This survey updates a 1985 study of automation in the Association of Research Libraries (ARL) in New England. These nine libraries (Brown University, Boston University, University of Connecticut, University of Massachusetts, Massachusetts Institute of Technology, Dartmouth College, Boston Public Library, Harvard University, and Yale University) are the largest in New England. The locations of these nine libraries are shown in Figure 1. Their experiences foreshadow those of other libraries and indicate trends in library automation.

Since 1985, due to vendor instability, some of these libraries have had to switch to com-pletely new systems or add modules from other vendors. While placing an emphasis on standard hardware (i.e., IBM or DEC), the libraries have expanded their systems personnel and developed closer relationships with campus computer service departments. Networking -- both on campus and external -- has become extremely important to library automation and will continue to play a major role in the future. The ability to mount external databases onto the local system appears to be a trend of increasing importance. Concomitant with this is the loading of externally cataloged data to the local systems. In this environment of mixed data and external networking, the user interface becomes extremely significant. Some libraries are placing an emphasis on this development. Finally, all the libraries are becoming aware that library automation is a continuous process. While loading one system, they already are planning for the next generation system.
 
 

2. RESEARCH METHODOLOGY

This research continues the case study method developed in the 1985 study of automation in New England ARL libraries (Tebbetts, 1985). A telephone interview of the Director and/or the Systems Librarian of each library was conducted. The open-ended interview technique was used which permits the interviewees to expand on their answers and provides more individualized information.

The major issues probed were:

• Whether the Library has the same system as in 1985?

• What changes have been made and why?

• Whether networking is available and what its impact is on the library?

• What internal organizational changes have occurred?

• What factors are involved in the decision-making process?

• What trends are discernible in library automation?

A review of the library literature revealed a telephone survey of twelve other ARL libraries which resulted in SPEC Kit 126 (Association of Research Libraries, 1986). There are some similarities in the results which will be discussed later in the paper. However, this study which updates earlier research revealed some changes in direction and emphasis.
 
 

3. STUDY RESULTS

From the telephone interviews, it is obvious that several changes have occurred in the last five years. In 1985, five of the nine libraries had begun implementation of automated integrated systems, one was on the verge of signing a contract for an integrated system, two had inhouse systems, and one had not made any plans.

In 1990, four of the five with integrated systems have had to either change systems or to add modules from other systems. This has resulted from vendor instability. The library on the verge of signing a contract did so but has since had to add modules from another system. Again, vendor instability is the reason for the switch to another company. Of the two libraries with inhouse sys-tems, one opted for a commercial system and one has continued development of its own system. The library that had not made any plans has since that time begun implementation of an integrated system.

Therefore, in five years, six of the nine libraries have had to make changes in their original plans, one has begun implementation of a system, and two are continuing development of their original systems.

In this volatile situation, certain factors and trends in library automation are discernible.
 
 

4. FACTORS AND TRENDS IN LIBRARY AUTOMATION

Why did so many of these libraries have to change systems or add modules from other sys-tems? When they changed systems what factors influenced their decision-making? What do these libraries view as important trends in library automation now and in the future. Results derived from the telephone interviews revealed the following factors as important to the process of library automation.

4.1. Vendor Instability

A major factor in the changes witnessed in the New England ARL libraries is vendor insta-bility. Four of the five vendors engaged by the libraries in 1985 have since either gone out of business or are in shaky circumstances. This has necessitated changes for the libraries involved. One library is in the process of getting a new system; another is continuing with the old system but cannot add any new modules. It also is looking at a new system. Two libraries have added sub-systems from another vendor, and one library with an incomplete system is looking at a distributed interfaced system with CD-ROM based networked catalogs.

Due to vendor instability, these libraries have not been able to fully implement the integrated systems as originally planned. They have been forced to go to new systems or to add subsystems from other vendors. The libraries with added modules have not interfaced yet these disparate sys-tems. Therefore, those libraries with online catalogs from one system have not linked acquisitions and serials control from other systems. Three do not have circulation systems linked to their online catalogs.

Only the library with the stable vendor has been able to fully integrate all the major library functions and proceed to further implementation and expansion of the system. One of the libraries with the inhouse system has continued development of the system by linking and interfacing modules from other vendors with the inhouse system.

It is obvious that vendor instability has been a major factor in delaying the development of library automation in several of these libraries. Of necessity, they have had to switch to new systems or mixed systems. The libraries are just in the process of switching so it is too soon to determine the problems encountered in the transition. However, it is safe to say that most of these libraries are finding it difficult to interface modules from disparate systems. Only the library with the inhouse system has done so successfully.

4.2. Hardware

Hardware has become a key factor in determining which system is utilized. Most libraries want standard hardware (i.e., IBM or DEC). This is very likely a result of the instability of vendors in the library automation field. There is a definite turning away from proprietary hardware and turnkey systems as shown in Figure 2.

