College and Research Libraries Cooperative Collection Developlllent at the Research Triangle University Libraries: A Model for the Nation Patricia Buck Dominguez and Luke Swindler The cooperative collection development programs of the Research Triangle university libraries are the oldest and most successful in North America. Analyzing their evolution and expansion over six decades, the authors identify the rationale and principles of successful cooperative collection development, the types of cooperation that work best for different subjects and kinds of materials, and the factors that promote cooperation over the long term. ooperative collection develop- ment is the flag, motherhood, and apple pie of librarianship. Everyone is for it. 1 But while library literature is full of attempts to describe what it is or explain how to do some aspect of it, there are no critical analyses of cooperation based on long- term case studies that document what has worked and why. The history of cooperation at the libraries of Duke Uni- versity, North Carolina State University (NCSU), and the University of North Carolina at Chapel Hill (UNC-CH), which together form the Research Tri- angle university libraries, provides the opportunity for just such a study. Librarians at Duke University, NCSU, and UNC-CH have cooperated for more than half a century. Recent statistics at- test to the success of their efforts. Com- parisons of nearly two million records in their shared online catalog revealed that 76% of the titles were found on only one campus, and only 7% were common to all three universities. 2 Applying this per- centage to their combined holdings of 9,536,556 volumes in the 1991/92 ARL Statistics, the number of unique volumes available to researchers at the three Research Triangle universities was 7,247,7~a figure probably exceeded only by the libraries at Harvard, Yale, Illinois, and the University of Cal- ifornia-Berkeley. Reflecting not only on the unique holdings but the coordinated, interde- pendent, and interlocked nature of the collections, a former provost at UNC- CH stated that the cooperative collection development effort of the Triangle Re- search Libraries Network (TRLN), the um- brella organization for library cooperation among the three universities, was the finest example of planning on campus.3 In congressional testimony on federal sup- port for libraries, the Research Triangle consortium was the only example of successful cooperative collection develop- ment cited.4 Why have observers singled Patricia Buck Dominguez is Humanities Bibliographer and Luke Swindler is Social Sciences Bibliog- rapher in the Collection Development Department, Davis Library, University of North Carolina at Chapel Hill, Chapel Hill, North Carolina 27599-3918. Their e-mail addresses are pat_dominguez@unc.edu and luke_swindler@unc.edu. They wish to thank fohn Abbott, Florence Blakely, Gary Byrd, Ginny Gilbert, fames Govan, foe Hewitt, fohn Rutledge, and fohn Shipman for .reading and commenting on earlier versions of this article. 470 out TRLN for special praise, and how has it managed to become the oldest and most successful large-scale cooperative collection development program among North American universities?5 Throughout six decades of trial and error, administrators, faculty, and librari- ans at Duke, NCSU, and UNC-CH have sought to identify the rationale and prin- ciples of effective cooperative collection development, the types of cooperation that work best for different subjects and kinds of materials, and the factors that con- bibute to successful cooperation over the long term. The lessons they have learned from their attempts to address these issues can help others around the country create effective cooperative collection develop- ment programs. THE EARLY 1930s: SETTING THE STAGE Historical and economic circumstances played a crucial role in the development of cooperation between Duke and UNC by limiting the options available to ad- ministrators, faculty, and librarians.6 After the Civil War, the South was the poorest region in the nation. At the turn of the century, North Carolina, which had still not recovered from the Civil War, was the poorest state in the region? Thirty years later, the Depression was reversing much of the economic pro- gress the state had made since then. Libraries reflected the state's economic fortunes. In 1901 the library at UNC, the largest academic library in the state, had one librarian, two student assistants, and about 40,000 volumes.8 A generation later, although a basic research library existed at UNC and the nation's largest tobacco fortune was building another at nearby Duke, neither institution possessed a great collection. Indeed, both libraries suffered budget reductions during the Depression, and a federal report issued in 1937 ranked the Chapel Hill-Durham area only thirty-fourth among the seventy-seven urban areas having li- brary collections in excess of 500,000 volumes.9 The second factor leading to coopera- tion was the ability of administrators, Cooperative Collection Development 471 faculty, and librarians to see beyond the limitations of their circumstances. Frank Porter Graham, president of UNC, and William P. Few, president of Duke, knew that they did not have the resources to build great universities in the conven- tional way, but they shared the New South vision of uplifting the region through planning and cooperation.10 To achieve their ambitions for their universities and the region, Graham and Few were willing to entertain unor- thodox solutions to the problems they faced. In 1933 they formed the Joint Committee on Intellectual Cooperation to determine how the two institutions could enhance and extend their re- sources. Two years later the group issued A Program of Cooperation, a re- markable document that asserted: The University of North Carolina and Duke University are confronted with obligations and opportunities which they can meet adequately only through a program of cooperative en- deavor.11 · Within the context of university co- operation, the presidents perceived the importance of library cooperation: Although these two libraries are al- ready the largest in the Southeastern States, neither has nor will be able to provide for a long time to come the materials for study and research which are to be found in the great libraries of the North and East. The opportunity of supplementing the resources of each li- brary by those of the other, offered by the physical proximity of the two in- stitutions, is one of which it is pro- posed to take advantage.12 This statement provided the philosophi- cal framework for library cooperation. Just as the presidents provided the vi- sion for cooperation at the university level, library directors Robert Downs (UNC) and Harvie Branscomb (Duke) provided leadership for library coopera- tion. Both men were willing to risk a cooperative approach to building library collections, despite the lack of models for doing so, because of the existence of a universitywide context favorable to . cooperation. For the same reason, faculty 472 College & Research Libraries on both campuses were willing to sug- gest and support cooperative projects. The third factor that encouraged cooperation was the availability of out- side funds. The General Education Board (GEB), a philanthropic agency that John D. Rockefeller endowed, played a crucial role. In the early 1930s it made the improvement of higher edu- cation in the South, particularly library and laboratory facilities, a prime objec- tive.tJ Influenced by sociologist Howard Odum and the Chapel Hill regionalists whose research it financed, the GEB hoped that funds spent enhancing col- leges and universities would translate into improved economic well-being and eventual rehabilitation of the region. 14 The Program of Cooperation echoed those sentiments and ambitions. Memos between administrators, faculty, and librarians at Duke and UNC mentioned the GEB, highlighted the opportunities it offered the two institutions "to assume leadership in this region," and expressed the fear that "if these two institutions can't get together, they [the GEB] seem to be seeking other institutions that might do this and their policy may be to assist some other institution more thoroughly than they would either Du~e or Carolina separately."15 To a large ex- tent, then, library cooperation came into being because a funding agency en- couraged it tangibly. 16 The fourth factor leading to coopera- tive collection development was shared bibliographic information about the col- lections and enhanced access to the materials. The GEB underwrote an ex- change program for main entry cards in 1934. This bibliographic information was essential to the success of the cooperative programs. Indeed, until fa- culty and librarians knew what both li- braries held, cooperation could not work. Special inter-library loan arrange- ments (including daily delivery service) and the extension of full library privi- leges to faculty and advanced graduate students at the other institution, in place by 1935, also facilitated cooperationY Interinstitutional cooperation there- fore began because visionary individu- November 1993 als in positions of authority saw it as a way to surmount dismal economic cir- cumstances and enable their institutions to compete successfully with richer uni- versities. Librarians, nurtured by grant funding, needed bibliographic and physical access to each others' holdings in order to cooperate in building collec- tions. Only when all these factors came together could cooperative collection development programs begin. THE LATE 1930s: SEARCHING FOR WAYS TO COOPERATE Library cooperation began in 1934, when the GEB granted Duke and UNC $12,500 for a joint catalog that "facilitates the interchange of books and makes possible a co-ordinated development of future book collections."18 Cooperative collection development dates from the following year. With the stage set, Downs and Brans- comb began to address the major issues of cooperative collection development: What are the rationale and principles of cooperation? How do libraries cooperate? Which academic disciplines, subjects, and types of materials make good candi- dates for cooperation? How do librari- ans, faculty, and administrators work together to develop effective programs? Following the themes outlined in A Pro- gram of Cooperation, the library directors agreed that the goals of cooperative collec- tion development were to achieve excel- lence and serve users by providing resources for research that the libraries could not afford otherwise, rather than to save money.19 They planned to reach these goals by creating coordinated, in- terdependent, and interlocked collec- tions that minimized the unnecessary duplication of materials.20 Mter determining the goals and objec- tives of cooperation, Downs and Brans- comb developed five principles of cooperation.21 In the first place, they agreed that cooperation would empha- size what a library, acting in self-interest, could contribute to cooperation. To this end they encouraged each institution to build on the strengths of its academic programs and library collections. Sec- ond, librarians did not restrict what their cooperative partners could acquire.22 As Downs later observed, "Libraries should not be asked to give up anything but rather to assume positive responsibili- ties and receive direct benefits."23 Third, the directors, who were sensi- tive to the potential use of items, decided to limit cooperation to materials needed for "graduate and research activities." They excluded instructional titles, whether for undergraduates or students in the professional schools, and con- sidered duplication of basic texts, sets, and periodicals desirable. Fourth, both agreed to maximize the number of unique research materials by avoiding unnecessary duplication. Finally, the directors recognized that if agreements were to be successful, they needed to be flexible and allow for adjustment and expansion. 24 Regarding the question of which sub- jects would lend themselves to coopera- tive collection development, members of the Committee on Intellectual Coopera- tion suggested two options: (1) concentration in each library of materials dealing with specialized problems or fields of knowledge in which one institution is primarily in- terested, and (2) subdivision of fields in which both institutions are inter- ested.25 In addition, they asked librarians to avoid duplicating specialized research materials, particularly expensive titles, large sets, and serials, where one copy in the area was sufficient, and to divide collecting. responsibility for state, fed- eral, and foreign documents. 26 Downs and Branscomb lost no time ap- plying to the GEB for a cooperative collec- tion development grant. In 1935 they received $50,000, which they divided equally between the two institutions. Although the Program of Cooperation presented two strategies for coopera- tion, the librarians decided that in this grant they would focus on materials re- quired by major disciplines that met the following criteria: (1) Strong departments in both insti- tutions should be chosen both because Cooperative Collection Development 473 such departments presumably are doing highly effective work, and be- cause the problem of coordinating the work of the two Universities must be solved in such areas. (2) The depart- ments must be ones which have shown an interest in and ability to correlate their programs with those in the other University. (3) The depart- ments should be · those which are believed to be of special importance to this region in an economic, social, or cultural direction.27 The disciplines they selected, following the recommendations of departmental chairmen at both universities, were botany, zoology, chemistry, physics, English, sociology, and economics. The librarians hoped that "these depart- ments will become an illustration and example to others in the two institutions, and the habit of mutual dependence on the other University induced by the ac- tive use of a considerable body of mate- rials in the other library will forward the whole movement of cooperation." Downs and Branscomb expected success in these key areas to lead to successful coopera- tion overall. They may also have recog- nized that the best strategy for winning a grant from the GEB, given its emphasis on uplifting the South, was to select dis- ciplines "of special importance to this region in an economic, social, or cultural direction. " 28 For this initial attempt librarians and faculty stressed two approaches. First, they divided materials on an ad hoc basis. The decision was "more or less arbitrary as regards basic sets, periodical files, and other material applying to the field as a whole." Second, they made the first of many efforts to cooperate system- atically on academic disciplines. In this case they divided responsibility for books, serials, and other library materi- als by the major subfields of each of these disciplines according to faculty research interests. 