Navigation

Reference Service Structure -- Centralized and Decentralized Schemes

Copyright 1990, 1998 by Leigh Kimmel

For permission to quote or reprint, contact Leigh Kimmel

As libraries, particularly in academic institutions, grow larger and larger, there comes a time when the issue of the structure of the reference department comes up. This issue centers around the question of whether the reference collection should be kept in one centralized unit, with all reference work done from it, or whether reference should be distributed into a number of subject-specialized departmental libraries.

The departmental libraries which are found scattered about the campuses of many major academic libraries have their roots in the seminar collections used in Germany. In the German system the professors would draw together collections of materials specially related to the seminars they were holding, and then would see to it that the students who needed the materials got access to them. In time these libraries, which started out more or less informally, became institutions in their own right with a place in the organizational structure of the university.1 When the modern American university began to arise, it used the German model as the basis for much of its structure, and this included the structure of the library. Thus many American university libraries have at least a few departmental libraries.

However there was at the same time a noticable push towards keeping all library materials in a central location. Each side proposed their own arguments as to why their scheme of library structure was the best, and why the other scheme was less than completely suitable. Even as early as 1939 there was an extensive article in ALA Bulletin which discussed the advantages and disadvantages of each scheme of organization, then concluded that each had its particular values.2 Another article from the same era suggested that the collection itself should not be divided and scattered about the university, but the librarians should each specialize in one particular literature in order to give more effective reference service.3

Others questioned just what kind of centralization was being discussed. Were they talking about physical centralization of the collection, emphasizing that all of the books be gathered into a single physical location? Or were they talking about administrative centralization, arguing that all reference activities should be placed under the control of a single head of reference services rather than be handled by the various departmental librarians who might report to a head of departmental services or even the head librarian. Or were they talking about operational centralization, arguing for the importance of keeping a single reference desk that everybody knew about and recognized?4

More recently there has been some questioning of the idea that reference librarians need extensive subject specialization in order to handle most reference questions. In fact it has been argued that specialization may actually limit the effectiveness of the reference librarian by tying one down to an unnaturally narrow collection.5 A librarian who is familiar only with the materials in his or her specialty, so the reasoning goes, will only use those materials and will be unlikely to turn to materials outside that narrow specialty, even when the other materials might be more appropriate.

Thomas Watts, in his "Brief for Centralized Library Collections" gives five arguments against departmental libraries.6 His first arguement is the growing interdependence of knowledge, which frequently makes it necessary to consult materials traditionally associated with a number of different disciplines. If the library is arranged along departmental lines, it may well be necessary to spend as much time walking all over the campus to the various libraries as it takes to actually find and examine the items in question. Also reference librarians who get these sorts of questions may well find themselves shuttling their patrons all over campus to get the answers.7 Equally, the patrons may well never even bother to go to some of the other libraries where relevant items are held, particularly if they perceive those libraries as being outside their field. From this he derives his second arguement, that of user convenience. As he sees it, the scattering of the collection all over the university in various specialized collections makes it harder instead of easier for the patron to use the library.

His third arguement is that the various collections may well become isolated. There have been cases in which the librarians of various departmental collections in a university have become so tied up with their own collections that they have lost sight of the mission of the library as a whole. Instead they have divided up into armed camps that8 spend their time trying to cut down the librarians of rival departmental collections in an ugly private war that leaves everybody losers.

His fourth arguement for centralization is that of expense. By running a number of departmental libraries, the library ends up spreading its staff too thin and duplicating materials needlessly. Also on many university campuses the departmental libraries are kept in grossly inadequate quarters and given castoff equipment from the main library. Thus the patrons end up being shortchanged by getting only halfway service, although the departmental library may well bring together all the materials they need for their work and have subject specialists as staff. It would be far wiser to staff one central reference desk adequately than to staff a dozen departmental ones halfway.

His final argument is that the existance of multiple departments scattered about the campus fosters problems of inadequate interdepartmental communication. This is of course related directly to his third arguement, that of the growing isolation of collections fostered by departmentalization. He claims that the fragmentation is also self-perpetuating, as the more departmental libraries exist on a campus, the more small disciplines will demand its own library in its hall.

Often problems with departmentalization tend to show up in the area of the hard sciences. Here the scientist's most important books are often kept in the laboratory in a special "lab collection" that frequently belongs to the academic department instead of to the library. Thus many science collections remain centralized, as there is little demand for a special collection for each of the disciplines within the hard sciences.9

However the people who champion the idea of the decentralized library with each special collection kept together have powerful arguements in their favor. For instance it is often pointed out that special collections can become "buried" if they are incorporated into the central collection.10 This has happened at the University of Illinois, where a number of special collections including at least one notable collection in library science have been allowed to get "lost" in the central bookstacks because no special record was kept of them.

Also many proponents of decentralization are quick to point out that simply putting all the materials together in one place won't necessarily overcome the internal barriers that may be set up by the patrons themselves who are unwilling to look outside their own fields.11 Often librarians become familiar with a certain group of materials they use frequently, and are extremely reluctant to use other sources, even if they are close at hand. If a person isn't willing or able to think of using a certain material, it might as well be on the moon for all the good it does.

