College and Research Libraries THOMAS F. PARKER Models and Methods: The Tools of Library Networking Several models and methods applicable .to university library networks are discussed as tools to increase understanding of cause and effect re- lationships in these complex organizations. Collection of data by un- obtrusive measures and display of information by graphic techniques are briefly described. The utility of the interventionist method, orga- nization development techniques, and pilot projects is outlined. Levinson,s organization diagnosis techniques and Meier's communica- tions model depict university libraries as stressed organizations. Other models applied to the university library context include a network model, Sloan,s exchange model, Thompson,s organization action model, and Saeger, s vender model. M ANAGEMENT OF LARGE HUMAN OR- GANIZATIONS is one of the great chal- lenges of this century. Lack of sufficient understanding of the cause and effect relationships operating in large complex organizations can demonstrably lead to such difficulties as the energy crisis, the inflation crisis, the unemployment crisis. Library networks will scarcely cause ma- jor crises in themselves, but networks in difficulty can affect the lives and for- tunes of hundreds of people and will involve institutions which represent large capital-intensive investments of communities, cities, states, or universi- ties in the range of millions of tax or gift dollars. Understanding cause and Thomas F. Parker is a library manage- ment consultant and served as project di- rector of the Library Study Project for the libraries of the California Institute of Tech- nology; University of California; Los An- geles; and the University of Southern Cali- fornia. This project was supported by the National Science Foundation (Grant GN- 33010). 480 I effect relationships in library networks is an undertaking worthy of consider- able -attention. Effective management of these enterprises is absolutely dependent on accurate knowledge of how library networks operate, the nature of their components and functions, the effective- ness of development methods, and which methods operate at least cost in human stress and scarce dollars. J;i'ortunately for those who seek un- derstanding of library networks a num- ber of models and methods are avail- able to increase understanding of cause and effect relationships and workable development approaches. The Library Study Project-a feasi- bility study of library cooperation among the libraries of the California Institute of Technology; the University of California, Los Angeles; and the University of Southern California- had an opportunity not only to survey available models and methods but to as- sess their worth- in actual field experi- ence. Among the models utilized were a network model, the exchange model, the organization action model, the com- munications model, and the vendor model. Useful methods employed in- cluded graphic data display, the inter- ventionist method, organization devel- opment, organization diagnosis, the use of pilot projects as change tools, and unobtrusive measures for data collec- tion. The models are all narrative-En- glish~language, nonmathematical mod- els. The methods have all been applied in the field · o{ management but have seldom, if ever, been utilized in an aca- demic library environment. COLLECTION AND DISPLAY OF DATA One of the most challenging tasks facing the inanager or analyst of large complex organizations is that of making the organization observable. The prob- lem has been effectively formulated in the field of urban analysis, and various methods are ·employed to make urban systems observable. Library administra- tors have long used a variety of personal methods to keep track of the variegated operations of their organizations, in- cluding various reporting structures, hierarchical staff meetings, and disjoint- ec;l incrementalism. Multiple operation- ism is frequently resorted to in a recog- nizable pattern identified as a manage- ment "style." The Library Study Project faced the problem of making three large, dispar- ate libraries observable within the con- straints. of limited project staff, project schedule limitations, and the necessity of avoiding disruption of library opera- tions. Data collection was carried out using the unobtrusive measures tech- niques formulated by Eugene Webb and others at Northwestern University.1 The approach assumes that the environ- . ment is open, that multiple operation- ism must be used to avoid methods that sha.re the same weaknesses, and that the analyst ·is always deaHng with approxi- matioris to .knowledge. : Sou·rces of mea- Models and M.ethods I 481 sure invalidity are explicitly recognized. The approach of choice utilizes simple observation, contrived observation, and physical traces that include . erosion and accretion, running records, and episodic records. The Library Study Project col- lected all publicly available library in- formation handouts, library guides, and various public reports. Limited use was made of internal reports, other than an- nual ones, although other reports were 'freely made available by the libraries. In addition, the project collected stu- dent newspapers, publi~ity releases, and local newspaper clippi11gs on a syste- matic basis. The publicity releases car- ried v.aluable information on substan- tive changes occurring in the universi- ties and the libraries, as did the news- paper clippings. Student newspapers carried invaluable reports of official and unofficial events, including campus con- troversies and any problems encountered in using the libraries. Project staff em- ployed simple observation te.chniques in interviewing library personnel in the dozens