29 Faculty and librarians assigned specific subfields to each library. For example, in chemistry Duke took responsibility for biochemistry, paper and cellulose chemis- try, agricultural chemistry (particularly 47 4 College & Research Libraries tobacco), and food chemistry; UNC em- phasized chemical engineering, petroleum products, electrochemistry, and the history of chemistry. For English, they developed complicated divisions based on chrono- logie~! periods, authors, and genres.30 This .attempt at the systematic division of responsibility for the publications of major disciplines seemed to make sense at the time. Faculty and librarians may have chosen this approach because they conceived of academic disciplines in terms of their subfields. However, the systematic division of responsibility for books, serials, and other library materi- als by major subfields proved im- possible, because it weakened library support for the discipline as a whole and jeopardized scholars' ability to do re- search in their .specialties. In addition faculty interests changed over time, further undermining the stability of sub- fields as units of cooperative collection development. Therefore, this type of sys- tematic cooperation did not survive the grant. Indeed, it apparently provided a model of how not to cooperate, because librarians never divided traditional dis- ciplines by their major subfields again. Although this division of responsi- bility did not provide a long-term model for cooperation, the grant was successful in other ways. Librarians learned they could cooperate on an ad hoc basis for specialized and costly titles, such as mul- tivolume sets, long periodical runs, and newspaper backfiles. Indeed, ad hoc cooperation has been one of the most successful forms of cooperative collec- tion development over the decades and has been responsible for extending the number of unique holdings in the TRLN libraries significantly. The grant also fostered a cooperative mentality. As Downs and Branscomb wrote, ''There is now general acceptance of the idea of cooperative collections, and it is becoming general procedure to limit duplication of rare and expensive items in all fields." 31 Cooperative collection development efforts continued and multi- plied because librarians became com- mitted to cooperation and kept searching for ways to expand it. Their efforts, in November 1993 turn, succeeded because faculty ac- cepted cooperation as a ·given. Building on the momentum of this two- year grant, librarians next considered the cooperative acquisition of foreign, federal, and state documents. Although faculty and librarians had based the cooperative proposals funded by the GEB on faculty research interests, librarians, acting on their own, proposed a systematic division of government documents in 1937. Downs suggested to Branscomb that foreign documents should be concentrated at Duke, "because of the excellent start you have made in this field." Both libraries were to remain depositories for current federal publications (with librarians at UNC taking responsibility for filling in gaps of older materials). Because of the strength of UNC' s holdings, its librari- ans would assume responsibility for state documents. 32 But the faculty disagreed. They pro- duced a report arguing that because re- searchers at both institutions were engaged in the study of both local and foreign problems, "a division of function can never be made which will allocate to one the responsibility for domestic and to the other foreign, it is the belief of this Council that a division of library materi- als on this basis should not be attempted. We believe that a more satisfactory plan would be to endeavor to divide each area between the two libraries."33 The faculty version prevailed. Ultimately, faculty and librarians put into operation a more complex plan that divided responsibility systematically according to geography, subject (which often corresponded to is- suing agency), and publishing format, such as legislative journals.34 The initial proposal, faculty reaction, and final agreement revealed the impor- tance of basing cooperative agreements on academic programs and including fa- culty in their development. It also marked the first time that faculty and librarians divided collecting responsi- bilities geographically, an approach that played a major role in later cooperative efforts. The agreements for government docu- ments worked well. Their success de- monstrated that systematic cooperation for materials of interest to faculty in many departments worked, if the items were not central to their teaching and research specialties and were distinct in format or method of acquisition. Al- though modified and expanded over the decades, cooperative agreements for government publications continue to be a major focus of cooperation among the Research Triangle university libraries. Beginning in the late 1930s, Duke and UNC received a series of grants from the North Carolina Division of Cooperation in Education and Race Relations to buy library materials on "all aspects of Negro history, literature, education, economic and social conditions, religion, health, etc." Within a few years these funds created a combined African-American collection of 10,000 volumes, with al- most no duplication except for recent books that would be in demand on both campuses.35 The grants demonstrated that new areas of interdisciplinary re- search-even those of special interest to faculty at both universities-could be fruitful areas of cooperation. Librarians included interdisciplinary cooperation in their next grant proposal, perhaps be- cause of their success here. In their application to the GEB in 1938, librarians recognized the tentative and experimental nature of the original cooperative agreements, the necessity of winning the support of all parties af- fected, and the importance of avoiding the appearance of arbitrariness. At the same time the application showed that they had assimilated important lessons from their earlier grant and the agree- ments for government documents. In their search for a systematic model of cooperation, librarians shifted their em- phasis from disciplines that were strong at both universities to subjects repre- senting unique academic and collection strengths. For other subjects they pro- posed cooperating on an ad hoc basis.36 On the basis of unique academic strengths, UNC took responsibility for geology, music, Indo-European linguis- tics, library science, and Romance lan- guages. Duke concentrated on forestry, Cooperative Collection Development 475 fine arts, mathematics, religion, and Oriental history, philosophy, and litera- ture. Where both supported strong pro- grams, librarians asked for funds to develop collections in fields involving more than one academic discipline. They chose social history, which was of inter- est to departments of sociology, econom- ics, and history; political science, which included international law, and federal, state, and local government; and classi- cal studies, which included history, lit- erature, and art. The multidisciplinary nature of these fields represented a different approach from the previous grant, which had focused on traditional academic disciplines. Finally, building on their earlier successes, Duke and UNC proposed using grant funds to con- tinue cooperation in government docu- ments, bibliography, and newspapers, which were of interest to the research community as a whole. 37 Although Duke and UNC did not receive this grant, the hope of securing outside funding pro- vided librarians with the impetus to develop approaches that would form the basis of future cooperation. By the end of the decade, librarians could look back on their efforts with a sense of accomplishment. They had es- tablished the rationale and principles of cooperation that continue to this day. They had identified the two major types of cooperation: ad hoc and systematic cooperation. Librarians had successfully applied the ad hoc approach to costly items and materials for special collec- tions. They had systematically divided books, serials, and other library materi- als of interest to many disciplines that were characterized by distinctive format or method of acquisition, particularly government documents and news- papers. In addition, faculty and librari- ans enjoyed enhanced bibliographic and physical access to each others' collec- tions and had developed a spirit of cooperation that would motivate them to maintain existing programs and create new ones. The librarians were aware, however, of what they had not yet accomplished. There was little intercampus communica- 476 College & Research Libraries tion, no ongoing coordinated growth of the collections, and they therefore had not built a well-rounded collection to be used by the whole region.38 In addition librarians had not yet developed a system- atic approach to cooperation for specific subjects over the long term. The creation of that model would be the achievement of the next decade. THE 1940s: CREATING THE AREA STUDIES MODEL In the 1940s an emerging interdiscipli- nary field suggested a systematic way to cooperate on a subject. As the Allies suffered reverses during the early part of the Second World War, Sturgis E. Leavitt, professor of Spanish at UNC, believed that "the hope of civilization lies in the New World. Cultural relations between the Americas are therefore more impor- tant now than ever before."39 He, his col- leagues, and librarians at Duke and UNC who were interested in this developing field, proposed expanding cooperative collection development to cover research materials from and about Latin America. They also suggested in- cluding a third institution, Tulane Uni- versity, which had already developed strong holdings on the area.40 On the basis of faculty interests and li- brary holdings, faculty and librarians initially agreed to divide collecting re- sponsibility by subject. Tulane would cover Caribbean archaeology, Indian (Na- tive American) languages, modernismo, and the influence of U. S. literature on Latin American literature. Duke would collect the cultural history of the colonial period and Brazilian studies. UNC, for its part, would acquire materials on bibliography, library science, Spanish American languages, Spanish American literature in the United States, folklore, constitutional and political history, the eighteenth century, and the cabildo. Of greater importance, however, were the provisions for each university to as- sume responsibilities based on geogra- phy, which both faculty and librarians considered "logical and fair." Building on the strengths of their collections, Tu- lane took the Middle American region, November 1993 including Cuba and the Antilles; Duke emphasized Brazil, Ecuador, Peru, Bolivia, and Colombia; while UNC ac- cepted responsibility for Chile, Para- guay, Argentina, and Uruguay.41 In 1940 the Rockefeller Foundation gave Duke, UNC, and Tulane a grant of $75,000 ($25,000 each) to be spent for Latin American studies over a five-year period.42 When faculty and librarians came to work out policy and procedures for implementing the grant, they aban- doned subject arrangements in favor of geographical divisions. In fact, the only mention of subject divisions was that ''North Carolina will develop its collec- tion of folklore without geographic re- strictions."43 The geographic model of cooperation eventually became preemi- nent among cooperative strategies for dividing foreign area studies. There are many explanations for the continuing success of this paradigm. The simplicity of administering the geo- graphic divisions was a major attraction. Faculty and librarians found it easy to remember such clean divisions. Another reason for success was the faculty's realization that neither institu- tion had resources to build major collec- tions for a new area of research. They saw cooperation as the best way to ac- quire a wide range of materials in an emerging field the libraries could not afford to support otherwise. The con- tinuing importance of Latin American studies over many decades ensured the survival of these cooperative agree- ments, even during times of limited funding. Perhaps the major reason for success, however, was inherent in the newness and interdisciplinary nature of Latin American studies. The materials were important to faculty and students in many departments, yet no academic de- partment had a vested interest in the area that corresponded to a standard dis- ciplinary subfield. As a consequence, librarians had freedom to interweave the collections, creating a coordinated whole. They anticipated this goal from the start: "Subject interests of faculty . . . which reach across the geographical line of di- vision will be met by the agreement that each library, in buying in its allotted field, will consider requests from the other faculty on the same basis as re- quests from its own." 44 They also insti- tuted a liberal interlibrary loan policy to mitigate any hardships users might ex- perience as a result of this geographic division of responsibility.45 Finally, cooperative collection develop- ment for Latin America worked not only because it had long-term faculty backing but also because library administrators hired staff to implement it. The first pro- vision for spending the Rockefeller money stated that "each institution will appoint a coordinator who will act as the central agent for his university. Through him all matters affecting the individual institution and the cooperating institu- tions will be cleared." 