College and Research Libraries FREDERICK G. KILGOUR Computer-Based Systems, A New Dimension to Library Cooperation* The computer has added a new dimension to library cooperations. The recent exper-iences of the Ohio College Library Center~s computer- based cataloging system illustrate how the computer can facilitate in- terlibrary cooperation. The paper also presents some of the difficult organizational problems which must be coped with in developing a computer-based cooperative system. THIS PAPER PRESENTS the concept that library cooperation through computer- based systems enables librarianship toes- tablish new substantive and economic goals. Computerized cooperation also opens up untrodden avenues of research and developn1ent, and by making unnec- essary the imposition of uniformity on library processes, the new cooperation creates hitherto upexplored opportuni- ties for intellectual development in the profession. However, computerized co- operation is, at the present time, in a highly dynamic and incomplete state of development and operation. Such activ- ities at the Ohio College Library Center have turned a major corner -into a new period of library evolution but even the OCLC has taken only a few steps into a vast and unexplored area. As is well known, it was the lure of the benefits of cooperation more than any other one factor that stimulated li- brarians in 1876 to establish the Amer- ican Library Association. One of the 0 Presented at the Annual Meeting of the As- sociation of Research Libraries, May 13, 1972. Frederick G. Kilgour is director of the Ohio College Library Center in Columbus, Ohio. first important committees of the Asso- ciation was the Cooperation Committee that instituted a variety of standards that enabled libraries to cooperate and to reduce costs if not expenditures. One of the early accomplishments of the Co- operation Committee was initiation of the establishment of a standard size for catalog cards. This accomplishment en- abled libraries to purchase cards much more cheaply because they were mass produced, and similarly to obtain inex- pensive cases in which to house the cards. Prior to the acceptance of a stan- dard size card, each library obtained a special size card from a paper manufac- turer and had a cabinet maker build special cabinets in which to house the cards. Moreover, the employment of a card of a standard size meant that it was possible to interchange cataloging infor- mation among libraries, a process that made possible the dissemination of catalog cards among libraries.1 The standardization that made pos- sible the interchange of cataloging in- formation, and thereby a reduction in cataloging effort, was a simple technolog- ical standardization. Today, another technological event, the advent of the digital computer, enables a new type . of cooperation-a cooperation which . for the first time will enable the profession I 137 138/ College & Research Libraries • March 1973 to attain some of the goals hoped for a century ago as well as other goals en- tirely unanticipated at that time. Despite the fact that cooperation has long enjoyed a limited success in librari- anship, there exists no definition of li- brary cooperation, and this paper will not present one. However, there are at least three qualities that characterize li- brary cooperation. First, and of most importance, cooperation makes possible the establishment of new objectives for a group of cooperating libraries as dis- tinguished from the classical goals of individual libraries. Altogether too of- ten a library cooperative is thought of by its members as supplying each mem- ber with a service to further its own goals. Such service centers have enjoyed only limited successes. A truly coopera- tive center establishes goals that are not achievable by individual libraries. A second quality of cooperation is the sharing of resources without cost to the institution providing the resources. An example here is the shared cataloging project of the Ohio College Library Center ( OCLC), wherein it does not cost the original cataloging library any- thing to have another library use the central files. Of course, to be truly cooperative each participating institu- tion must do its small share of con- tributing cataloging information to the central data store. A third quality of cooperation is the pooling of human and financial re- sources to achieve a system unattainable by individual libraries; another example from OCLC will illustrate this charac- teristic. It is most doubtful that any li- brary in Ohio could afford to operate the computer system that OCLC main- tains. Moreover, there are few individ- uals qualified to do the research, de- velopment, and maintenance necessary to sustain the operations of such a cen- ter. The parent institutions of a num- ber of academic libraries participating in the center do not even possess a small computer, but OCLC makes available to their libraries computer power that the entire institution could not afford. Although computers possess a variety of capabilities which enhance library co- operation, only two of those capabili- ties will be discussed. The...Jlrst is the computer's enormous power to treat an individual person as a person, and an in- dividual event as a unique event. Sec- ond, a computer-based system has the capability of operating an information system that can supply information to a person when and where that person needs that information. NEW 0BJECI'IVES Cooperatives which possess expensive, manual union catalogs have sought to improve the bibliographic accessibility to resources within a region. Such union catalogs have attained only limited suc- cess because