Also a very large collection may well be overwhelming to the user, and even to a librarian who is trying to be aware of all the possible materials in a field. Many people have spoken of their sense of being lost when confronted with the central bookstacks or the main reference room at the University of Illinois. By contrast, a well- selected departmental collection that is located in a place convenient to the users might well be more comfortable in consulting.12

Also departmental libraries are able to serve groups with needs that are distinct from the majority of users and might be ignored in a centralized library. The number of small informal libraries scattered around the campuses of many universities are evidence that somebody isn't being served by the library, and is seeking alternative ways to fulfil those information needs.13

Further, the presence of departmental libraries can serve a valuable social function as well. Having all the materials dealing with a given subject gathered together into a departmental collection tends to bring their users together as well. Thus the students and faculty in a given field of study are brought together to socialize and work on problems co-operatively, while in a centralized library situation they may well never notice one another for being lost in among the multitude of people from other disciplines. This social function is particularly important in such disciplines as law where the professors want their students to spend time discussing questions with one another.14

Proponents of decentralization are also quick to point out that the problems inherent in departmentalization can be overcome by a little thought. For instance the problem of the interdependence of knowledge can be solved by using cross-references to link various literatures that have traditionally been regarded as separate.15 And computer technology, in particular online catalogs, can solve one of the greatest problems of the traditional departmental library -- how to find out if other departments might have the item that is sought if it is not in that department. In this way it is ironic that the centralized automation allows greater decentralization of library service.16

The use of computers, in particular the online catalog, help to break down the barriers that have often impeded interdepartmental communications in very large libraries, and has also improved access to very large and unwieldy libraries, whether they are centralized or departmentalized.17 Most notably, many of the files that were traditionally kept on paper in a single site, particularly certain files used in the cataloging and acquisitions departments, become accessable everywhere when they are put online. While it would have been impractical if not downright impossible to duplicate these files in paper form at all the sites where they might be used, when they go online all that is required to access them at remote sites is a computer, a modem and a telephone line that connects with the central computer on which the files are found.18

However these online catalogs need to be designed properly or they can end up being worse than the old paper forms. For instnace, it may be difficult to zero in on items in one's own departmental collection through an online union catalog that doesn't have the ability to use location as a limiter on its searches.19

Also departmentalization has an administrative benefit. The larger an organization becomes, the more time and effort must be devoted to keeping track of everybody in the organization instead of doing productive work. When the organization becomes large enough, there may even be people whose sole duty is to keep track of the duties of other employees. In this sort of situation the interaction between employees becomes very formal, and people communicate primarily by memos instead of talking to one another. By contrast, if one can keep the size of the organization small (under twelve staff members), things can be kept on a less formal level and people can communicate directly and spend most of their time on productive work instead of bureaucratic monkey-business. In fact it is known that workers in big central cataloging departments often envy the people in the branch libraries.20 Also it is known that patrons like small units rather than larger ones. Perhaps this is due to the atmosphere of informality that is fostered by the smaller organization. Patrons may well pick up on the stuffy formality of the organization of a large centralized library and be put off by it, while they feel that they are being treated personally in a smaller unit.

However the decentralized system puts a heavy demand on the head librarian, who must be able to work collegially with the staff and co-ordinate library activities to prevent conflicts between the departments.21 This sort of juggling act requires vast ability on the part of the head librarian, and not many people are up to that kind of challenge. The great geniuses of librarianship have been able to do this with consummate skill, keeping large numbers of departmental libraries working together toward a common mission. However less able administrators have been unable to handle this balancing act and have ended up instead setting one department off against another and fostering rivalries that end up with lines of battle drawn across the organizational structure and blocs of hostile departments spending more time fighting one another than serving their public.

If we are to departmentalize library service, how are we to go about it? There are a number of ways in which library services can be decentralized. There is operations-oriented decentralization, in which the library is divided by kinds of material. For instance all microforms may be gathered together into a microform library, and maps may be gathered into a special map library, while rare books are gathered into their own special collection. This has its distinct advantage in that the librarians at such a collection are able to understand the peculiarities of the material, as they spend all their time working with that one kind of material. Another scheme is user-oriented decentralization, in which the libraries are divided as to what group of users, such as undergraduates or graduate students, are most likely to use it. But the most common means of decentralization by far is the subject departmental scheme, by which materials dealing with a particular subject are gathered together into a particular library.22

Subject specialization seems to improve service the most of any of these schemes of decentralization. It divides the library's clientelle into manageable segments. Thus accuracy and efficiency of service are improved.23

The administration is usually centralized even when the actual service is departmentalized. However medical or law libraries are frequently administered separately, some times even being a part of the law or medical school that they serve. In these cases the libraries have arisen independently of the main library, instead being sponsored by the educational department. If one is to try to forcibly bring them under the administration of the main library, there is often hostility and the library ends up being worse off by having them in than by having them independent.24