46 As part of the agreement, UNC sought "to employ a library assistant ... [to] facilitate the handling of exchanges, of purchases from South American dealers, and in coordinating the work with the qther two cooperating libraries."47 Later, UNC hired a Latin American bibliographer, the first full-time collection develop- ment officer with specific subject re- sponsi,bilities in the Research Triangle university libraries. While developing the cooperative model for Latin America, librarians at Duke and UNC continued to search for strategies that would work for other sub- jects. As part of their efforts, they divided collecting responsibility for a number of fields in the early 1940s. Many of these subjects represented unique academic strengths. Duke, for example, had the only programs in religion, medicine, and for- estry, while UNC had unique programs in library science, public health, geology, folklore, and linguistics. Other divisions were based on the strengths of library collections. Duke had exceptional hold- ings of American literature, for example, while UNC had a special collection of North Caroliniana. A few of these sub- jects represented the librarians' continu- ing efforts to find a way to divide subject disciplines of interest to both institu- tions. In such cases they did not assign Cooperative Collection Development 477 responsibility for subfields based on fa- culty interests, as they had in the 1930s, but on broader categories such as early German literature (UNC) and late Ger- man literature (Duke).48 Librarians continued the systematic division of responsibility for publica- tions with distinct formats and methods of acquisition that were of interest to faculty in many departments or the aca- demic community as a whole. These materials included state government documents and the catalogs and annual reports of colleges and universities. For documents they based cooperative agreements on geography, subject/issu- ing agency, and format. For colleges and universities, they used issuing agency. Duke collected catalogs and annual re- ports from private institutions; UNC, those from public ones.49 The 1940s were extraordinarily success- ful. During this decade, faculty and librar- ians at Duke and UNC developed one of the major types of systematic cooperation, the area studies approach. Librarians also learned that they could continue to build complementary holdings based on unique academic or collection strengths. In addi- tion they continued agreements for many types of materials of general interest that were distinct in format or method of ac- quisition. Finally, these years demon- strated that the subjects and kinds of materials identified in the previous de- cade as good candidates for ad hoc cooperation were indeed appropriate choices and worked over the long term. A series of grants from the Carnegie Corporation in the early 1940s helped libr~rians maintain and solidify these cooperative collection development agreements. They used these funds not only to honor Duke and UNC faculty requests, as they had done with the ear- lier Rockefeller grants for Latin America, but to meet the needs of faculty throughout the region within their re- spective areas of responsibility. 50 THE 1950s: EXTENDING GENERAL COOPERATION When grants for cooperative collec- tion development ended in the late 1940s 478 College & Research Libraries and no new outside funds replaced them, enthusiasm for new cooperative initiatives waned, as well, although ex- isting forms of cooperation continued. 51 Then, in 1953, the presidents of Duke and the UNC system (which included the State Agricultural and Mechanical College at Raleigh, later NCSU, the Woman's College at Greensboro, and the Chapel Hill campus) appointed faculty and librarians from each of their institu- tions to an Inter-University Committee on Library Cooperation. Their purpose was to reinvigorate and expand the cooperative programs. 52 Representatives from the State Library joined them soon after. Librarians from the five institutions tried to coordinate their acquisitions policies. 53 They contributed information about their holdings to union lists of pe- riodicals and agreed to allow faculty and graduate students to borrow books directly from each other. 54 Staff from the four universities also agreed to meet reg- ularly to implement the policies of the Inter-University Committee. 55 Despite these initiatives with nearby libraries, only Duke and the Chapel Hill campus were involved in cooperative collection development programs. In 1956 librarians at the two institutions codified their existing agreements. The results, which were remarkably similar to those existing in the early 1940s, un- derscored the success of the original principles and types of cooperation over two decades. Librarians retained a systematic divi- sion of responsibility for government documents but revised specific com- ponents of the agreements. Because "UNC has since developed more aggres- sive and extensive collecting," it took responsibility for all state documents. Duke, which had recently established a Commonwealth Studies Center, agreed to be responsible for Canadian govern- ment documents, with the exception of geological publications (which UNC continued to collect comprehensively). 56 This division is remarkably similar to the one the librarians proposed in the 1930s, but which faculty did not accept, be- November 1993 cause it seemed too arbitrary and was based only on collection strengths. This time, however, faculty could accept such a division, because it matched academic programs as well as library collections. There were also some new initiatives. Librarians divided depository responsi- bility for the publications of various in- ternational agencies. In addition they attempted to divide responsibility for census statistics and ethnographic pub- lications geographically, following the successful model for Latin American stu- dies. Duke was to collect material for the Far East west to India, the British Isles and the Commonwealth (again building on its Commonwealth Studies Center), its Latin American countries, and the USSR. UNC agreed to collect titles from the Near East (west of India to Europe), its Latin American countries, Africa, and the areas of Europe not covered by Duke.57 These agreements were super- seded by more comprehensive arrange- ments, as area studies gained in importance in the 1960s. The rapid growth of microform pub- lishing in the 1950s presented librarians with a financial challenge that they turned into a major cooperative success. They coordinated the purchase of major microform sets on an ad hoc basis to reflect faculty interests and collection strengths. In a few cases, such as the Landmarks of Science, they shared the cost of a set and placed it in the most appro- priate library. The cumulative results of these cooperative efforts became evident when the TRLN union list of microform collections appeared in 1986. Only 3 per- cent of its nearly 1,200 entries repre- sented materials held at more than one institution, and only about 1 percent was held at all three. Librarians' attempts to cooperate sometimes ran afoul of the faculty's need for materials. Faculty at Duke vetoed a proposal to divide responsibility for ex- pensive foreign government serials, in this case, the British sessional papers and the French Journal Officiel. They said they needed both subscriptions on their own campus. But Gertrude Merritt, chief of the Processing Division at Duke, over- ruled them. She urged Harry Bergholz, the chief bibliographer at UNC, to con- tinue subscribing to the Journal Officiel, while Duke subscribed to the sessional papers. 58 This incident illustrates one of the major dilemmas of cooperative collec- tion development. Is it better to allocate resources to meet cooperative responsi- bilities and thereby build a more com- prehensive joint collection, or to satisfy immediate faculty needs by giving pri- ority to local needs? In this case librari- ans were able to realize broader cooperative objectives. In other in- stances, faculty pressure has been so strong that librarians have had to dupli- cate expensive materials. On the whole, however, faculty have been willing to support cooperation. THE 1960s: EXTENDING AREA STUDIES COOPERATION The growth of national programs for cooperative acquisitions in the 1960s led librarians to review existing agreements between Duke and UNC. In particular they weighed ~heir obligations to con- tinue local cooperative programs against participation in national endeavors, such as the Farmington Plan. In 1961 Benjamin Powell, university librarian at Duke, wrote to Jerrold Orne, his counterpart at UNC, asking him whether the two libraries should jettison their cooperative agreement for Latin America in favor of a national program. According to their existing agreement, Duke and UNC covered all the Latin American countries selectively. Under the Farmington Plan, they would work with only a few countries, but in greater depth, and rely on other libraries in tile United States for research materials from other nations. Orne's response il- lustrated the value that librarians in the Research Triangle placed on their local arrangements: ... I do believe that we both have, first, a responsibility for mutual ac- cord on the division of fields in the Latin American countries closely tied to our teaching programs and, secondly, that any participation in a national pro- Cooperative Collection Development 479 gram must be related to our individual responsibilities first. ... If what we do fits into the Farmington Plan, I will be happy to be named with it, but if it does not, I cannot be too much concerned. 59 Duke and UNC did not participate in the Farmington Plan, which lacked roots in the participating institutions and ulti- mately withered away.60 By contrast, the Duke/UNC cooperative program for Latin America met faculty needs at both universities and thrived. Indeed, librari- ans at Duke and UNC joined the Latin American Cooperative Acquisitions Pro- gram two years later, because they could build their national contribution on local cooperative agreements.61 The different fates of these projects demonstrate the importance of the principle of self-inter- est as the foundation for cooperation. Cooperative ventures that do not grow out of the academic programs or collec- tion strengths of individual institutions will not survive. During the 1960s, new area studies programs came into existence at both universities. Faculty and student needs for materials from and about Africa, East Asia, Eastern Europe, and South Asia, in particular, strained available funds. In meeting these new demands for re- sources, librarians drew on their ex- perience with a geographical division of Latin America as a model for successful cooperation. Faculty developed a joint Duke/UNC- CH graduate program in Russian and East European history in the early 1960s. Librarians supported it by dividing re- sponsibility for Russian and Soviet mate- rials in the humanities and social sciences, while limiting the acquisitions of bOOks and serials from other East European countries to titles related to Russian ·stu- dies.62 Later, they divided responsibility for the Slavic countries of Eastern Europe. Librarians at Duke took responsibility for Polish materials; their colleagues at UNC did the same for Czech publica- tions; while librarians at the University of Virginia agreed to cover titles in South Slavic languages for certain subjects. About the same time, librarians for- malized agreements for Africa. As in the 480 College & Research Libraries case of Latin America, they based their cooperative responsibility on academic and collection strengths. Because Duke had supported a Commonwealth stu- dies program since the mid-1950s and its libraries held many publications from these countries, librarians there took re- sponsibility for the English-speaking areas of Africa. Librarians at UNC com- plemented Duke's efforts by collecting specialized materials for the Arab north and some of the French-speaking areas of sub-Saharan Africa. Eventually, they assumed responsibility for nearly all the non-Anglophone countries of the conti- nent in order to divide costs equitably.63 By the end of the decade, librarians began to cooperate on Asian materials. Here again the geographic model pre- vailed. Although they decided that both institutions would acquire titles to sup- port East Asian studies in Western lan- guages, librarians divided responsibility for materials in Chinese and Japanese.64 Eventually, librarians at UNC accepted responsibility for acquiring and pro- cessing titles in Chinese, while those at Duke did the same for Japanese. Librarians also formulated agree- ments for other areas of the world. Be- cause of Duke's commitment to acquire Commonwealth materials and its large- scale participation in the PL-480 program, its librarians assumed responsibility for building research collections in South . Asian studies and hired a South Asian bibliographer during this period.65 Fol- lowing the