of library units encompassed in the study. The amount of data collected during the life of the project was substantial. To increase utility of the material, the information ·collected · was summarized and displayed in graphic fashion. -Graphic data display is a simple exten- sion of the familiar use of · maps, How :charts, and organizational charts. How- ever, in the project graphic display was used as a deliberate tool in organiza- tional problem solving. The project employeq a wide range of graphic display options, including circle graphs, bar graphs, symbolic- graphs, maps, and symbolic library- process How charts devised by the proj- ·ect. Bar charts were created for operat- -ing statistical' data. Hours· of library operation were shoWn~ and unit ·and ser- vice .loqations and library-to-library ·unit intenictions were mapped~ Administra- tive org~nization was charted. Library 482 I College & Research Libraries • November 197 5 staff size and personnel distribution were developed from analysis of. staff lists. NETWORK FEASIDILITY The potential for a library network among the three libraries was explored using the interventionist method, orga- nization development techniques, and a variety of pilot projects. The interventionist method has been formulated by Chris Argyris at Yale University, one of the foremost au- thorities on organizational behavior in the country. 2 The method focuses on three tasks: generating valid and useful information for the client systems, helping the clients make free and in- formed choices among their options, and generating firm client commitment to their choices. The interventionist aims at aiding the client system in collecting and commu- nicating valid and accurate information about the realities of the situation, without distortions from the biases and defenses of the client systems, other de- cision makers, or the interventionist. The interventionist works with the cli- ent systems to generate free choice, i.e., choice made with minimal internal de- fensiveness, based on valid information related to the client's main goals and needs and incorporating a realistic, chal- lenging level of aspiration. The inter- ventionist strives to represent the whole client system in his efforts, not the point of view of any group or individual, or his own. Once free choice has been made, the last task is that of aiding the client systems in developing internal commit- ment to the path selected. Anyone em- barking on a new venture is beset by doubts, and so are library administra- tors. Therefore, the interventionist pro- vides support for the clients~ free choice. When these three tasks have been completed, the interventionist then terminates association with the client systems, which have been strengthened by the intervention. The contrast be- tween the interventionist method and the more frequently encountered "sell- ing" of system change is striking, since the interventionist method is devoid of manipulative aspects. ORGANIZATION DEVELOPMENT In addition to the interventionist method, the Library Study Project uti- lized techniques of the field of organi- zation development, or OD. Organization development is a rela- tively new field in the area of manage- ment. It grew out of work done at Case Western Reserve University, Harvard University, Massachusetts Institute of Technology, University of California at Los Angeles, and Yale University. Some of the pioneers in developing the field include Chris Argyris, Richard Beckhard, Kenneth Benne, Warren Ben- nis, Robert Blake, Jane Moun ton, and Herbert Shepard. An excellent summary of the field has been published by French and Bell. a Organization development focuses on organization behavior interventions de- signed to build action teams. In OD work considerable attention is paid to role and goal clarification, interpersonal and interunit relationships and conflicts, client underutilization of each other as resources, and the processes of commu- nication, decision making, and task allo- cation. Typically an organization devel- opment effort is based on an action-re- search model that begins with data gathering and situation assessment and moves on to feedback to the client sys- tems, followed by discussions and work on the feedback and on emerging or up- dated information, joint action plan- ning, action, and further data gathering and feedback, with reiteration of the cycle. The action-research model was prominent in the activities of the Li- brary Study Project. Early in the study the project team took on the role of a neutral buffer operating in · a constant information-feedback mode. In · the two-year life · of the project, team mem- bers held over 400 meetings with librari- ans, library staff, and university staff and administrators. PILOT PROJECTS IN NETWORK DEVELOPMENT To enhance the· information being developed and fed back to the partici- pating libraries, a series of pilot · projects were developed -and implemented. For libraries to make free and informed choices on the feasibility of resource sharing, it is desirable for them to have first-hand information on the risks, po- tentials, and benefits; Pilot projects pro- vide opportunities for libraries to ef- fectively explore mutual arrangements, the impacts, benefits, and difficulties in a controlled, safe, · but real-world en- vironment. In the feasibility study, pilot projects were developed, imple- mented, and monitored on the basis that they could be terminated at any time, without prejudice, by any participant who wished to withdraw. The pilot projects included a delivery and passen- ger service, which proved to be a useful interlibrary link, and test utilization of teletype and telecopiers, which were not successful as such, but were useful