufficiently to engage in such unfettered resource sharing without the most powerful incentive, usually in the form of money and external pressure. This consortial model tends to be laid onto somewhat skeptical libraries, which succumb because otherwise unattainable goals can be realized (a new system, some collections funding. . . . .). The mix of libraries of varying sizes common to these consortia bring some interesting lessons, often counterintuitive. Indeed, Hugh Atkinson, librarian at the University of Illinois in the 1970's, and a mover in the development of Illinet, proved to all of us highly skeptical librarians at major research institutions, that the largest libraries can be gainers in a consortium of so-called non-peer libraries - truly counter-intuitive!

The second consortial type is based on linking local systems with different technologies, often a mix of systems from several different vendors. The success of these systems will depend upon the application of standards such as Z39.50 to enable communication among disparate systems. A goal is to allow consortium participants maximum local autonomy while gaining some of the benefits of direct patron borrowing. The CIC initiative, CoPY (Columbia, Penn, Yale), and the recently-formed Pennsylvania consortium PALSI are examples of this approach.

Implementing either consortial model requires battling institutional inertia. A quality implementation requires the attitudes common to any major change: Question everything; Give up a sacred belief a day; Be willing to abandon beloved practices. This is difficult for many institutions. Our staff can be disconcertingly certain of the rightness of their approach and oblivious of the need to change.

Developments and Challenges for Interlibrary Loan

The author believes that interlibrary loan will continue to have a critical role to play in libraries for the foreseeable future. If we are lucky, direct patron borrowing through consortia will reduce interlibrary loan traffic rather than just slow the rate of growth. If that is true, we will be able to reallocate we have put into ILL to other, more pressing functions. Whether ILL traffic is reduced, or the rate of growth merely slows, we can assume that direct patron borrowing will skim off the easy citations, leaving interlibrary loan operations to deal with the difficult and obscure.

Those who still believe interlibrary loan to be a necessary service have been working to improve it. Interlibrary loan as practiced in the beginning of this decade compared badly with other units in our libraries, particularly our cataloging departments which had made dramatic improvements in work flow and productivity over the previous years. Compared to the post-industrial electronic service that cataloging has become, interlibrary loan looks like a pre-industrial cottage industry. The last major technological improvement in interlibrary loan was the introduction of ILL subsystems by the major bibliographic utilities in the 1970's. Modest enhancements had been made in those subsystems over the years, but on-going inadequacies led to the development of enhancement software packages by disgruntled ILL staff.

Experience with the demands and frustrations of ILL led this author to conclude that the time was ripe for library directors to be as engaged in ILL. An involvement comparable to directors' previous intense interest in technical services technology and workflow issues was required. You will remember the 1970's and early'80's, when library directors were obsessed with the online catalog, authority control, and AACRII!

In the early 1990's, the author and key colleagues were successful in outlining resource sharing issues for research libraries in a white paper4 which led to an Association of Research Libraries initiative to re-conceptualize interlibrary loan. Within a short period of time, ARL's North American Interlibrary Loan/Document Delivery (NAILDD) project emerged, with three major goals. Those goals were to improve interlibrary loan through the development of financial software to handle billing, the development of management software to provide decision support information for both practitioners and library administrations, and the implementation of standards to allow communication among electronic systems without re-keying of data.

The NAILDD project quickly engendered a Developers/Implementers Group (the DIG), where software vendors could come together to discuss libraries needs and technological solutions. An early result was OCLC's implementation of their Interlibrary Fee Management (IFM) software, which allows charging among libraries without individual bills being issued. This software had the pleasant effect of facilitating international interlibrary loan. Librarians from Australia, from Hong Kong, and other countries have praised the convenience of being able to borrow from US institutions and pay charges without cutting checks in foreign currency.

The most recent progress has been in the implementation of the international interlibrary loan protocols, ISO 10160* and 10161*. In 1995 NAILDD's Developers/Implementers Group accepted National Library of Canada's challenge to implement the international standard. A small group of vendors members interested in implementation assembled at the Coalition for Networked Information meeting in Portland in November 1995and another group, the International Protocol Implementers Group, with the euphonious acronym IPIG, was formed. Progress is occurring faster than expected, with participation by vendors large and small (OCLC, The Library Corporation, III. . .) and significant international interest. There are now British and Australian members of the IPIG and interest from Scandinavia, Central Europe, and Asia.