same logic, Duke's librarians eventually assumed responsibility for Australasia, Canada, and the English- speaking countries of the Pacific and West Indies. In response, librarians at UNC reduced their collecting of materi- als from all these Commonwealth coun- tries to a basic level. The geographical model of coopera- tion worked .as well for these areas as it had for Latin America, and for the same reasons. One measure of the extent of the success of this model is evident in the latest union list of current foreign news- papers at Duke and UNC-CH, which dates from 1988. It revealed that only 21 percent of the 192 subscriptions were November 1993 duplicates-and these tended to be heavily used items such as Le Monde. Moreover, the duplication rate drops to only 4 percent when West European newspapers are excluded. THE 1970s: PROVIDING A STRUCTURE FOR COOPERATION The decade began inauspiciously with minor elaborations and expansions of the agreements for area studies. This sit:.. uation changed a few years later, when library administrators and staff created a new framework for cooperation, and outside agencies contributed major funding for cooperative projects. In response to inflationary increases in serials prices and concern about whether collections could support research in the rapidly growing Research Triangle Park, · university librarians James Govan, UNC-CH, and Connie Dunlap, Duke, appointed a committee to explore addi- tional cooperative ventures.66 The group soon invited librarians at NCSU to par- ticipate as full partners, and together they established the Triangle Univer- sity Libraries Cooperation Committee (TULCC). Within a few years TULCC be- came the Triangle Research Libraries Network (TRLN), the current umbrella organization governing all cooperative endeavors, including collection develop- ment, bibliographic' and physical access, and automation.67 These organizations provided a struc- ture for regular communication that nur- tured cooperation. During the course of increasingly frequent joint meetings, librarians became aware of shared inter- ests ·and opportunities for cooperative action. As a consequence, they believed they were in a strong position to secure grants for collection development, bibli- ographic control, user studies, and pro- gram evaluations.68 Librarians rec~ived two cooperative collection development grants for $250,000 each from the Title 11-C pro- gram; one during 1978/79 and another for 1980/81. They followed the success- ful cooperative models of the past in spending these funds. Librarians pur- chased materials in areas of unique aca- demic and collection strengths. At UNC- CH, for example, they purchased special- ized grammars and dictionaries to support research in linguistics. Librari- ans also made a number of ad hoc pur- chases of expensive titles, especially microfon:ri collections and newspaper and periodical backfiles. At NCSU, for example, they purchased the U.S. patents collection in microform. Finally, librari- ans at Duke and UNC-CH used the funds to enhance foreign area studies holdings, concentrating on the countries for which their institutions were re- sponsible. Following the pattern of ear- lier cooperative collection development grants, these acquisitions represented unique additions to the consortium's collections. The importance of these joint collec- tion development grants for advancing cooperation between Duke, NCSU, and UNC-CH cannot be overstated. They led to a reaffirmation and refinement of pre- vious agreements, the extension of cooperation to new areas, and the full integration of NCSU into the coopera- tive programs. They also helped make cooperation a central concern of collec- tion development. In order to implement the grants most effectively, a broad representation of selectors from all three campuses met quarterly to discuss their projects. In the past, cooperation had been the preroga- tive of library administrators. For the first time, as a result of these grants, librarians at the operational level began to participate directly in planning cooperative programs. By coincidence the TRLN institutions were installing new collection development staff around this time. The Title 11-C grants enabled these individuals to develop a cooperative mentality that they now consider a normal-rather than excep- tional-way of going about their collec- tion development duties. Finally, these grants enabled coopera- tion to proceed at a much faster pace than would have been possible other- wise. As John Shipman, university bibli- ographer at UNC-CH, pointed out in his final report on the second Title 11-C Cooperative Collection Development 481 grant, "there have been few periods during which [cooperative] activities have reached the level of those of the past three years." 69 In order to guarantee continued cooperation, Shipman has allocated an average of $50,000 annually since the first Title 11-C grant, solely for this pur- pose. These funds and others that have since become available for cooperative purchases have proved to be an excel- lent, continuing incentive. Over the past dozen years they have totaled close to a million dollars at UNC-CH alone. 70 The availability of this money heralded the intensification of coopera- tive collection development efforts during the next decade. THE 1980s: EXPANDING COOPERATION TO CORE AREAS The success of the cooperative pro- grams for area studies and the enthusi- asm generated by regular meetings led the bibliographers with major responsi- bility for Western Europe and the United States to develop cooperative programs for their areas of the world. Because materials published in Europe and North America are so central to the scholarly enterprise in this country, cooperative decisions for publications from these areas have been more complex. The cooperative ventures for Western Europe took place during the middle of the decade. The bibliographers for Western Europe at UNC-CH and Duke planned cooperative programs for French regional history and German literature.71 The first program, established in 1984, covered French regional materials for the Triangle by assigning collecting re- sponsibility based on a geographic divi- sion of France.72 It applied only to lower priority titles. UNC-CH accepted re- sponsibility for departements in the southern half of France and Paris; Duke, for the rest of the country. A proposal to cooperate on German belleslettres also dated from that year. In order to expand the coverage of contem- porary German literature, John Rutledge, bibliographer for Western European re- sources at UNC-CH, suggested that both 482 College & Research Libraries universities collect major authors, but that Duke acquire works by secondary authors whose names began with the letters A-Land Austrian writers, while UNC-CH took responsibility for those whose names began with M-Z and East German and Swiss writers.73 Both programs ran into problems. By the late 1980s, when funds could no longer cover higher priority titles in major fields, librarians stopped buying minor French regional histories. They revised the pro- gram, however, to divide responsibility for major regional publications along the same geographical lines. The proposal to collect German authors cooperatively foundered when Helene Baumann, West European bibli- ographer · at Duke, recognizing that Duke did not have the academic pro- grams to justify such a broad scale of collecting, stated that her "primary man- date is to buy what Duke faculty and students need now and in the future." In the same letter she suggested building to strengths at each institution, with Duke buying specialized materials on German Baroque literature and German-Ameri- cana, because of the library's strong holdings in these areas, while UNC-CH emphasized German language, pedagogy, and folklore, which built on its academic and collection strengths. Rutledge agreed with her suggestions, and cooperation on this basis has worked.74 The success of the revised agreements for German language and literature and for French regional history once again revealed the importance of - tying cooperation closely to academic pro- grams and collection strengths rather than using abstract or arbitrary criteria. The experiment in French regional his- tory also demonstrated that successful long-term cooperative programs cannot include subjects and materials that are too marginal to survive periods of tight funding. During the late 1980s librarians made their first attempts to cooperate in a major way on materials related to the United States. The need to increase coverage of the American South arose when faculty and administrators at November 1993 UNC-CH proposed an institute of Southern studies modeled on the foreign area studies programs. Realizing that UNC-CH did not have the funds to ac- quire all the relevant materials its re- searchers would need, librarians turned to their colleagues at Duke and NCSU for help. Their common goal was to build a joint collection for Southern stu- dies that would become the major center for scholars and students undertaking comparative and multistate research on the region. Because this initiative covered all sub- jects and formats and involved dozens of selectors in many disciplines at three universities, staff met together for two years to exchange information and dis- cuss possible agreements for various subjects and formats. They learned the strengths and weaknesses of each other's collections, where they dupli- cated one another, and where there were gaps. Once again, the prospect of outside funding acted as a powerful incentive for them to agree on divisions of re- sponsibility for materials from and about the region. Several factors complicated the dis- cussions. In the first place, for historical and cultural reasons librarians in the Re- search Triangle have always collected in- tensively on the region. In addition, faculty and students at the three institu- tions have had strong research interests in the South for decades. Whenever people care deeply about an area or sub- ject, cooperative collection development agreements are more difficult to ne- gotiate. Logistically, cooperative agreements for the South presented a challenge, be- cause most of the scholars doing re- search on the region were at UNC-CH. By contrast, Duke had the largest en- dowment with which to purchase South- ern Americana, but fewer faculty studying the South. NCSU wanted to be involved, but was not sure how its em- phases on science and technology would fit in with the usual cooperative focus on the social sciences and humanities. The organization of the libraries, their selectors, and selection sources also complicated the negotiations. Up to this point, formal cooperative agreements had been limited to collections in the main libraries. Because of the all-encom- passing na~ of collection development for Southern Americana, cooperative ef- forts had to involve librarians in both cen- tral and branch libraries. In developing these agreements, librarians needed to be sensitive to the complex relationships between faculty and staff in branch li- braries and their lack of experience with cooperation. The types of selectors at the three in- stitutions further complicated the process. Until this project, cooperation had involved primarily full-time collec- tion development officers who covered many fields. The scope of their responsi- bilities gave them a broad perspective on subjects, users, and overall library re- sources. They also had enough autonomy and authority to develop cooperative agreements. Most librarians involved with Southern Americana were part-time selectors responsible for one discipline. Because of the nature of their responsi- bilities, their perspectives, and their lack of experience with cooperation, they were also less aware of the ways it could benefit the larger community. In addition, the sources that selectors used to identify items for acquisition had an impact on the materials they could cover. Librarians at UNC-CH and NCSU used Library of Congress proofslips and cataloging-in-publication forms, which encompass a broad array of materials related to the South and include many nontrade and other specialized titles. Duke's selectors relied primarily on ven- dor forms, book reviews, and user sug- gestions, which provided narrower coverage of the universe of publications, but met their collection development needs. Finally, collecting priorities differed, resulting in varying commitments from each institution. Duke's selectors em- phasized special collections-and had the endowed funds to afford such mate- rials. NCSU' s collection development of- ficers preferred to concentrate on a limited number of academic and collec- Cooperative Collection Development 483 tion strengths, such 1:ts climatology, and a few formats, such as dissertations. Librarians at UNC-CH decided to focus on title-by-title selection, because of their ability to identify a broad spectrum of publications and because of the wide range of topics their faculty and students were researching. The cooperative agreements for the American South that emerged from these meetings covered all subjects and formats and incorporated lessons librar- ians had learned over the decades. The divisions of responsibility met the pri- orities and needs of each institution and were therefore likely to continue. Where a university had strong or unique aca- demic programs or collecting strengths, librarians based responsibilities on them. Because NCSU had a college of textiles, for example, librarians there as- sumed responsibility for materials on this topic. Their colleagues at UNC-CH took responsibility for folk music, be- cause of that library's special collection of those materials. Where more than one institution had academic or collection strengths, librari- ans divided responsibility on an ad hoc basis for expensive titles, such