to the libraries as additional opportunities for joint interaction and decision mak- ing. LIBRARIES AS STRESSED ORGANIZATIONS Large academic libraries are no dif- ferent in many aspects from other large, complex organizations. In the course of their development they en- counter stressful conditions with which they must cope to survive. Assessment of the stress levels in organizations is as important to library network develop- ment as the activities of a doctor exam- ining the health of a football team are to team performance. If one team member is severely stressed, the team ·· · ·Models and Metholk I 483 may inadvertently place an overload on that position (or node i'ri a network), and team failure may result. A technique for · assessment of organi- zational stress has been developed by Harry Levinson at Harvard University.4 Levinson divides activities that organiza- tions undertake as stress responses into adaptive and maladaptive categories. He classifies some 110 specific activities as normal, exaggerated normal, maladap- tive, disruptive, or self-destructive. The Library Study Project found that none of the three libraries studied was com- pletely free from activities in the range from riormal through disruptive. In addition to Levinson's work, the communications model of a university library developed by Meier identifies communications overload stress respon- ses and the consequences.5 Meier found that ·severe tests of the library resulted in spatial consequences, status structure modifications, changes in the communi- cations structure, and alterations in the value and relationships structures of the library. Again, the project team found that all three libraries evidenced severe· stress responses. It is clear from the standpoint of large organizations or the· viewpoint of libraries as communica- tions systems that academic libraries are likely to show substantial scars of severe stress. MoDELS oF LIBRARIES IN AcnoN Besides Meier's model, the Library Study Project utilized four other socio- logical noriquantitative models to en- hance understanding of the complex- cause and effect relationships at work in university libraries. A network model was formulated to provide insight into the structure of li- brary-to-library interactions. Several for- mulations of the network model are available in library literature. The mod-· el synthesized by the Library Study Proj- ect viewed a library network as having five basic· components: resources (books~ 484 I College & Research Libraries • November 1975 periodicals, other library materials, and staff expertise), directories to the re- sources (card catalogs, printed catalogs, holdings lists, in-process files), commu- nications (telephones, teletype, material delivery, passenger service), users ( fac- ulty, graduate students, undergradu- ates), and management (consortium management, library administrations, committees, task teams). In the course of the feasibility study the project team paid considerable at- tention to the skills required for suc- cessful library interaction. The team found that a library network clearly re- quires · a specialized and not widely rec- ognized set of skills that are boundary- spanning in nature. These boundary- spanning skills are different in type from the skills utilized inside a single, even multiunit, library. The skills in- volve the ability to carry out effectively the interactive coordination of network activities; regular information exchange; development of an understanding of each member library's policies, proce- dures, and problems in a nonjudgmental fashion; and development of an under- standing of each member library's larg- er organization, its parent campus, and campus administrative policies and problems. Boundary-spanning skills also include a capability to accommodate role changes: former rivals are now sudden- ly to be treated as part of the home team. Absence of an adequate level of bound- ary-spanning skills will prevent effective ongoing network operation. The study team analyzed the manage- ment requirements of a library network. The findings indicate that network man- agement is unexpectedly complex. Li- brary network management includes operations management, service supplier management, service user management, and service broker management func- tions; and these are usually dispersed. Network operations management is re- quired for network facilities which must be obtained and installed and for protocols which are developed for op~ eration, maintenance, and feedback on stability. Network management must identify · service suppliers, assess their capabilities, and define the type of ser- vice they are willing to supply. Potential network users must be identified and frequently aided in effective use of net- work services available to meet their needs. The service broker function must be performed to bring potential users and suppliers together and assist them in making arrangements for service ex- change over the network In library net .. works, management functions may be dispersed or · concentrated, performed entirely by member libraries, by a cen- tral group, or by some combination. ExCHANGE MonEL OF LmRARY ·RESOURCE SHARING The exchange model of library re- source sharing . developed by Sloan was found to be of particular use in assess- ing the readiness of libraries to form a network. Sloan based her model on work by Thompson (described below), with modifications drawn from small- group interaction and coalition forma- tion theory. 