the catalog exists in but one place. What was needed was a technol- ogy that would make union catalog in- formation available throughout the re- gion. Until the advent of the computer, the only solution to this problem was an extravagantly costly printed book cata- log which never contained the listings of publications most in demand-name- ly current publications. The computer has made possible a less costly bookform catalog that can be kept more up to date than could manual bookform catalogs, but even this process does not thorough- ly solve the problem of providing infor- mation on news publications. A cooperative based on computeriza- tions, like OCLC, can establish the new objective of making up-to-date catalog information available to persons throughout the region-a heretofore un- attainable objective and, therefore, an undefined objective. At the present time, the OCLC system enters one-quarter of the institutional-holdings information into a central data base at the time cata- loging is performed, and the other \ j l "' 1 l three-quarters on the evening of the day the cataloging is done. At sometime in the future, the system will be modified so that all institutional-holdings infor- mation will be entered into the data base within seconds after cataloging has been completed. Libraries have long had the objective of participating in the intellectual ac- tivities of its users. In small libraries it is possible for the librarians to know their collection in detail and to know their users as individuals. In this attrac- tively human situation, the librarian can often participate in the educational, re- search, or informational activities of in- dividual users. The same type of partici- pation also occurs in intelligence agen- cies which are designed to obtain and communicate information to a decision maker before the decision maker knows the information exists. However, intel- ligence agencies have relatively few users, and in terms of number of users and amount of information processes, their expenditures are high. Of course, high expenditures are justified because of the extreme importance of the infor- mation. The library . lost the ability to partici- pate in the educational, research, and informational activities of users as its professional staff grew in numbers. But cooperation based on computerization can make it possible for large libraries once again to participate in the infor- mational activities of their users. In- deed, an exciting assemblage of tech- niques to achieve such participation is operating successfully already or is vis- ible immediately ahead. In information processing, it is SDI services which have done most to demonstrate the ability of the computer to treat a user as an indi- · vidual; SDI services have even demon- strated their value to individual librar- ians by enabling them to keep up with the publication of new findings in li- brary research and development. Computerized cooperation also makes Computer-Based Systems I 139 possible establishment of a new econom- ic goal for libraries. Traditionally, the economic goals of libraries have been to design and operate efficient proce- dures that eliminate unnecessary costs. The technique is sometimes described as "saving money," but it is doubtful that any library has been able to reduce overall expenditures by improvement in the efficiency of its procedures. The in- exorable rise in library salaries caused by rising living standards in the commu- nity as a whole causes an increase in ex- penditures despite economies of proce- dure. This circumstance can be expressed as a rate of rise of per-student costs in academic libraries. In the past couple of decades, this rate has been 5 percent as compared with a somewhat more than 2 percent increase in per man-hour cost in the economy as a whole. 2 Until the advent of computerized cooperation, it was impossible for librarianship to de- velop a strategy which would decelerate the rate of rise of per-student library costs and to bring that rate in line with the rate experienced in the general econ- omy. It is now clear that it will be neces- sary to invoke computer technology to increase continuously library productiv- ity in order to decelerate the rate of rise of per-student costs. For the entire li- brary population, it now appears that the only way that computer technology can be invoked is through the coopera- tive use of computers. The highly efficient manual proce- dures developed by libraries in the past century have caused an imposition of uniformity that restricts the thinking of librarians about library processes. James Duff Brown, pioneer British li- brarian, detected as early as 1903, the deleterious consequence of uniformity when he wrote, "The hampering effects of too much uniformity are to be seen in full operation in France and the United States."3 A decade later, Bro,\rn exhorter1 the librarians of New Zealand 140 I College & Research Libraries • March 1973 to avoid "paralysing stagnation." He went on to say, "One thing to be said for the standardized American methods of library management is, that they do produce a general level of efficiency, which, at least, assures an all-round me- diocrity," and then pointed out that, "Standardization ... tends to stifle all future progression."4 A simple and trivial example of stag- nation caused by uniformity is the tech- nique of employing unit cards in card catalogs. Such cards are uniform except for call numbers and entries typed on them. However, there is no need for tracings to appear on all cards, but until recently, many librarians accepted their presence in an essentially unthinking manner. With computerization it