as micro- form sets, or systematically, by geogra- phy (for newspapers) or format. In the case of regional belleslettres, for example, Duke agreed to collect small press mate- rials, while UNC-CH concentrated on little magazines. The agreements repre- sented an equitable division of costs, as they had for the area studies programs. More broadly, the cooperative agree- ments for Southern Americana revealed that librarians could cooperate in inter- disciplinary areas of intense interest to many constituencies and do so even in times of financial austerity. Indeed, when programs are organic and build on academic programs and collection strengths, library priorities, and organi- zational structures, they are more likely to be successful in the long run than are arbitrary divisions of responsibility that ignore these crucial factors. The success of the cooperative efforts for Southern Americana bore fruit in 1991/92 and 1992/93, when the three 484 College & Research Libraries libraries received two Title 11-C grants of nearly $600,000 to acquire materials documenting the contemporary South. In particular, readers of the grant liked the cooperative nature, detailed plan- ning, and comprehensiveness of the pro- posal. Librarians are expanding on this success by pursuing other grants for Southern Americana. The ability of librarians to work to- gether on cooperative projects for West- ern Europe and the American South was significantly enhanced by a shared on- line catalog that became operational mid-decade. Just as library cooperation in the 1930s owed its success to bibliographic and physical access to the collections, cooperative collection development in the 1980s advanced for similar reasons. A joint online union catalog made the re- sources of the three libraries available to all their users. During this period TRLN librarians also extended direct borrow- ing to undergraduates; expedited inter- library borrowing, including the faxing of priority requests; and wrote special lending agreements for East Asian ver- nacular materials related to cooperative programs. , Advances in shared automation also made ad hoc cooperation possible for a wider range of materials by significantly lowering the cost of determining what each library held. These developments contributed to the increasing importance of collection strengths in influencing cooperation. Finally, they made library cooperation more acceptable to faculty, students, and librarians, and helped users and selectors view the TRLN col- lections as ultimately one. THE 1990s: LOOKING TO THE FUTURE Although the sciences had been part of the first cooperative collection develop- ment grant in 1935, they vanished al- most immediately as an area of cooperative endeavor. For fifty years cooperation remained confined to the humanities and social sciences. In re- sponse to a lack of funding for acquisi- tions and the tremendous increases in the number and cost of scientific, techni- November 1993 cal, and medical serials in the mid-1980s, librarians took a renewed interest in cooperative collection development in the sciences. The pressures generated by these forces led selectors of scientific materials to begin meeting together in 1988. Cooperation in the sciences received a further boost when administrators at NCSU, the university with the strongest focus on science and technology, appointed full-time science bibliographers with re- sponsibilities for large subject clusters. Like the subject and area bibliographers for the humanities and social sciences, these full-time collection development of- ficers assumed a leading role in planning and coordinating cooperation. Science selectors have been supported in their efforts by the creation of a struc- ture for incorporating specialized areas into the cooperative collection develop- ment organization. In order to broaden the scope of cooperation, the TRLN Col- lection Development Committee added roundtables covering non print materials and government documents in 1990. The following year it established a round- table for medical, scientific, and techno- logical fields. Now the science librarians have a forum and context to develop cooperative agreements. Cooperative collection development in the sciences received additional en- couragement from a two-year grant the Council on Library Resources (CLR) awarded TRLN in 1991. Under this grant administrators, faculty, and librarians are identifying the obstacles to coopera- tion in the sciences, determining how to overcome them through advanced tech- nology, creating new organizational ar- rangements that ensure ongoing faculty participation, and discovering the kinds of strategies that might enable TRLN to provide advanced electronic informa- tion services.75 Another aspect of the CLR grant in- volves the development of administra- tive structures to formalize cooperative agreements. When the cooptrative pro- grams began in the 1930s, they were part of an overall institutional emphasis on intellectual cooperation. Since then, the heads of the universities have continued to encourage library cooperation. The Memorandum of Understanding establish- ing TRLN bears the signatures of the uni- versities' presidents and chancellors, and the provosts serve on its governing board. Although university administrators have supported all general cooperative agree- ments, librarians have never asked for- nor received-faculty or official admin- istrative approval for specific cooperative collection development programs. Over decades of cooperation librari- ans have run into problems on two counts because they lacked faculty in- volvement and formal administrative approval. In the first place, faculty have occasionally exerted pressure to change agreements that did not match their re- search needs. In the second place, admin- istrators and faculty have established academic programs in areas that librarians had ceded to cooperating institutions and therefore could not support adequately. In such cases, university administrators would have been better served if they had been aware of the cooperative agreements and the economic consequences of abro- gating them. TRLN librarians are using the CLR grant to create a way for faculty to participate in the development of coopera- tive agreements and for university admin- istrators to endorse them formally. This type of faculty and administrative involve- ment should increase the likelihood of successful long-term cooperation. The continuing proliferation and growing importance of interdisciplinary research throughout the academy pre- sents librarians with many new oppor- tunities for cooperation. Librarians at Duke, NCSU, and UNC-CH, for example, are using the CLR grant to discover if the recently created Center for World Environ- ment and Sustainable Development- which involves over 150 faculty from all three Research Triangle universities- might provide a model for cooperative collection development in the sciences. OBSERVATIONS ON SUCCESSFUL COOPERATION For more than half a century librarians at the Research Triangle universities Cooperative Collection Development 485 have wrestled with the key issues of cooperative collection development: Why should librarians cooperate? Which aca- demic disciplines, subjects, and types of materials make good candidates for cooperation? How do librarians, faculty, and. administrators work together to develop viable programs? In this article we have analyzed our efforts to answer these questions. We offer the following synthesis of the insights we have gained as a guide to help others create equally effective cooperative collection develop- ment programs. Rationale for Cooperation The goals of cooperative collection development are institutional excellence and enhanced service to users. Adminis- trators, faculty, and staff rarely have the resources to support academic programs and library collections at the level they envision. They must therefore seek in- novative approaches to advance local aspirations and meet local needs over the long term. Cooperative collection development is the best-and increas- ingly the only-way to realize these goals. If cooperation is to succeed, it must therefore emphasize institutional ad van cement and enhanced service to users rather than saving money. 76 Librarians can achieve these goals by developing cooperative programs ·that build interlocked collections. This strategy extends the number of unique titles available to users. Materials that librari- ans at one institution cannot afford or think are inappropriate may be available from other members of the consortium. This approach also minimizes the unnec- essary duplication of materials. By coordinating their collections, librarians do not need to duplicate specialized research materials and can use their funds to buy titles that are more central to academic programs and collection strengths. The resulting interdependent collec- tions provide a breadth and depth of coverage that would be impossible for individual institutions to achieve on their own. Eventually, cooperating li- braries become resources both for their 486 College & Research Libraries institutions and the entire country. These ideas have been central to cooperative collection development among the Re- search Triangle university libraries from the beginning. 77 Principles of Successful Cooperation Librarians at the Research Triangle universities have identified several prin- ciples that have served their cooperative programs well. They include institu- tional self-interest, academic and collec- tion strengths, audience and level of use, the centrality of subjects and materials to the local scholarly enterprise, and the way programs change over time. Librarians have learned that coopera- tion must spring from institutional self- interest and that agreements must grow organically out of academic programs and collection strengths. Only by grounding cooperative responsibilities in this way can librarians create viable programs. 78 If they divide responsibilities too ab.., stractly or arbitrarily and do not tie them to programs or collections, cooperation will not survive.79 It follows, then, that because each participant must believe that cooperative programs serve its self- interest, cooperative programs must be viewed as mutually advantageous by all involved, although the benefits do not have to be absolutely equal. Librarians should therefore accept collecting re- sponsibilities within regional or national cooperative programs only when they base them on the needs of their local insti- tutions, because only then can their insti- tutions be held truly accountable for fulfilling their obligations. 5° Following this principle, librarians have discovered that they need to build agreements on what their library can and wants to contribute to cooperation. Colleagues at cooperating institutions cannot force each other to assume ob- ligations nor restrict what they can ac- quire.81 Librarians have also learned to limit cooperative efforts to research materials. They specifically have excluded under- graduate and heavily used graduate titles, and considered the duplication of basic texts, sets, and serials desirable. 82 November1993 Through decades of trial and error, librarians have come to realize that the subjects and materials covered by cooperative agreements must not be so central to research that faculty insist they be available locally, nor so marginal to it that tight funding jeopardizes a pro- gram's existence. If librarians accept cooperative responsibility for areas that are too peripheral to academic programs or library collections, the agreements will not survive the hard financial times that institutions periodically face. No matter how well intentioned, when funding cuts threaten major programs, cooperative agreements for materials at the periphery perish. Finally, if the cooperative programs are to remain viable, librarians have rec- ognized that they must be flexible. 83 As programs on campus change, new fa- culty research interests develop, or new collecting opportunities arise, coopera- tive agreements require modifications. Types of Successful Cooperation Over the years librarians have iden- tified two major kinds of cooperation. The ad hoc approach is one of the most basic forms of cooperative collection development; it is also one of the most successful. Systematic cooperation is more complex and more limited in its applications. 54 The earliest attempts at cooperation used the ad hoc approach. While it can be applied to all subjects and kinds of materials, ad hoc cooperation works best in exceptional cases, primarily for ex- pensive titles. Appropriate candidates include large microform collections, costly periodical subscriptions, domes- tic and Western European newspapers, extensive serial backfiles, substantial multivolume sets, and items for special collections.85 The high cost of materials in these categories justifies the time librarians must spend negotiating the decision to purchase them. The ad hoc approach to cooperation is not efficient for the regular, ongoing selection of books and serials, however. Systematic cooperation for books, se- rials, and other library materials works where institutions have unique aca- demic programs or library collection strengths.86 It is also viable in instances where more than one institution sup- ports strong academic programs or li- brary collections that are of interest to many disciplines but not central to any single one. Because these rna terials are important but not crucial to disciplinary subfields, it is politically possible for librarians to build cooperative programs for them. Materials that lend themselves to systematic cooperation include those that are distinct in format or method of acquisition, those that support foreign- area studies, and those that are inter- disciplinary in nature. In all these cases, once librarians agree to cooperate, they do not need to consult with their co- operative collection development part- ners on each title. One of the models for systematic cooperation consists of materials that are distinct in format or method of ac- quisitions. Government publications are excellent examples of this type of cooperative collection development. Li- brarians can divide responsibility by geography, subject, format, or issuing agency. Electronic resources may also provide opportunities for systematic cooperation. Area studies materials also make ex- cellent candidates for systematic coopera- tion, particularly titles published in foreign countries. librarians can accept- or avoid-responsibility for these areas, based on academic programs or collection strengths. If they decide to share responsi- bility with another institution, a geo- graphical division works well, because it is clearly defined and easy to remember. Indeed, with a few minor adjustments, the geographical division of responsi- bility for materials from and about Latin America, for example, has been success- ful for half a century. 