Five of the nine libraries are on or are looking at systems run on IBM equipment and two are using DEC equipment. One library is looking at versatile software that will run anything that is UNIX-based, and one has DG hardware. In 1985, three had systems running on proprietary hardware. In a time of vendor instability, proprietary hardware becomes especially unattractive. These libraries have all had to look at other systems or add modules from other vendors. Upgrad-ing of hardware and software in this situation becomes virtually impossible. Therefore, standard computer hardware becomes much more attractive in this volatile situation.

4.3. Systems Staff

The most dramatic changes in the internal organization of the libraries have occurred in the systems departments. All of the libraries now have systems offices in the library. All of the aca-demic libraries have seen increases in staff with the addition of programmers, operators, and training/documentation specialists. Only the public library reports no increase in systems staff, but this is due to financial constraints rather than need. Some of this increase is attributable to the vola-tile situation in the vendor arena. With libraries unable to rely on the vendors for programming and maintenance, they have had to expand their inhouse capabilities. The proliferation of mixed sys-tems has necessitated an increase in systems staff to handle the disparate systems.

Reorganization in the systems office is evident in two of the libraries. With added staff, service and first line support have been emphasized in these two institutions. This is also true in the libraries as well. Inhouse programming is also becoming more common. Directors of the systems offices are becoming more responsible for overall automation planning while other sys-tems staff handle operations and training. Systems offices tend to have an average staff of four persons with two libraries having much larger staffing and one having a smaller staff. If this survey is indicative of the future, other libraries can expect to increase systems staff.

4.4. Campus Computer Services

Along with the increase in library systems offices have much closer relations with the campus computer services. The libraries have come to rely on campus computer services for additional expertise, programming, and operations support. With vendors becoming less stable, the libraries have turned to campus resources to develop their systems.

Three libraries are using mainframes located in and maintained by campus computer services, two have future plans which involve close cooperation with campus computer services, three are in consortium arrangements, and one has a huge library systems office which handles the computer functions. It is evident that the trend is for the libraries to rely on local expertise to develop and implement their systems rather than turning to vendors and their specialists.

The importance of networking is another factor which is making it essential for the libraries and the campus computer services to work together. Connecting the libraries' systems to the campus networks and to external networks requires expertise and close cooperation.

Therefore, it is becoming essential for the libraries and the campus computer services to co-operate to provide resources to their faculties and students.

4.5. Networks

Networking is becoming extremely important on all of the campuses in this study. Three of the campuses are fully networked with the online catalogs available on the network, five campuses are partly networked with three libraries on the network and two not connected. The public library is serving as a regional center but it is not networked yet.

It is becoming obvious that the libraries' databases are high demand items on the campus networks. Faculty and students want access to these databases from their offices and dormitory rooms.

In addition to campus resources, faculty and students want access to other materials available on the nationwide networks, such as Internet. Libraries are making their online catalogs available on these networks. The ability to interface these systems is becoming a crucial factor on campus requiring the combined expertise of libraries and computer specialists.

Libraries have the databases while networks provide the mode of access. Each needs the other to complete the system and provide the best service to their clientele.

4.6. Externally-developed Databases

As library systems mature and resources permit, these libraries are adding externally deve-loped databases to their computer systems. Although all of these libraries have CD-ROM based indexes, most consider CD's an interim technology with locally mounted databases the preferable alternative. Two libraries have mounted externally-developed databases ranging from Medline to Shakespeare's plays. One has loaded additional tapes and is looking at other databases while three are attempting to link CD networks to the campus networks. The other libraries are studying the possibilities. The ARL Spec Kit 126 also indicated that the mounting of externally-developed data-bases was a trend for the future.

Copyright is still a problem in all of these considerations, and libraries are finding publishers reluctant to sell their raw data (see Klemperer, 1989). However, trends clearly indicate that locally mounted externally-developed databases, both bibliographic and full-text, are the next phase once the online catalog is fully developed.

4.7. Decentralized Cataloging

Once the library's catalog is completely online and all the records are in machine-readable form, the next step is to enrich the online catalog with records derived from sources other than catalogers. The two most fully developed libraries are proceeding in this direction. One systems librarian estimates that only half of the records have been input by catalogers. Some have been added by other library staff, such as acquisitions and circulation, while others are from external sources, such as CRL (Center for Research Libraries) tapes and major microforms data.

Therefore, data for the library's database is coming from a variety of sources with traditional catalogers being only one avenue.

It appears that the combination of externally-developed databases and decentralized cataloging will continue to grow as libraries expand and enrich their databases.

