LIBRARY COOPERATION IN GUATEMALA:

LIBRARY COOPERATION IN GUATEMALA:
THE CCEBU AUTOMATION PROJECT

COOPERACION INTERBIBLIOTECARIA EN
GUATEMALA: el Proyecto de Automatización CCEBU

Sheila A. Milam

University Libraries
Arizona State University
Tempe, AZ 85287, USA

Automation of library operations depends on the economic and technical parameters of the participating institutions, as well as on the level of expertise found in the personnel. The manifes-tation of these three factors is seen in the automation project carried out by the universities in Guatemala in agreement with the United States Agency for International Development and facili-tated by the Committee for Cooperation among University Libraries (Comité Cooperativo entre Bibliotecas Universitarias, CCEBU). Examination of the CCEBU project illustrates how unique each project is by looking at its goals, participants, funding, components, technical specifications, and future.

The funded project took place from September 1987 to August 1988 and was considered the first phase of a potentially two-phase project. The result is a microcomputer-based cataloging system in each participating university library which creates a database and generates catalog cards and other print products. The new system improves the access to and exchange of bibliographic information among the libraries.

The purpose of the CCEBU project was to improve library service to students, faculty, and other library patrons at the participating libraries by making bibliographic information more availa-ble to them. Not only was service at the individual libraries improved, but their ability to carry out cooperative efforts grew also. The primary objectives of the project were: access to holdings information about reference and theses collections at the five libraries, generation of new acquisi-tions lists, efficient production of catalog cards, and creation of bibliographic databases.

In 1987, the five universities in Guatemala were serving more than 70,000 students and held more than 300,000 monographic titles, theses, and serial titles. All of the institutions are located in Guatemala City, a fact which made meeting and communicating easier.

The Comité Cooperativo entre Bibliotecas Universitarias was the essential organization which provided the impetus, means of communication, and coordination to the automation project. Lines of communication were already established among the participating institutions and working rela-tionships were functioning. The committee was formed in 1983 and began implementing projects. The record of agreement and coordination among the institutions increased the probability of suc-cess for the automation project.

The committee has two representatives from each of the universities, which are Universidad de San Carlos (USAC) (Central Library only), Universidad Francisco Marroquín (UFM), Univer-sidad Mariano Gálvez (UMG), and the Universidad Rafael Landívar (URL), and the Universidad del Valle de Guatemala (UVG). Representatives to CCEBU at the time of the project were Dr. Jeannette de Criado and Silvia de Díaz Durán of UFM, Edna Cifuentes de León and Dr. Samuel Berberián of UMG, Lourdes Bendfeldt of URL, Victor Castillo of USAC, and Dr. Jorge Arias de Blois and María Emilia López of UVG.

In 1987, the committee saw the automation of cataloging operations as the means to improve access to bibliographic information in the university libraries. Fortunately for CCEBU, the assis-tant director of USAID at that time was interested in higher education and university linkages. He had heard of this committee linking the university libraries and was interested in its work. His support of CCEBU's endeavors was essential to the funding of the project.

Following discussions between university representatives and USAID, CCEBU presented a preliminary proposal to USAID in July 1987. It proposed the goals be accomplished by automa-ting technical services, creating an information network among the five university libraries, and accessing bibliographic databases to aid in cataloging and online reference searching. Proposed activities were: A study tour to observe automation systems, the purchase of appropriate hardware and software, technical assistance, and linkage with a university in the United States.

In August 1987, the formal funding proposal was drafted by a USAID- provided project development specialist, Peter Boynton, and a librarian from Arizona State University, Sheila Milam. The university rectors and USAID officials signed the agreement and the work began.

The proposal set out two phases of the project. The second phase would depend on the success of the first. The first phase would include automation of cataloging functions, provision of related technical assistance, and training. Phase two could grow out of the experience of the first phase and might include further development of the catalog, establishment of a communication network among the participating universities, and access to bibliographic databases in the U.S.

Components of the project were the evaluation and recommendation of software and hard-ware, a study tour to observe automation in university libraries, training, and installation and implementation. The proposal's implementation schedule with target dates and actual dates is shown in Table 1.

In USAID budgetary terms, this was a very small project. It yielded benefits out of propor-tion to the financial investment because of human resource investments of time, energy, and com-mitment. The anticipated total cost was $95,000, with $70,000 solicited from USAID for the hardware and software,the study tour,the onsite observation visit, and contingencies. The univer-sities were responsible for the personnel, space, maintenance, and supplies at an estimated cost of $25,000. In some cases, the estimates were rough due to the lack of exact information, yet the year-long project stayed within budget.