87 Systematic cooperation is more diffi- cult for Western Europe and the United States, because materials from and about these parts of the world are more central to the scholarly enterprise in this country. One possibility is to develop an interdisciplinary approach to books, se- Cooperative Collection Development 487 rials, and library materials that divides coverage according to academic and col- lection strengths, format, and geogra- phy, as we have done with Southern Americana and hope to do for en- vironmental studies. As new areas of in- terdisciplinary research become more prominent, librarians will have more op- portunities to explore this type of cooperation. By contrast, agreements based . on major academic subfields and specialties will not work, except on an ad hoc basis. Faculty need to have materials that are closely related to major sub- fields available locally. For conven- tional disciplines, then, there is still no successful model for systematic co- operation. In short, it seems to be im- possible to divide academic disciplines in an academic way. Factors Contributing to Successful Cooperation Looking back over decades of coopera- tive effort, we have identified seven major factors that promote successful cooperative collection development. They include propitious circumstances, visionary and committed individuals, supportive organizational structures, appropriate staff participation, biblio- graphic and physical accessibility to col- lections, outside funding, and a history of successful cooperation. First, circumstances have to be con- ducive to cooperation.88 When the economic, social, political, cultural, or academic environment limits an institu- tion's ability to provide resources, a joint effort becomes the best way to meet local needs. The situation at Duke and UNC- CH in the 1930s provided the impetus for cooperative collection development. Given the South's poverty, administra- tors, faculty, and librarians knew they did not have the resources to build major research libraries competitively, so they decided to meet the need for materials cooperatively. Over the decades, each major new cooperative initiative has begun for similar reasons-a need for library resources without adequate funds to acquire them locally. Now, al- 488 College & Research Libraries most sixty years later, the rising costs of library materials, the appearance of new electronic formats, and inadequate fund- ing create new imperatives to expand cooperative collection development. Second, key individuals must share both a vision of what cooperation can accomplish and a commitment to pursue cooperative options. While administra- tors, faculty, and librarians understood the limitations imposed by circum- stances in the 1930s, they also had a vi- sion of what they could accomplish through cooperation-not only for their own institutions but also for the region. Since then, library staff have continued to search for new ways to cooperate, while library administrators have sup- ported them. Their vision and commit- ment have been crucial to success.89 Third, administrators must establish formal organizational structures that en- courage cooperation.90 Library coopera- tion in the Research Triangle began in the context of "cultural relations between the two institutions,"91 and involved uni- versity administrators, faculty, and librarians. University administrators and faculty have continued to partici- pate, but only up to a point. Librarians have never asked faculty or university administrators to ratify specific coopera- tive collection development agreements. Intrainstitutional structures that pro- vided for greater faculty involvement and specific administrative endorse- ment would lend more credibility to cooperative agreements, because all par- ties concerned would have worked to- gether to create them. These groups would therefore have a greater stake in maintaining them. ' Interinstitutional structures are also important, because they foster an en- . vironment in which cooperation can take place. Cooperative collection development among the Research Tri- angle institutions began and has been periodically revitalized and expanded because university or library adminis- trators created new organizations to pro- mote it. The regular meetings of collection development staff, which began under the auspices ofTRLN in the November 1993 1970s, provide opportunities for librari- ans to maintain old cooperative pro- grams and create new ones. These meetings also encourage honest and open communication between librarians from different institutions, help selectors coordinate practices, and thereby social- ize staff for cooperation.92 Fourth, the involvement of staff at the operational level is essential. No matter how much administrators promote cooperation, the key to success lies ulti- mately with individual selectors. They, and not administrators, create and operate the actual cooperative programs'. Selec- tors therefore need to be intimately in- volved in all aspects of the cooperative process for their areas of responsibility. 93 The resulting participatory relationship among selectors ensures they will make realistic commitments and meet their ob- ligations to each other.94 Selectors also need support and time.95 A major reason the area studies pro- grams have been successful is because full-time · bibliographers have overseen their development from the beginning and have devoted considerable intel- ligence, creativity, and energy to main- taining them. Where cooperative programs for Western Europe and the United States exist, it is because full-time collection development officers have taken the initiative and worked with their part-time colleagues to bring such programs into existence. Cooperation in the sciences has not yet emerged. If it does, it will be partly because recently appointed full-time science bibliog- raphers can nurture its development.96 Fifth, the experience of TRLN and other cooperative consortia demon- strates that librarians must provide in- formation about the holdings of cooperating libraries and maximize the availability of their collections.97 Biblio- graphic accessibility, faculty and student access to collections, and special docu- ment delivery have been critical to successful cooperation. In the 1930s, librarians duplicated main entry cards and created a union catalog. Shortly thereafter, they added direct faculty bor- rowing and daily document delivery. During the 1980s, they created a joint online catalog, expanded borrowing for faculty and all students, and improved interlibrary loan (including the faxing of rush requests and free or subsidized photocopies of articles). New technolo- gies offer even greater opportunities to link libraries in cooperative endeavors in the 1990s.98 Indeed, although the prox- imity of the TRLN libraries aided coopera- tion in the past, advances in tele- communications and the appearance of electronic library resources reduce the sig- nificance of distance, both for the Research Triangle university libraries and for other institutions around the country.99 Sixth, librarians need to recognize the importance of outside funding both for initiating new ventures and revitalizing old ones. Although the TRLN coopera- tive programs began during the Depres- sion when the economy could not have been worse, financial need alone did not lead to cooperation. As a matter of fact, a recent survey of cooperative collection development programs among mem- bers of the Association of Research Li- braries found only one other program dating from the 1930s.100 Rather, outside funding was the catalyst that brought cooperation into being and contributed to its success. 101 From their beginning in the 1930s, through the development of cooperative programs for area studies in the 1940s, 1950s, and 1960s, to the revi- talization and expansion of cooperation in the 1970s, 1980s, and 1990s, every major cooperative initiative by TRLN librarians has come about because of the existence of outside funds used either as Cooperative Collection Development 489 seed money or to lock in embryonic cooperative agreements. Finally, a history of successful cooperation encourages its continuance and expansion.102 In the case of TRLN, the cooperative programs are solidly es- tablished, well-known throughout the universities, and widely accepted by ad- ministrators, faculty, and librarians. After half a century members of the con- sortium have built formidable com- plementary collections. Any attempt to abrogate these arrangements would en- tail significant political and economic costs. Therefore, just as historical cir- cumstances provided the impetus that led to cooperation in the 1930s, they are now influential in ensuring its survival. Approximately two-thirds of a cen- tury ago, during the depths of the De- pression, administrators, faculty, and librarians at Duke and UNC-CH real- ized that they would never have enough money to build two separate compre- hensive collections. By working to- gether, however, TRLN librarians have built coordinated, interdependent, and interlocked collections of far greater breadth and depth than they could have achieved alone. Currently librarians across the country face similar problems. They cannot af- ford to acquire all the materials scholars need for research, nor will they be able to document fully contemporary civili- zation. By cooperating, however, librar- ians can build local, regional, and national collections that serve both their institutions and the world. REFERENCES AND NOTES 1. Indeed, Joe Hewitt and JohnS. Shipman found that "the theoretical necessity and potential benefits of cooperative collection development are almost universally recog- nized," and even "libraries reporting no cooperative collection development activity seemed as favorably disposed to the idea of cooperative collection development as those in libraries actively engaged in programs." "Cooperative Collection Develop- ment among Research Libraries in the Age of Networking: Report of a Survey of ARL Libraries," Advances in Library Automation and Networking 1 (1987): 191, 198. 2. Database Overlap Study (Chapel Hill, N.C.: TRLN, Oct. 31, 1988), and TRLN Database Overlap Study (Chapel Hill, N.C.: TRLN, May 1992). 3. Quoted in Minutes of the Department Heads Meeting (Academic Affairs Library, UNC-CH), Oct. 24,1989, 2. 490 College & Research Libraries November 1993 4. U.S. House, Subcommittee on Postsecondary Education, Hearing on the Reauthorization of the Higher Education Act of 1965: Library Programs (Serial No. 102-35) (Washington, D.C.: Govt. Print. Off., 1991): 42. 5. Hewitt and Shipman found only one cooperative collection development program among ARL institutions older than that of the Research Triangle university libraries- and it is confined to the field of religion. 202. 6. Before the creation of the university system in 1963,"UNC" consisted only of the Chapel Hill campus; after that date the formal designation of "Chapel Hill" was added. 7. Robert Estall, Population Change: The American South (London: John Murray, 1989), 5. 8.. Louis R. Wilson, The Library of the First State University: A Review of Its Past and a Look at Its Future (Chapel Hill, N.C.: University of North Carolina Library, 1960), 23, 27. 9. For statistics on collection size, growth, and expenditures during this period, see Robert E. Molyneux, The Gerould Statistics, 1907/08-1961/62 (Washington, D.C.: Association of Research Libraries, 1986) and U.S. Advisory Committee on Education, Library Service, report prepared by Carleton B. Joeckel (Washington, D.C.: Govt. Print. Off., 1938): 91. 10. For concise discussions of regionalism and the New South ideology, see Charles Reagon Wilson and William Ferris, Encyclopedia of Southern Culture (Chapel Hill, N.C.: Univ. of North Carolina Pr., 1989): 1113-15, 1121-22. 11. (Durham-Chapel Hill, N.C., 1935), 7-8. 12. Program of Cooperation, 9. 13. General Education Board, Annual Report 1947/48, 45. See also its Annual Report, 1933/34,23-25, and 1934/35,33. 14. Raymond B. Fosdick, Adventure in Giving: The Story of the General Education Board: A Foundation Established by John D. Rockefeller (New York: Harper, 1962), 269-73. 15. E. Morrell, letter to Frank [Porter] Graham, Apr. 6, 1934, and R. B. House, letter to [Robert Diggs Wimberly! Conner, May 14, 1934, Chancellor's Records, R. B. House Series, Faculty Affairs, Special Committees: Intellectual Cooperation with Duke, 1929- 1934, University Archives, UNC-CH: 1. 16. Proposed Principles of Procedure for Cooperative Buying Program on the Part of the Libraries of Duke University and the University of North Carolina, n.d., Academic Affairs Library, Librarian's Records, Duke University-University of North Carolina Library Cooperation, 1930-1933, UniversitY, Archives, UNC-CH. 17. UNC, Report of the Division of Library and Ltbrary School1935/36, 1. 