6 The exchange model iden- tifies the prerequisites of library re- source exchange, discusses the determi- nants of exchange (range, direction, de- gree of dependency, structure of units, and domain ·consensus), and develops five indices of readiness to cooperate .. In these the degree of domain consensus is seen as crucial, with important influ- ences emanating from the existing moti- vation to coalesce, the amount and type of resources relevant to the exchange, the values placed on alternative expecta- tions, and the influence of nonutilitari- an strategy choices. These last include prestige, influence on the group, power or leadership elements, and desire to ex- pand domain. The Library Study Proj- ect found that the exchange model cast important light on several aspects of co- operation potential, providing valida- tfon or indicating the need for further checking of project hypotheses on po- tentially fruitful areas of cooperation. THE ORGANIZATION AcnoN MoDEL The organization action model devel- oped by James D. Thompson proved to be quite applicable to university li- braries.7 Thompson~s model regards complex organizations as operating a certain type of technology (which can be manual) to perform an instrumental task. The organization is made up of a set of components (institutional, man- agerial, cote technology, and boundary- spanning), each with specific functions, problems, and a finite range of coordi- nation options. The organization per- forms its instrumental task in an inter- active task environment containing sup- pliers, clients, competitors, and regula- tory groups. The organization faces ex- ternal and internal contingencies and constraints that threaten the organiza- tion's survival and must be managed. To manage these threats the organiza- tion develops internal and external in- teraction strategies to achieve homeosta- sis in the face of environmental fluctu- ation. The organization is dependent on human discretion to operate the core technology, span the boundaries, man- age the activities, and implement the survival strategies. To obtain this cru- cial human element the organization ne- gotiates inducement-contribution con- tracts with individuals with the neces- sary skills or capabilities. The Library Study Project found that the organization action model pro- vides an effective overview of university libraries operating in complex environ- ments. The model supplied clear defini- tion of the similarities and differences of the libraries studied. The model has great potential for supplementing the experience and intuition of library ad- ministrators by defining the appropriate internal structures to meet external con- Models and Methods I 485 tingencies, external interaction strate- gies, and effective internal coordination as required by circumstances. The mod- el will be useful in predicting crises in the external environment, in identify- ing areas in which the library is vulner- able, and in predicting the impact of proposed changes in library policies and procedures. THE VENDOR MODEL In addition to the :models identified and retrieved from the literature on or- ganizational behavior, one model was developed inside the project. The proj- ect's senior systems analyst, JoAnn Sae- ger, successfully formulated the vendor model, which defines the complex inter- action that takes place between a vendor and a customer. 8 The model describes four time-sequenced stages: initial con- tact, test contract negotiation, contract implementation, and contract evalua- tion. The model identifies the key issues and key individuals at each stage and predicts individual behavior as the se- quence progresses. The model is partic- ularly accurate in predicting the be- havior of the contact person inside the customer (read library) organization, with goal ranges, task types, and posi- tions at contract evaluation time. The model also outlines the main tasks of both the vendor and customer organiza- tions, with implications for customer responsibility. The model highlights the crucial role that must be played by the library systems analyst as intermediary, with analytical, specification writing, legal, and program-management func- tions. The vendor model will be useful in procurements of any system of such complexity that evaluation is difficult (e.g. , a computerized acquisitions or cir- culation system). SUMMARY The Library Study Project was fortu- nate in having the opportunity to iden- 486 I College & Research Libraries • November 1975 tify, examine, and apply a number of organizational behavior methods and models in a real-world environment. Un- obtrusive measures, graphic data dis- play, organization development meth- ods, pilot projects, and stress analysis are effective tools for network development or · improvement. The network model, the exchange model, the organization action model, and the communications and vendor models contribute substan- tially to an understanding of the vital cause and effect relationships in the complexities of library networks. REFERENCES 1. Eugene Webb arid others, Unobtrusive Mea- sures: Non-Reactive Research in the Social Sciences (Chicago: Rand-McNally, 1966). 2. Chris Argyris, Intervention Theory and Method: A Behavioral Science View (Read- ing, Mass: Addison-Wesley, 1970 ). 3. Wendell L. French and Cecil H. Bell, Jr., Organization Development: Behavior Sci- ence Interventions for Organization Im- provement (Englewood Cliffs, N.J.: Pren- tice-Hall, 1973). 4. Harry Levinson, Organization Diagnosis (Cambridge, Mass. : Harvard Univ. Pr., 1972). 5. Richard L. Meier, "Information Input Over- load," Libri 13:9-44 ( 1963 ). 6. Elaine F. Sloan, "Toward an Understanding of Library Cooperatives as Organizations" (ERIC, 1970-). ED 047 752. 7. James B. Thompson, Organizations in Ac- tion (New York: McGraw-Hill, 1967). 8. JoAnn Saeger and Thomas F. Parker, "A Bridge between Two Worlds: The Interac- tion between a Library and a Vendor of Ad- vanced Technology" (unpublished).