be- came possible to treat each card individ- ually, the tracings need be printed only on specified cards. It has not been easy for many librarians to divest themselves of beliefs imposed by such simple uni- formity, largely because until recently there was no opportunity to think about such a matter. Nevertheless, there are li- brarians who insist that the tracings should be on all cards largely because tracings have "always" appeared on all cards, at least throughout their profes- sional careers. Computerized cooperation will destroy the tyranny of uniformity which has more and more narrowly confined the effective imagination of librarians as li- brarians have increasingly imposed uni- formity on their manual processes. Once again, a librarian can experience the de- licious euphoria which springs from the creation of a completely novel idea which can be transformed into action. Once again, librarians will be able to function as intellectuals in the true sense of the word. INSTABll..ITY Cooperative librarianship does not, however, bask in the security of time- tested principles of organization, can- not afford the false facade of self-suffi- ciency, cannot relax in simplistic tech- niques of communication and the low technology of typewriters, and is utterly destitute of qualified manpower. At the present, there is no organiza- tional design of choice for computer- ized library cooperatives. The Ohio Col- lege Library Center is a corporation made up of constituent members that are academic institutions, not academic libraries, in Ohio which have paid a membership fee for the current academ- ic year. This type of organization has worked well for five years. However, it must be pointed out that it was the pres- idents of the institutions, not the librar- ians, who brought OCLC into being, and that the presidents and librarians had been talking and working with each other, off-and-on, for a decade and a half prior to the incorporation of OCLC in 1967. These two circumstances have certainly contributed to some ex- tent to what appears to be organization- al effectiveness. OCLC' s five years of experience is in- adequate to warrant recommendation of its type of · incorporation as a model Other regional cooperatives should ex- amine existing organizations and adopt a policy of eclecticism. Some regional cooperatives have suf- fered difficulties in obtaining exemption from the payment of income tax. OCLC, for example, had to change its charter at the behest of the Internal Revenue Service to qualify for exemp- tion. Other cooperatives have been un- able to obtain exemption or have re- ceived only partial exemption. Orga- nizers should be certain that the form of organizational structure selected will qualify the cooperative for exemption from income tax payments. Computerized cooperatives should not presume to be self-sufficient, just as indi- vidual libraries, no matter how large or small, can afford such a presumption. In- trinsic to cooperation is recognition on Computer-Based Systems I 141 the part.- of an · institution ·of .the fatal- knowledge, patience, experience, and ex- ity of seeking self-sufficiency, and just cruciating pressure can yield a telephonic as individual libraries should cooperate network operating effectively. with each other so should regional or- Similarly, librarians participating in ganizations cooperate. Indeed, coopera- the development and operation of com- tion among regions will extend coopera- puterized cooperatives must possess some tion nationwide and internationally, and knowledge of the operation of comput- will enhance the availability of re- ers and of the characteristics of the sources while at the same time will fur- world of computation. The extent of ther reduce duplication of effort. In oth- this knowledge must be sufficient to en- er words, cooperation among regions will able the librarian to understand the po- further decelerate rate of rise of per- tential of the computer as well as the student costs. potential of libraries, for only by Extension of cooperation among com- thoughtfully combining these two po- puterized regional cooperatives will not tentials can a librarian make a contribu- be achieved easily. First it will be neces- tion to the new librarianship. Altogether sary to design a complex, computerized too often library computerization con- interregional communication system sists of a computer programmer com- in order to solve hitherto unsolved tech- puterizing · an existing procedure. Such nical problems. There is no reason to computerization is tragically inadequate, ·think that these obstacles cannot be for it exploits neither the potential of overcome, but it will take time and spe- the library nor that of the computer. . . cial competence ·to ·'successfully elimi- A major source of instability in . com- nate present ignorance. Moreover, it puterized cooperation stems from the will be necessary to establish some kind fact that computer hardware which will of a national library network agency be necessary to support cooperative sys- to develop, operate, and maintain the terns in the foreseeable future does not network. The informal library networks presently exist. Dependence on others to of the past and present are completely develop the new equipment required for inadequate for the immediate future. continued operation does not impart a On-line, computerized, telephonic comfortable sense of security to which communications employed by comput- librarians have long been accustomed. erized cooperatives have no predecessors Nevertheless, computerized cooperatives in classical librarianship. Hence, the can exist only by taking such calculated new librarianship has an entirely new risks. area of technology about which it must OCLC requirements for secondary become effectively knowledgeable; such storage memories furnishes an example knowledge involves at least a smatter- of such a calculated risk. The OCLC on- ing of telephony, an acquaintance with line catalog is growing at such a rate complex rate structures, legal restric- that present equipment will, by early tions, and telephone company policy. 