Harvard is an example of a highly decentralized library. It has ninety units (by contrast the University of Illinois has only thirty- five) which are united by a union catalog of all their items, as well as informal contacts between the librarians that fosters a sense of common purpose. The library publishes the monthly periodical Harvard Librarian and supports the Harvard Library Club, which gives the librarians a basis upon which to socialize. The large number of departments do create some problems, particularly with inconsistency. There are several classification systems and cataloging codes in force, and policies on circulation tend to differ from library tto library. However there are strong benefits. In particular for a private institution such as Harvard which does not receive state money, the departmentalization facilitates recognition of special collections and thus of obtaining the all-important grants and other gifts. Also it improves cataloging and collection development.25

Overall, no matter which form of library organization one chooses, there is a balance of costs and benefits. A centralized library may be able to offer more hours and more materials, but it will often be unable to offer the degree of personalized service that a departmental library can. And a decentralized library with a large number of specialized departments can give more personal attention from staff who understand the special problems of the literature of the discipline, but may not be able to offer the hours of service or such fine facilities as a large central library.26 We will simply have to decide which we would rather give our patrons.

Notes

Bibliography

Atkinson, Hugh C. "A Brief for the Other Side" in "Centralization or Decentralization of Library Collections: A Symposium" The Journal of Academic Librarianship 9:196:202 (Sept 1983)

________ . "The Impact of Closing the Catalog on Library Organization," in Closing the Catalog: Proceedings of the 1978 and 1979 Library and Information Techonology Institutes. D. Kaye Gaspen and Bonnie Jurgens, eds. Phonix, AZ: Oryx 1980

________ . "The Impact of New Technology on Library Organization," in The Bowker Annual of Library and Book Trade Information. 19th edition, New York: Bowker 1984

________ . "Optimum Speed of Library Access as Related to Optimum Size of Library Collections," Washington, DC: U.S. Department of Health, Education and Welfare, Office of Education, Bureau of Research 1970

Blanchard, J. R. "Departmental Libraries in Divisional Plan University Libraries: College and Research Libraries 14:243-48

Bruno, J. Michael. "Decentralization in Academic Libraries," Library Trends 19 (January, 1971): 311-317

Bryant, Douglas. "Centralization and Decentralization at Harvard," College and Research Libraries 22: 328-334 1961

Hibbard, Michael. "Centralzied Library Collections? Well, Maybe: A Response," in "Centralization or Decentralization of Library Collections: A Symposium" The Journal of Academic Librarianship 9:196:202 (Sept 1983)

Holley, Edward G. "Reaction to 'A Brief...'" in "Centralization or Decentralization of Library Collections: A Symposium" The Journal of Academic Librarianship 9:196:202 (Sept 1983)

Hurt, Peyton. "Staff Specialization" ALA Bulletin 29: 417-21 (July 1935)

Maizell, Robert E. "The Subject-Departmentalized Public Library," College and Research Libraries 12: 255-60, July, 1951

McAnally, Arthur M. "Departments in University Libraries," Library Trends, 7: 448-64, January 1959

Miller, Robert A. "Centralization versus Decentralization," ALA Bulletin 33: 75-79, February 1939

Phelps, Rose B. "The Effect of Organizational Patterns on the Adequacy and Efficiency of Reference Service in the Large American Public Library" Library Quarterly 17: 281-95 (1947)

Shoham, Snunith, "A Cost-Preference Study of the Decentralization of Academic Library Services," Library Research 4: 175-94 (1982)

Tauber, Maurice F., ed. "Centralization and Decentralization in Academic Libraries: A Symposium" College and Research Libraries 22:327 (Sept 1961)

"The Configuration of Reference in an Electronic Environment (Move from Decentralized to Centralized Library Service at Louisiana State University) College and Research Libraries V. 48 p. 302- 13 July 1987

Thompson, Lawrence S. "The Historical Background of Departmental and Collegiate Libraries" Library Quarterly 12: 49-74, January 1942

Watts, Thomas D. "A Brief for Centralized Libray Collections" in "Centralization or Decentralization of Library Collections: A Symposium" The Journal of Academic Librarianship 9:196:202 (Sept 1983)

Woodard, Beth S. And Gary A. Golden "The Effect of the Online Catalogue on Reference: Uses, Services and Personnel." Information Technology and Libraries 4: 338-45 (Dec 1985)

Wordsworth, Anne. "Decentralization is the Best Principle of Organization Design Where It Fits" in "Centralization or Decentralization of Library Collections: A Symposium" The Journal of Academic Librarianship 9:196:202 (Sept 1983)

Copyright 1990,1998 by Leigh Kimmel

For permission to quote or reprint, contact Leigh Kimmel

This paper was originally written for a class in Advanced Reference, taught by Assistant Professor Bryce Allen, Graduate School of Library and Information Science, University of Illinois at Urbana-Champaign.

Last updated November 5, 2010.