Arizona State University provided technical assistance. The dynamic growth of this universty over the past decade is matched by the rising position of ASU Libraries in the Association of Re-search Libraries ranking. ASU Libraries is one of the most automated academic libraries in the United States. Its greatest innovation is the online system using CARL (Colorado Alliance of Re-search Libraries) which includes databases and indexes developed both locally and commercially.

ASU's major responsibilities to the CCEBU project were the evaluation of and recommen-dation on software, the planning and coordination of a study tour in the United States, and the project evaluation.

The Head of the Library Technology and Systems Department, George Machovec, and the

Latin American Catalog Librarian, Sheila Milam, assisted the project. Their networks of collea-gues in the areas of library automation and Latin American studies librarianship were invaluable in finding potential software. For the study tour, colleagues from SALALM (Seminar on the Acqui-sition of Latin American Library Materials) provided information and hosted portions of the visits.

The two-week study tour was an orientation in automated systems of academic libraries. Representatives of the five Guatemalan university libraries, the directors in most cases, traveled to the U.S. in November 1987 and visited four cities: Phoenix/Tempe, Arizona; Flagstaff, Arizona; Austin, Texas, and Washington, D.C.

The first two days were spent at Arizona State University observing cataloging, acquisitions, circulation, online searching, and interlibrary loan. Meetings with the heads of collection develop-ment, library instruction, and librarians council provided a broader view of operations in a large academic library in the U.S. At Northern Arizona University, a university library more similar to theirs in size (400,000 monographic titles and 1 million volumes), equipment that accesses the catalog from classes held miles away on the Indian reservations was demonstrated. The Benson Collection at the University of Texas in Austin, one of the best Latin American library collections in the U.S. was developing a local online catalog and discussed the process. In Washington, D.C., the Library of Congress led a tour and presented demonstrations of special uses of techno-logy, such as the optical disk pilot project of movie stills, graphics, and print. The United States Information Agency (USIA) library and the Columbus Memorial Library of the Organization of American States (OAS) were visited also.

Libraries in the project needed a software that would be compatible with Anglo-American Cataloging Rules, second edition, included topical subject headings from a controlled vocabulary suitable for university libraries, and accommodated a Dewey Decimal or the Library of Congress call number. Technical and financial capabilities dictated a microcomputer-based technology. Telecommunications costs prohibited connection to an existing bibliographic database. The fund-ing source required visible results within one year. This would be the first automation project for the libraries and library directors involved. All factors suggested a microcomputer-based system would be the best choice.

LogiCat software stood out as the package to supply what the project needed. Three repre-sentatives from CCEBU and the two advisers from ASU went to Mexico City to observe the use of LogiCat. They visited Universidad Iberoamericana, Instituto Francés sobre América Latina, Banco de México, Centro Universitario de Investigaciones Bibliotecológicas (UNAM), and Biblioteca Benjamin Franklin. The evaluation team recommended the purchase of LogiCat. They saw many advantages to selecting LogiCat. It would function in Guatemala's university library environment. Record creation and card production would require less time per item and increase the number of records and cards produced. The memory was sufficient for projects CCEBU planned. The for-mat had the desired balance between being a structured format and a flexible format. The program was in Spanish. It ran on standard equipment. It had been used long enough and by enough ins-titutions to have proven its worth. Sistemas Lógicos updated the software according to advances in the technology and to the feedback from its clients. Finally, the format is based on the inter-national standard, the MARC format.

LogiCat (version 3.0) was selected. It is produced by Sistemas Lógicos in Mexico City and, at that time, was installed in more than forty libraries in Mexico and other Latin American coun-tries. It was designed and produced by people who had had experience with university libraries and online catalogs. The company offered training and other services to clients. LogiCat creates a database of bibliographic records and produces catalog cards, labels, and other print products. The records are in a simplified MARC format that includes the most common fields in books format and uses mnemonic tags, instead of numbered tags. The information could be converted to mag-netic tape and loaded on a mainframe with some manipulation. It was programmed using dBaseIII and operates with the MS-DOS operating system on any IBM/PC, XT, AT, or most compatibles. LogiCat supports up to 100,000 records per database and a virtually unlimited number of databases can be supported, limited only by the hard disk storage. The system offers multiple databases, keyword data retrieval, full Boolean operators, a wide variety of printed reports, statistics, back-up procedures, and routines to merge or segment databases by almost any criteria identified in the records. An important additional feature is that all or parts of a database can be written to disk in machine- readable-form for exporting to a word processor, sharing data with another LogiCat system, or exchanging data with other online cataloging systems, text retrieval packages or data-base managers.