1973 be inadequate for reasons of Unhappily, even a thorough knowledge equipment expense and available floor of these topics cannot provide commu- space (file organization is entirely ade- nication effectiveness because telephone quate to support a much larger file than companies do not have the necessary present equipment and floor space can know-how and techniques to design and house). During the first third of 1973, operate a large network without falling ·. ; OCLC will receive new, as yet unavail- into .an unknown number of self-de- able, secondary memory equipment signed · traps into · which the computer- which will more than double memory i~ed coopera_tive is also cat~pulted. Only capacity per square foot of floor space. -:~: ~ :·, · . . : : ~ -: ·-- • .. . ~ · · : , . ~ ··-- . ~ ; ··--· · ·- - · ·· · ' ·~ . . . '· · -• . . ·- · ... .. ·r . 142 I College & Research Libraries • March 1973 The equipment will be adequate for a year or so at which time still higher den- sity memories are promised. When the file overflows this equipment several years hence, it will be necessary to ob- tain wholly new types of memories now in the early stages of development but not yet in operation except in purely ex- perimental laboratories. Fortunately, there remain changes in file maintenance that can be invoked if the equipment development timetable slips further into the future than is now estimated. Computer software contributes more instability to library automation than does hardware. Computerized coopera- tives must attract the most highly quali- fied, imaginative, and effective program- mers who can be located and persuaded to join the cause. Until recently, there has been no measure against which pro- grammer performance could be judged, but by now, enough cost-beneficial li- brary programs exist so that specific goals can be set for programmers. In- competent programmers, of whom there are altogether too many, program dis- aster. One instance is known in which a catalog-card program producing unit ~ards required several hundred times the amount of computer processing as the OCLC program that puts out catalog cards in final form, ready-to-file. It is ab- solutely necessary that such disastrous programming be eliminated from li- brary automation. The highest obstacle in the path of evolving computerized cooperatives is the near total absence of libra rians pos- sessing an effective knowledge of com- putation and of system programmers with experience in designing and pro- gramming complex, character manipula- tion systems. Librarians must be suffi- ciently and intimately familiar to be able to combine the potentials of librar- ies and computers into new systems. The technique of using truncated search keys to access the computerized catalogs as employed at the Ohio State University libraries and OCLC is an example of a type of catalog searching unfamiliar to many librarians. 5 The prospect of com- puterized descriptive cataloging pushes the new librarianship a considerable dis- tance from classical librarianship.6 To make effective decisions in the area of library automation, librarians must know much more about computation than they think they must know. The bright side of this dismal picture is the vista ahead, for surely part of man's greatest happiness is learning. SUMMARY The combination of cooperation and computation makes possible the estab- lishment of new library objectives un- attainable by individual libraries. Per- haps the most important of new goals is that of economic viability-a goal to which individual libraries cannot aspire. Computerized cooperatives do not, however, enjoy the comfortable stability of classical libraries. Indeed, their in- stability, not to be confused with in- security, can be a frightening experience for librarians. Moreover, librarians, as have physicians, engineers, and members of other rapidly developing professions, must redirect themselves to become per- petual students-an enterprise in learn- ing that produces the most pleasurable of human gratification. REFERENCES 1. F . G . Kilgour, "Standardization for Inter- change of Cataloging Records-MARC II," Proceedings of ihe Third lnterootional Con- gress of Medical Librarianship, Amsterdam 5~. May, 1969, p . 103-109. 2. William J. Baumol, et al., "The Cost& of Li- 4 ' t ! brary and Information Services," Libraries at Large (New York: R. R. Bowker, 1969), p. 168-227. 3. W. A. Munford, ]ames Duff Brown, 1862- 1914; Portrait of a Library Pioneer (London: The Library Association, 1968), p. 60. 4. Ibid., p. 91-95. 5. Gerry P. Guthrie and Steven D. Slifko, "Analysis of Search Key Retrieval on a Large Bibliographic File." (In press); F. G. Kil- gour, P. L. Long, and E. B. Leiderman, "Retrieval of Bibliographic Eatries from a Computer-Based Systems I 143 Name-Title Catalog by Use of Truncated Search Keys," Proceedings of the American Society for Information Science 7:79-82 ( 1970); F. G. Kilgour, P. L. Long, E. B. Leidennan and A. L. Landgraf, "Title-Only Entries Retrieved by Use of Truncated Search Keys," Journal of Library Automa- tion 4:207-10 (Dec. 1971). 6. F. G. Kilgour, "Concept of an On-Line Com- puterized Library Catalog," ] ournal of Li- brary Automation 3:1-11 (Mar. 1970) .