Full implementation of the standard protocol will mean that requests can be moved from one local system or bibliographic utility to another, without re-keying, until the request is filled. This will significantly reduce staff time, the largest cost component in interlibrary loan or document delivery.

Indeed, a 1993 ARL/RLG interlibrary loan cost study showed $30 as the cost of an ILL transaction ($19 for the borrowing institution and $11 for the lender). Three-quarters of this is staff cost. If the staff time involved in an interlibrary loan can be reduced, costs will drop accordingly. Productivity improvements of 100% are within reach.

The 1993 cost study has just been reproduced with eighty-five ARL libraries and twenty-three libraries from the Oberlin Group of colleges. This recent Mellon-funded survey also attempts to measure performance - turn-around time, fill rate, and user satisfaction. The analysis of survey responses is underway and results should be available widely by summer.

From the latest survey, we hope to see the effects of changes made by libraries which have participated in both. We also hope to identify best practices and disseminate them widely, so that libraries wanting to make improvements will have models of the most cost effective and user-satisfying operations. The desire to identify best practices grew out of the results of the first survey. Directors are anxiously awaiting the results of the second. (Mary Jackson, ARL's Access and Delivery Services Consultant, is the principle investigator on this study.)

What next?

Call it global, call it international. Whatever, that is the watchword for tomorrow. Business interests are global, education is increasingly international. Who among us hasn't heard some senior administrator orating on the need to make (our company, the curriculum, our organization) international? Who hasn't had to deal with a user community that is multi-national and multi-lingual? With library budgets in a losing battle with increased publication prices and volume, what library will own the full range of international materials to meet local demand? None that this author can envision.

While there are some valiant attempts to bring a larger proportion of international materials into our libraries, there will be a continuing and growing need to engage in international resource sharing. To make this viable, several things must happen.

For starters, we must set a good example by increasing our willingness to lend outside our borders, building the sound practices which will alleviate some of the real or feared problems. We must start now changing staff attitudes and encouraging taking calculated risks.

We must encourage the availability of bibliographic information from all quarters of the globe. It appears that OCLC and RLG are engaged in an exciting competition to see which can bring the largest number of international libraries into their databases. (OCLC's fee management system and RLG's shares program remove one of the obstacles to lending - the difficulty of paying in different currencies.)

We must improve delivery options to include those as reliable and safe as UPS or FedEx, so that we can 'mail' our resources, sure that they will get there, and get back. We must support the enhancement of electronic delivery mechanisms so that non-returnables can go directly to the end-user, at light speed. We must support the enhancement of our country's electronic infrastructure, so that it can handle our traffic at something that approximates light speed rather than (as happens occasionally) at rush hour gridlock pace.

We must complete the conversion into machine readable form of records for all the holdings - microforms, government publications, manuscripts - of all of our libraries - large and small, North American and international.

Locally, we must work every day to change attitudes, to nurture those who think regionally, nationally, and internationally. We must reward those who can recognize when professions of institutional autonomy are really an excuse for parochialism. Our parent organizations are having to work more collaboratively. We as librarians are often the first to make that collaboration real. We should continue to keep ahead of our colleagues and set good examples for our patrons, our administrators, our legislators.

Footnotes

Kohl, David, Resource Sharing in a Changing Ohio Environment, in Library Trends, Winter 1997, pp. 435-447.
Lynch, Clifford A., System Architecture and Networking Issues in Implementing the North American Interlibrary Loan and Document Delivery (NAILDD) Initiative in The Future of Resource Sharing, Shirley K. Baker and Mary E. Jackson, editors, 1995, The Haworth Press, pp. 145-168.
Lynch, Clifford A., Building the Infrastructure of Resource Sharing: Union Catalogs, Distributed Search, and Cross-Database Linking, in Library Trends, Winter 1997, pp. 448-461.
Baker, Shirley K. And Mary E. Jackson, Maximizing Access, Minimizing Cost: A First Step Toward the Information Access Future, Association of Research Libraries, Washington, DC, 1993.

Shirley K. Baker
Vice Chancellor for Scholarly Resources
and Dean of Libraries
E-mail: shirley.baker@wustl.edu
Washington University
Campus Box 1061
St. Louis, MO 63130
Phone: 314-935-5400
Fax: 314-935-4045