One of the most attractive features of LogiCat was its adaptability for further automation, a major consideration. LogiCat could be operated in a multiuse environment via a local area network or with other microcomputer-based file servers. Since the machine readable records from LogiCat could be exchanged via floppy disks, removable hard disks or by telecommunications (via mo-dem), it would be possible for these Guatemalan libraries to share their valuable databases with other libraries worldwide. Sistemas Lógicos also produces several other products compatible with LogiCat: LogiPres (circulation control), Periódicas (serials control), and LogiCom (acquisitions control). These would offer possibilities for future automation programs. The hardware required by LogiCat is the industry standard and would be easy to integrate into a larger program with no difficulties.

A centralized file of the bibliographic information in all five libraries could be accomplished in one of several ways.

A large multi-user PC with between 70-150 megabytes of storage on a Winchester hard disk could be obtained to merge all of the LogiCat files into one large database. The one multi-user PC would have several terminals connected to the one workstation. In this environment, the existing microcomputers could be used for data entry. On a periodic basis, each participating institution would hand over floppy disk copies of their new records to the institution operating the larger PC which could then merge the information into a central database. This could provide the basis for a centralized union catalog for the holdings in the five Guatemalan university libraries.

Several smaller microcomputers could be put in a local area network at one of the institutions with a central file server of 70-150 megabytes. The difference with this option as compared with that described above, is that one file server would support several microcomputers rather than dumb terminals. Each microcomputer could access the central file server and a master database could be built from all institutions as described in option 1.

Eventually, if large amounts of funding were available, it would be possible to export (trans-fer) the LogiCat databases into a large minicomputer which could provide a fully integrated auto-mated library system. Several packages in the United States now offer Spanish language inter-faces. A fully integrated system could support circulation, bibliographic maintenance, and public access catalogs.

In addition, the software used with a word processing program can successfully provide access to other parts of the collections by creating indexes and bibliographies.

The CCEBU database could be mounted on the CARL online system at Arizona State University as a separate database. It would be available to scholars and students at ASU, and elsewhere through dial-in access.

Versions 3.1 and 3.2 appeared after the funding year ended. Version 3.2 was available in November 1989 and brought increased speed to LogiCat functions. For example, printing five sets of cards took 119 seconds in 3.1 in comparison to 94 seconds in version 3.2. A "special" search by title took 75 seconds in 3.1, and only 29 seconds in 3.2. The organization of the title index was greatly speeded up: from 178 seconds in version 3.1 to only 47 seconds in version 3.2. Besides the advantage of faster operations, version 3.2 also presented a local area network capa-bility that allowed use of the database from various workstations at the same time. For example, at one PC, a librarian might be answering a reference question, while at another PC, a clerk could be inputting new cataloging records. In addition, the circulation software, LogiPres, could operate without the duplication of the bibliographic database.

LogiCat version 4.0 was available in January 1990 and presented users with improved searching. Searching all fields of the bibliographic record was possible, including all subjects, added entries, series, notes, and all words in the author, title, and publishing fields. Response time for these searches became immediate. This version would enhance the databases at CCEBU libraries.

Since LogiCat was developed in an MS-DOS environment using dBaseIII it will operate on any IBM microcomputer or most compatibles. It requires a minimum of 512K RAM although 640K is recommended for improved performance and efficiency. An average MARC record on LogiCat is 1,000 characters. Five megabytes is required on the hard disk as a work area for sorting and temporary operations. A 30 megabyte hard disk will therefore support 25,000 MARC records at each Guatemalan library. Considering that the five libraries would like to share their databases with each other (initially reference works and theses) in machine readable form, 30MB was the minimum recommendation for hard disk storage at each library. The IBM/PC clones procured have a hard disk with two sections. One has 30MB and the second has 10MB for a total of 40 megabytes.

Each microcomputer workstation has two floppy disk drive units. The 1.2 megabyte floppy disk drive will support database back-up. The 640K drive is more than the minimum required for the intentions of this project.

The printers need to feed from the bottom so the the heavy card stock used for catalog cards does not have to feed around the platen as in back feed printers.

The hardware vendor is located in Guatemala City and therefore can service the equipment in case of problems.