18. General Education Board, Annual Report 1933/34,33. 19. Memorandum to President Graham on Application to the General Education Board for Funds for Library Research Materials, May 2, 1935, 3, Chancellor's Records, R. B. House Series, Committee on Intellectual Cooperation with Duke University, University Ar- chives, UNC-CH; and Request for Aid in Developing Reference and Research Facilities in the Libraries of the University of North Carolina and Duke University (and its appendices) [May 4, 1935], 15, Libraries Records, Cooperation with UNC Library, AEB 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. Acquisitions, University Archives, Duke University. Robert B. Downs and Harvie Branscomb, "A Venture in University Library Coopera- tion," Library Journal 60 (Nov. 15, 1935): 877-79; and Harvie Branscomb and Robert B. Downs, "A Plan for University Library Cooperation," School and Society 42 (July 13, 1935): 64-66. These principles and the specific guidelines for interpreting them are embodied in the Proposed Principles; see also articles by Downs and Branscomb, cited in n.20. "Desiderata," Library Notes (Duke University), no. 15 (Dec. 1945): 13. Robert B. Downs, "American Library Cooperation in Review," College & Research Libraries 6 (1945): 415. Ibid. Program of Cooperation, 10. Ibid., 10-11. Request for Aid, 12. Ibid. Ibid., 13. Request for Aid appendices. Report to General Education Board on Expenditure of Grant to Duke University and the University of North Carolina Libraries for Research Materials, Nov. 16, 1937, 2, 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. Cooperative Collection Development 491 Libraries Records, Cooperation with UNC Library, AEB Acquisitions, University Ar- chives, Duke University. [Robert Downs],letter to Harvie Branscomb, July 6, 1936~ Duke Documents, 1936-1959, University Archives, Duke University. Cooperation with the University of North Carolina Library in the Collection of Public Documents, Duke Documents, 1936-1959, University Archives, Duke University. See UNC, Report of the Division of Library and Library School1936/37, 10; Memorandum of Conference with Mr. Nuremberger of Duke University Documents Department, Sept. 21, 1939, Duke Documents, 1936-59, University Archives, Duke University; and C. E. Rush, Concise Outline: Agreements and Understandings between the libraries [sic] of Duke University and the University of North Carolina Relative to Library Cooperation [Dec. 1, 1941], 2, Academic Affairs Library, Librarians' Records, Duke- UNC Cooperation: Memoranda, University Archives, UNC-CH. See also G. F. Sher- perd, Jr., Carolina-Duke Cooperative Plan for Acquisitions of Documents, Apr. 10, 1942, Academic Affairs Library, Librarians' Records, Duke University-UNC Library Cooperation: 1942, University Archives, UNC-CH, in which the geographical division for state documents becomes more prominent. UNC, Report of the Division of Library and Library School1936 /37, 2; and Academic Affairs Library, Librarians' Records, Carolina-Duke Cooperation, [1941 ?], Committee on Intel- lectual Cooperation, 1934-52, University Archives, UNC-CH, 1; see also Howard E. Jensen, "The Race Relations Collection of the Duke University Libraries," Library Notes (Duke University), no. 23 (Jan. 1950), 9-10. Request for Aid in the Further Development of Materials for Advanced Study in the Libraries of Duke University and the University of North Carolina, [Jan. 1938], 2-3, Chancellor's Records, R. B. House Series, Committee on Intellectual Cooperation with Duke University, University Archives, UNC-CH. Request for Aid in the Further Development, 5-6. Carolina-Duke Cooperation, [1939?], 1-2, Academic Affairs Library, Librarians' Rec- ords, Duke-UNC Cooperation, University Archives, UNC-CH. Quoted in Friends of the Library (n. d.), 1; reprinted from The Alumni Review (UNC) (July 1940). See Sturgis E. Leavitt, "University Cooperation," Paper delivered at the Bibliographical Conference, Washington, D.C., Feb. 21, 1942, in his Papers Presented at Various Meetings, 1935-1955, North Carolina Collection, UNC-CH. "Cooperation in Advanced Instruction and Research," in Development of Library Re- sources, 42; and Cooperation in Latin American Purchases Duke-UNC-Tulane, [Oct.? 1939] and Cooperation in Library Purchases between Duke University, Tulane Univer- sity, the University of North Carolina: Policy and Procedures [1940?], Academic Affairs Library, Librarians' Records, Latin American Cooperative Program, 1938-41, Univer- sity Archives, UNC-CH. "A Grant for Books on Latin America," Library Notes (Duke University) no.9 (Oct. 1940), 4; and Rush, 5. Cooperation in Library Purchases, 1-2. "Grant," 4. Cooperation in Library Purchases, 1-2. Ibid., 1. UNC, Report of the Division of Library and Library School1939 I 40, 2. University of North Carolina and Duke University Program of Library Cooperation, [May 1, 1942], 2, Libraries: Records of Cooperation with UNC Library, University Archives, Duke University. University of North Carolina and Duke University Program of Library Cooperation, [May 1, 1942], 2-3. See also E. Carl Pratt, "Library Cooperation at Duke and North Carolina Universities," College & Research Libraries 2 (1941): 142-45. "A Grant from the Carnegie Corporation," Library Notes (Duke University), no.10 (May 1941): 5. Memorandum on University of North Carolina and Duke University Program of Library Cooperation, Oct. 8, 1952, 1-2, Chancellors' Records: R. B. House Series, Committee on Intellectual Cooperation with Duke, University Archives, UNC-CH. 492 College & Research Libraries November 1993 52. See B. E. Powell, "Library Cooperation between Duke University and the University of North Carolina," Library Notes (Duke University), no. 31 (Nc;>v. 1955): 13-15. 53. Duke-Carolina Cooperation and Its Extension (DCCE), July 1, 1956, 2, Cooperative 54. Programs Files, Collection Development Department, Academic Affairs Library, UNC- CH [source hereafter cited as CPF]. Cooperative Lending Program of the Libraries of Duke, North Carolina State College, the University of North Carolina and the Woman's College of the University of North Carolina, Aug. 26, 1955, 1, Libraries Records, Cooperation with UNC Library, Univer- sity Archives, Duke University. 55. DCCE, 2-3. 56. Ibid., 3. 57. Ibid., 3 and 5. After Tulane dropped out of the cooperative program for Latin America, Duke assumed responsibility for materials from Central America; UNC, for those from the Caribbean islands. 58. See correspondence between Gertrude Merritt (Duke) and Harry Bergholz (UNC) in the fall of 1959, CPF. 59. Benjamin E. Powell, letter to Jerrold Orne, June 9, 1961, and Orne, letter to Powell, June 22, 1961, CPF. 60. See Hendrik Edelman, "The Death of the Farmington Plan," Library ]ournal98 (Apr. 15, 1973): 1253. 61. Duke University, The University Libraries: Report to the President 1963/64,9. 62. JohnS. Curtiss, letter to Benjamin E. Powell, Apr. 17, 1962; and University of North Carolina-Duke University Cooperative Program in Russian and East European His- tory, n.d.; both from Slavic Bibliographer's Files, Collection Development Department, Academic Affairs Library, UNC-CH. 63. Duke University-University of North Carolina Working Agreement on Africana Li- 64. 65. 66. 67. 68. 69. brary Materials, (June 1, 1965], CPF; see also, Duke University, The University Libraries 1964/65, 5-6. Tentative Selection Guidelines for Asian Materials, Mar. 15, 1972, Duke-UNC Coopera- tion, Administrative Offices, Academic Affairs Library, UNC-CH. These agreements were originally formulated in the fall of 1970. Duke University, The University Libraries 1960/61, 6; and 1961/62, 4; and 1971/72,2-3. Duke-UNC Committee on Cooperation, Report of Meeting, Apr. 30, 1976, 1-2, CPF; see also Joe Hewitt, "Triangle Universities Library Cooperation Committee," in his Selected Unpublished Papers and Presentations 1976-1980 ([Chapel Hill, N.C.], n. d.). On the la~t aspect, see Gary D. Byrd et al., "The Evolution of a Cooperative Online Network," Library ]ournalllO, no.2 (Feb. 1, 1985): 71-77. Minutes of UNC-CH/Duke Meeting, Jan. 27, 1977, 2, CPF. Memorandum to Miss Florence Blakely, from JohnS. Shipman, Re Final Report, Title ll C, Strengthening Research Libraries Resources Program, Oct. 28, 1981, 1, CPF. 70. "Notes from the Hanes Meeting of October 15, 1981," 1, CPF, and John Shipman, "Total Materials Funds Associated with Cooperative Purchases, FY 1979/80-FY 1991/92" (Chapel Hill, N. C.: UNC-CH, Academic Affairs Library, Collection Development Department, Jan. 8, 1992). 71. SeeJohn Rutledge's articles: "Collecting French Regional History Cooperatively," Collection Management 8, no. 2 (Summer 1986): 63-77; and "Collecting Contemporary European Literature for a Research Library," Collection Management 5, no.1/2 (Spring/Summer 1983): 1-13. 72. Program for Cooperative Collection of French Regional Materials, Mar. 2, 1984, 1, West European Bibliographer's Files, Collection Development Department, Academic Af- fairs Library, UNC-CH. 73. James Rolleston, letter to John Rutledge, May 17, 1984; and Proposal for the Coopera- tive Collecting of Contemporary German Belles Lettres between UNC-CH and Duke University, (July 23, 1984], West European Bibliographer's Files, Collection Develop- ment Department, Academic Affairs Library, UNC-CH. 74. Helene Baumann, letter to John Rutledge, Jan. 11, 1989, and Rutledge, letter to Baumann, Jan. 27, 1989, West European Bibliographer's Files, Collection Development Department, Academic Affairs Library, UNC-CH. Cooperative Collection Development 493 75. See the CLR grant proposal: Cooperative Information Resources Development: A Constituency Based Policy Analysis [Chapel Hill, N.C.], Oct. 10, 1990. 76. David Starn agrees, citing the experience of the Research Libraries Group (RLG). "Collaborative Collection Development: Progress, Problems and Potential," IFLA Jour- nal, 12, no. 1 (1986): 18. Richard M. Dougherty also shares this view: "To a financially strapped administrator, cooperation may be seen as a way to generate real dollar savings or to justify future budget reductions. But dollars saved is the wrong measur- ing rod-library cooperation rarely generates identifiable dollar savings. Cooperative programs in resource sharing and/ or shared collection development are better viewed as strategies to enlarge the universe of titles available to library users and/ or to speed up the delivery of documents through interlibrary lending/borrowing systems. These two criteria, availability and delivery, are more appropriate measures of the success of coopera- tive programs." "A Conceptual Framework for Organizing Resource Sharing and Shared Collection Development Programs," Journal of Academic Librarianship, 14, no. 5, (Nov. 1988): 287. See also Sheila T. Dowd, "Library Cooperation: Methods, Models to Aid Information Access," Journal of Library Administration, 12, no. 3 (1990): 66. 77. Hewitt and Shipman found that the expansion of the range of materials available to users, followed by a reduction in duplication, were also major objectives of cooperative programs among research libraries. 207. 78. Paul H. Mosher and Marcia Pankake in particular stress that "programs must be responsive and minimally threatening to local priorities" and that "emphasis should be on nonthreatening models which protect and recognize substantial and long-term institutional program commitments and seek to build on these." "A Guide to Coordi- nated and Cooperative Collection Development," Library Resources & Technical Services 27 (1983): 425. Donald Simpson agrees. "Library Consortia and Access to Information: Costs and Cost Justification," Journal of Library Administration, 12, no. 3 (1990): 96. Librarians in New York successfully established their cooperative collection develop- ment efforts on the assumption that "what libraries were actually doing in collection development in their institutions' self-interest, they would be willing to continue to do in the region's interest .... No monitoring or enforcement had been built into the State's program. Enlightened self-interest was, therefore, both the only motivation for following the regional plan and a very appropriate one in a cooperative system." Joan Neumann, "Impact of New York's Collection Development Funds on Resource Shar- ing," Bookmark 45 (Fall1986): 26-29. Self-interest, coupled with financial incentives, has also been crucial to cooperation even for institutions within a system. See George J. Soete and Karin Witten borg, "Applying a Strategic Planning Process to Resource Sharing: The Changing Face of Collaborative Collection Development among the University of California Libraries," Advances in Library Resource Sharing 2 (1991): 56-57. 79. George Jefferson says that one of the early attempts at library cooperation in the British Isles failed because "allocation of purposely narrow subjects fields to encourage participation was done arbitrarily," and that "large libraries found ... it was difficult to reconcile this obligation with their duties to local readers." A more successful attempt in Wales assigned subject groups "after participating libraries had submitted their choice of subject." Eventually, however, there were so many objections "to the arbitrary allocation of subject fields," that librarians eventually retreated to an ad hoc method of acquisition. Jefferson attributes the success of a later experiment in the Newcastle area to the fact that "cooperative projects [were] founded on the realism of local circumstances and characterized by pragmatism rather than neat theoretical abstractions." Library Co-operation, 2d ed. (London: Andre Deutsch, 1977), 35-36 and 123. Librarians in Australia are also basing their cooperative efforts on collection strengths as related to local university programs, an organic and therefore successful strategy. See Margaret A. Cameron, "Evaluation and Inter-institutional Cooperation in Collection Development," Australian Academic & Research Libraries 20 (Mar. 1989): 23-28. 