4.8. Integrated Versus Interfaced Systems

The fully integrated system with one database serving as the basis for all library functions, such as acquisitions, circulation, serials, online catalog, and cataloging, still appears to be the ideal system most libraries want. However, vendor instability has made this extremely difficult to achieve. Figure 3 shows the changes which have occurred during the five-year period. Five of the six libraries that purchased what were designed to be integrated systems were unable to fully implement them, because the vendors did not develop all the modules. Only one library has a fully integrated system. The other libraries have had either to switch systems or to add modules from other companies. Those that have added other modules have found them ex-tremely difficult to interface. Two libraries that had no system in 1985 have acquired integrated systems. One library began with an interfaced system and has been able to link successfully acquisitions, serials, and circulation with its online catalog.

The ARL Spec Kit 126 reveals the same debate over integrated versus interfaced systems with integrated systems being favored for the most part.

Of the two most fully automated libraries in this survey, one has an integrated system and one has an interfaced system. However, at this time, it would seem that it is extremely difficult to interface disparate systems. Likewise, it is extremely difficult to find a fully developed integrated system. Both of these options require skilled systems personnel and close coordination with campus computer services.

4.9. User Interface

As library systems become more complicated with added databases and access to external networks, the need to provide a user interface becomes more essential. Users need to be assisted in their access to and search for information on the library's databases. They cannot be expected to remember specific commands and varying protocols. It is important for the computer system to aid the user by providing the interface from a set of common commands visible to the user into the unique commands required by each database. This user interface becomes the link between the patron and the computer.

In this survey, three of the libraries are proceeding on workstation-based systems with the user interface located on the workstation. One of these libraries is engaged with a major computer company to develop this approach. Two other libraries envision the user interface as a shell script resident on the mainframe computer. Whichever approach becomes most widely accepted, it is clear that the user interface is becoming an essential feature in the complex world of library auto-mation.

From this survey, it appears that the user interface will be one of the next major areas for development in library automation, and the user interface and workstation design also appeared as a consideration for future development in the ARL Spec Kit results.

4.10. Continuous Automation Process

From the results of this study and the 1985 study, it is becoming increasingly clear that auto-mation is a continuous process not a one-time effort. Just as owners of personal computers have to upgrade software and hardware to enhance and expand their capabilities so do libraries. Old software often is very difficult to connect to networks. New enhanced software capable of net-working very often require upgraded hardware. All of this requires programming and develop-ment.

In this survey, one systems librarian said that while she was purchasing one system she was already planning for the next system. Another librarian stated that she knew she had purchased a package that was obsolete in many ways, but that it was the best available at the time and that auto-mation is a continuous process.

Although many of these libraries have had to make changes in systems and others will have to do so in the future, they never lose what they have accomplished. All have machine-readable databases that can be transferred to other systems, and all have provided their users with improved access to information. Future developments will most likely require new hardware and software, but just as libraries are dynamic, changing institutions so their automation is a continuous process.
 
 

5. CONCLUSION

In conclusion, this study reveals some differences in the factors involved in decision-making from those in the earlier survey (Tebbetts and Chen, 1987). Although some of the factors appear in both surveys, the emphasis differs. In the earlier study, vendor instability in the library auto-mation field was beginning to appear, but this has intensified and become a crucial factor affecting how libraries develop their automation programs. In the earlier survey, some libraries still thought they could rely on the vendors for computer expertise and support while implementing and deve-loping their systems. With vendor instability increasing, the libraries have developed their own systems offices and increased their ties with campus computer services. They have turned away from turnkey systems and proprietary hardware to standard hardware, because of vendor insta-bility and the need for campus support.

Networking was a factor in the earlier survey, but it has become increasingly important in the last five years, especially with the development of nationwide networks. The addition of exter-nally-developed databases and decentralized cataloging are new factors to be considered, along with the increasing importance of the user interface.

The debate over integrated versus interfaced systems continues. Most libraries still seem to want integrated systems but have been unable to achieve them, while interfaced systems continue to be difficult to link.

Finally, it is evident that libraries are beginning to realize that automation is a continuous pro-cess, and that we are only in the beginning stages of development.
 
 

REFERENCES

ARL, Office of Management Studies. SPEC Kit 126: Automated Library Systems ARL Libraries. Washington, DC: Association of Research Libraries, 1986.

Klemperer, Katharina, "New dimensions for the online catalog: The Dartmouth College Library experience," Information Technology and Libraries 8 (2), 138-45 (1988).

Tebbetts, Diane R. Integrating Microcomputers in ARL Libraries: Management Considerations. Field Research Project Report for the Doctor of Arts Degree. Boston, MA: Graduate School of Library & Information Science, Simmons College, 1985.

Tebbetts, Diane R. and Chen, Ching-chih, "Administrative considerations in selecting automated library systems," Proceedings of the First Pacific Conference New Information Technology, June 16-18, 1987, Bangkok. West Newton, MA: MicroUse Information, 1987. pp. 299-306.