In July 1988, the system was installed in the five university libraries. As planned by CCEBU, the catalog records of reference items were input into LogiCat (retrospective conversion). Then the libraries began converting local theses. In September 1990, the reference and thesis collections were converted in four libraries and current receipts were cataloged upon arrival. USAC, the largest of all, has converted the reference collection to machine readable format, and continues to work on the theses. Retrospective conversion of the general collection is almost complete at UFM, while the other universities will take longer, based on size of collection and personnel support.

Librarians participating in this project agree unanimously that LogiCat is very beneficial to their library users. Cataloging is faster and demands less personnel. The quality of catalog cards is better. Before acquiring LogiCat, the libraries used mimeograph machines to produce catalog cards, which was messy and labor-intensive. In USAC, for example, less than 10,000 catalog cards were produced in 1986. In 1987, with less than six months use of LogiCat, 33,704 cards were produced. In 1987, 3946 titles were cataloged compared to 5340 in 1988, the first year when LogiCat was used the full year. These are significant increases, especially in libraries with limited staffing.

The purpose of this cooperative automation project was the better exchange of bibliographic information. The LogiCat databases created by each university library of certain collections can now be shared with other CCEBU libraries by exchanging diskettes. A librarian at any of the universities can inform a user whether another library holds a book he or she seeks. Librarians can also search other libraries' databases before purchasing expensive or seldom used items. The generation of "new acquisitions" lists is easier and less time-consuming.

At the end of the year project, recommendations were made to the project participants. Stan-dardization is essential in cooperative projects. Bibliographic information must be recorded uni-formly in order to be interpreted by users in the participating institutions. Adherence to AACR2 and establishment of bibliographic input standards that list criteria for creating a new record and obligatory fields are very important. Quality control ensures that the information contained in the databases meets the agreed standards.

The alliance between university administrators and library directors deserves to be nourished. These two halves need each other in order to provide the best library service to users. Library directors and computing center engineers need to share their respective areas of knowledge in order to build the best system.

This project is an interesting combination of factors that led to its occurrence. It was a truly collaborative effort that consists of a small bit of funding and large amounts of institutional and individual commitment and cooperation that is yielding very significant results, with the possibility of yielding even greater results.

It was truly a cooperative project. The goal was the improvement in library service in each of the five universities through greater knowledge of the other libraries holdings. All participating libraries agreed to carry out certain activities so that this collective body of information could be built.

It created a micro-computer-based system that fit into the financial and technical parameters dictated, yet it is sophisticated enough for academic libraries.

The project takes place in Latin America and the chosen software is produced in Latin Ameri-ca and in the Spanish language. The founders of the company worked in academic libraries with mainframes prior to designing software for microcomputers and understood the needs of academic libraries.

It is rare for USAID to fund such a small project. It is also rare for USAID to fund a project in higher education. CCEBU's track record, good contacts (relationship of mutual respect) between a member of CCEBU and a USAID officer, and that individual officer's interest in the project helped the project receive funding. This shows that, sometimes, funding comes from unlikely sources.

The existence of CCEBU, the Committee for Cooperation among University Libraries was another key factor to the success of this project. It had worked together for several years in planning and implementing projects that required institutional and individual cooperation. Trust and professional relationships were firmly established before this project began.

All participating libraries were in the same city, which facilitated planning meetings, all communication, and the exchange of bibliographic information.

This project has a national impact on higher education and access to information because every one of the universities in the nation participated.

The CCEBU cooperative automation project is a success. Goals were met, and met within the guidelines set out by the funding agency. Guatemala's university libraries have an automated cataloging system which creates a database and generates catalog cards and other print products. The new system improves the exchange of bibliographic information among the libraries and facilitates resource sharing.

This project was originally envisioned as phase one of a two-phase project with the second phase contingent on the success of the first phase, and growing out of its accomplishments. Not only were the goals of the first phase accomplished, but they were accomplished within the bud-getary and time requirements of the agreement. The project is seen as a success by all participants and can securely serve as the basis for phase two. The Committee for Cooperation among University Libraries are formulating the goals and objectives of further cooperative automation projects.

An extension of these accomplishments could include: upgrading the LogiCat software to

its newest version; a local area network or other microcomputer-based file server would allow a multiuse environment; implementing the circulation, serials, and acquisitions control packages produced by Sistemas Lógicos; increasing and facilitating access to the bibliographic databases among the five libraries; and, sharing machine readable records via removable hard disks or telecommunications (via modem) with other libraries worldwide.

Library automation projects involving bibliographic information come in all shapes and sizes. They do not have to be large or expensive, or even to use the most avant garde technology to bring innovation to library. The success of a new technology in a library is measured by the quality of service it offers information users and their improved access to information.