80. When librarians base cooperative commitments on local needs, they obviate Maidel K. Cason's concerns about accountability in national efforts. "Accountability in Coopera- tive Collection Development: The Elusive Ingredient," in Academic Libraries: Myths and 494 College & Research Libraries November1993 81. 82. 83. 84. 85. 86. 87. 88. 89. 90. Realities: Proceedings of the Third National Conference of the Association of College and Research Libraries, ed. Suzanne C. Dodson and Gary L. Menges (Chicago: Association of College and Research Libraries, 1984): 245-48. David Starn points out that all cooperative efforts by the founding members of RLG "were entirely voluntary, thereby recognizing the continued autonomy of each insti- tution." "Collaborative," 10. This approach has been essential to the success of other cooperative programs. See Karen Krueger, "A System Level Coordinated Cooperative Collection Development Model for Illinois," in Coordinating Cooperative Collection Development: A National Perspective, ed. Wilson Luquire (New York: Haworth Press, 1986), 53-54; Soete and Wittenborg, 56; and Mosher and Pankake,425. Such a user-oriented approach allows cooperative programs to work even: when institutions vary greatly in the breadth and depth of collections, as in the case of ILLINET libraries. See Krueger, 50-51 . In fact, Mosher and Pankake state that cooperative agreements should be reviewed every three to five years and modified to reflect any changes in practice. 429. Martha Smith stresses the same point. "Cooperative Collection Development for Rare Books among Neighboring Academic Libraries," College & Research Libraries 46 (1985): 160-67. Hewitt and Shipman found ad hoc agreements on expensive research materials and the selection and cancellation of serials to be the most common form of cooperation among research libraries, while systematic divisions of responsibility based on subject, language, country of. origin, or format were rare, and when they existed, they were narrow in scope. 191 and also 211-15. . Other cooperative ventures have had similar experiences. See Soete and Wittenborg, 53, 55, and 57, and Elizabeth Roberts, "Cooperation, Collection Management, and Scientific Journals," College & Research Libraries, 48 (1987): 247-51. Elizabeth P. Roberts gives an example of such a program between the libraries of Washington State University and the University of Idaho, where UI has responsibility for journals in forestry and mining; WSU, for veterinary medicine, because the other institution does not have academic programs in those areas. "Cooperation," 247-51, and "Cooperative Collection Development of Science Serials," Serials Librarian 14, no. 1/2 (1988): 19-31. More broadly, David Starn, among others, states that the North American Collections Inventory Project's Conspectus was designed "to present a composite picture of collection strengths and current collection practices in participat- ing libraries." Librarians are using information about these unique collection strengths to build systematic cooperative programs on a national scale. "Collaborative," 11. This was the approach that the Farmington Plan and its national-level successors later adopted. More recently, local and regional cooperative programs, such as the North- west Regional Consortium for Southeast Asian Studies, have assigned specialized collecting responsibilities for specific foreign countries to their member libraries. On the latter, see Marian Ritter, "Four Paradigms for Sharing Library Resources," College & Research Libraries News 52 (1991): 367. David C. Weber noted that timing is often critical to successful cooperation and that economic motives are found in every example he cites. "A Century of Cooperative Programs among Academic Libraries," College & Research Libraries 37 (1976): 215, 219. More specifically, Starn states that "RLG was founded in a time of relative financial austerity in US libraries." "Collaborative," 18. Smith also found that administrative support of cooperation is essential, because it helped to ensure the continuance of the policy. "Cooperative," 160-67. Conversely, one of the major reasons for the failure of cooperation in Louisiana was the lack of support from library administrators and governing bodies. Beverly E. Laughlin, "Barriers to Regional Collection Development," Louisiana Library Association Bulletin 52 (Fall 1989): 45-50. George Jefferson, analyzing the Newcastle approach to cooperative activity, states that "co-operation in the last analysis depends for success upon personalities who induce the wish to co-operate." He also points out "that the authority for projected co-operation should come from a broader base than just the actual libraries concerned and involve the highest executive level of the institutions." Library Co-operation, 122. The administrative link of the libraries of the University of California system via the Office of the President, coupled with an official policy of "One University, One 91. 92. 93. 94. 95. 96. 97. Cooperative Collection Development 495 Library," was the key factor in bringing about cooperation in that state. Soete and Wittenborg, 52. Based on a review of papers presented at a conference on cooperative collection development, Carl W. Deal concluded that "a governing authority should be established to coordinate efforts and respond to and set priorities." "A Model Criterion for Statewide Plan/Process/System," in Coordinating Cooperative Collection Development: A National Perspective, ed. Wilson Luquire (New York: Haworth Press, 1986), 217. Conversely, Hewitt and Shipman found that the lack of an appropriate organizational mechanism was one of the major reasons that research libraries failed to cooperate. 198. Program of Cooperation, 5. On the importance of these processes to successful cooperative programs, see Kurt Pond and Dwight F. Burlingame, "Library Cooperation: A Serials Model Based on Philosophical Principles," College & Research Libraries 45 (1984): 299-301. Conversely, communication breakdowns have contributed to the failure of many cooperative endeavors. Weber, 211. The involvement of those actually selecting materials was also crucial in expanding cooperation within the University of California/Stanford consortium from a one-time, ad hoc Shared Purchase Program to the Shared Collections and Access Program, which also includes on-going and systematic ventures. Soete and Wittenborg, 56-58. Moreover, Mosher has observed that "collaboration is achieved by working ahead, planning, reflecting, and talking with both users and colleagues about the collections, the programs they serve, and about aspirations for the collections of the future. The accomplishment of working collaboration among people doing selection and making collection management decisions is more central to effectiveness than distribution of subject, language, discipline or format." "Collaborative Collection Development in an Era of Financial Limitations," Australian Academic & Research Libraries 20 (Mar. 1989): 12-13. See also his "Cooperative Collection Development Equals Collaborative Inter- dependence," Collection Building 9, no. 3/4 (1988):·29-32. Mosher cites psychological studies showing that "effective cooperation is most readily achieved by forming small working teams" and that "such groups tend to foster cooperation rather than competition, and collaboration has been shown to strengthen such groups and encourage them to complete more challenging tasks." "Cooperative Collection Development Equals Collaborative Independence," in Collection Manage- ment: Current Issues, ed. Sarah Shoemaker (New York: Neal-Schuman Publishers, 1989), 31. See also his "Collaborative Interdependence: The Human Dimensions of the Con- spectus," IFLA Journal 16 (1990): 329. This atmosphere of honesty and trust helps promote accountability on the personal level and thereby addresses Cason's concern about the lack of accountability in cooperative collection development. "Accountabil- ity," 245-48. On the importance of allocating staff and time for cooperative activities an<;! providing means of continuous interaction through formal and informal meetings, see Mosher and Pankake, 425, and Deal, 217. Conversely, one of the major problems of cooperative programs is the lack of communication between partners. Hewitt and Shipman, 221. Administrators at other libraries have recognized the importance of full-time staff to run cooperative programs. In 1986 the Illinois State Library established the position of Coordinator of Cooperative Collection Development to supervise cooperation in Il- linois. "The establishment of this office has directly influenced the course of coopera- tive collection development in Illinois." Terry L. Weech, "Networking and Cooperative Collection Management-The Illinois Experience," Collection Building 10, no. 3/4 (1989): 55. Bibliographic and physical accessibility have been central to the success of every cooperative program. See, for example, Deal, 219-20. According to Hewitt and Ship- man, 95 percent of the ARL institutions provided special physical access or interlibrary loan privileges to users of partner libraries as part of the cooperative collection development agreements. 219-20. In fact, Mosher and Pankake state that cooperation "presumes easy bibliographic access and delivery in a time frame rapid enough not to have detrimental effect on the work of institutional users." 428. 496 College & Research Libraries November 1993 98. Hewitt and Shipman consider advances in national bibliographic networks in the late 1970s to be one of major factors behind the surge in cooperative programs that occurred during that time. 190 and 203. On the other hand, based on visits to nearly four dozen charter members of OCLC, Hewitt concluded that "coordinated collection develop- ment does not arise automatically simply because of the existence of a successful network," but that "strong independent initiatives are necessary." "Impact of Net- works on Collection Development," Library Acquisitions 1 (1977): 213. 99. As an indication of how significant they might be, RLG's Conoco Study revealed that selectors in the humanities were willing to change 40 percent of their selection deci- sions "and rely on collections at other institutions if they could be reasonably sure of both bibliographic access and physical availability of items in those collections (max- imum of seven days for delivery of materials)," while science selectors were willing to change up to 50 percent of their decisions if items could be obtained within three days. Mosher, "Cooperative Collection Development," 31. 100. Hewitt and Shipman, 202. 101. In his survey Kraus considers outside financial assistance to be essential. 179. Deal also emphasizes the importance of seed money in initiating cooperative collection develop- ment, but considers that "long-term maintenance of programs of cooperative collection development depends upon incorporating their support into ongoing budget alloca- tions." 218-19. 102. See, for example, Bernard G. Sloan, "Resource Sharing among Academic Libraries: The LCS Experience," Journal of Academic Librarianship 12 (1986): 28. A BETTER WAY To SEARCH DATABASES W e started in 1985, database searchers committed to better search software design. We became the premier vendor of Medline, then expanded our catalog to other databases. Last year we won Information World Review's PRODUCT OF THE YEAR for faster, easier search software. But a better way means meeting the evolving needs- individual and campus wide- of today's library users. Announcing OVID: a database interface so flexible it molds itself to your search environment. With OVID you're free to move /rom one operating system to another without retraining. OVID's Common User Inter/ace assures identical functionality in DOS, Windows and UNIX. A haven/or beginners, OVID's Easy Mode has on screen prompts. The more experienced can pull-down menus showing an array of search options. Experts will /eel at home using online syntax. Search with natura/language if you like. OVID mapping cuts through the mystery of controlled vocabularies, homing in on precisely matching subject headings. There's an unprecedented array of search tools - indexes, thesaurz; limits and fields - many never be/ore available in an inter/ace. They're all standard OVID features. .- HELP /or every search /unction is context-sensitt've and on screen, never more than a key- stroke or mouse click away. OVID. A better way to search ERIC, Current Contents, PsyciNFO*, Medline, Readers' Guide Abstracts, EMBASE and more. CD PLUS Technologies New York 800-950-2035/212-563-3006 London 44-(0)81-748-3777 Amsterdam 020-672-0242 ANNOUNCING The Human Development Race explains why some developing countries have readily succeeded at improving the health, education, and income of their citizens while others seem mired in failure. After examining economic and social factors in the performance of ninety countries, Lindenberg focuses on the experiences of six Central American countries, furnishing wide-ranging, and sometimes surprising, conclusions. Marc M. Lindenberg is currently senior vice pre_sident of programs for CARE, on leave of absence from his position as a public policy lecturer at Harvard's John F. Kennedy School of Government. 1993, 233 pages, Cloth ISBN 1-55815-277-6, $29.95 Paper ISBN 155815-278-4, $14.95 PUBLISHED BY II:§ PRESS 'The Human Development Race combines human resource issues and quality of life issues in a creative fusion that allows a richer analysis than the usual economic approach. Lindenberg successfully tackles the thorny problems of measuring improvements in a practical and theoretically compelling way." - William Ascher Professor, Sanford Institute of Public Policy Duke University