College and Research Libraries DONALD A. REDMOND, MICHAEL P. SINCLAIR, and ELINORE BROWN University Libraries and University Research Since a university library is a service agency,. it is unable itself to "ra- tionalize" the potentially limitless demands made upon it by the sum of requests from individual researchers. Theoretically its budget de- mands on the institution could potentially be limitless. The library is involved in two "information cycles": a "publication cycle" of pro- duction of !!ew knowledge, its formalization and its storage and use; and a "demand cycle." Both of these cycles involve institutional cost. Rationalization decisions can be made at a program level, when the institution decides whether to enter a subject area; or at a recruitment level, when the institution and a candidate for a staff position decide whether there is a "match'' of interests. Upon recruitment, the institu- tion has a responsibility to support the staff member's research in- terests. OUR CONCERN IS THE RELATIONSHIP be- tween the development of research re- sources in university libraries and the initiative of the researcher toward proj- ects. Arising from this is the cost, to both individual institutions and to re- gional systems of libraries or institu- tions, and ultimately to the (govern- mental) funding agency. Our concern is the conflict which occurs, so often as to be the rule, between the location of This paper was originally presented as a brief to the Commission to Study Ra- tionalization of University Research, of the Association of Universities and Colleges of Canada, at a hearing December 10, 1971. The authors gratefully acknowledge the valuable assistance of Ms. Lin Good, Ru- dolf C. Ellsworth, William F. E. Morley, and Theodore D. Phillips. All are members of the staff of Douglas Library, Queen's University at Kingston, Ontario. The opin- ions expressed are those of the respondents and not those of the university. research and the location of resources: the one ever varying, the other often underused. Our contention is that unless rationalization decisions are applied in two ways, university libraries are faced . with a limitless expansion and needless duplication of efforts to build resources. Those two ways are: ( 1) rationalization decisions at the institutional level to de- limit areas of interest and acceptance by the individual of limits to the re- sources which can be provided for him; and ( 2) the cooperative interinstitution- al approach to information resources, which will make possible rationalization decisions about the relocation of library resources. Library resources are gathered, in the first place, in response to current teach- ing and research demands. They neces- sarily represent commitments to re- search. The cost of acquiring and or- ganizing a collection sufficient for even the smallest research project is very large, in time as well as money. A col- / 447 448 I College & Research Libraries • November 1972 lection once gathered, though it be un- used because of changing research in- terests, represents a continuing charge to the institution for storage and ser-. ) VICe. Instances are rare in which an individ- ual or a research group is drawn to an institution because its library research resources are attractive. So few institu- tions are fortunate enough to own col- lections attractive to researchers in many disciplines that the phrase "Harvard ef- fect" is obvious in meaning. By-and- large, established collections are not the impetus to research. Rather, the re- searcher generates a demand post facto upon the library. The alternative pos- sibility-that the individual's research will be limited to available materials- is much less 1ikely. The unpredictable, "mosaic" nature of the choice of individual research topics means that many very small areas may be intensively studied, and re- sources on those topics in demand for a time; while other areas are u~touched, and sparse or fallow in the library. The informational material used for re- search is typically so specialized that the statistically predictable frequency of use, well documented by Ash, Trueswell, and many others, is very low-on the order of once in several years. 1 • 2 Wheth- er large or small collections of research material are gathered they will both have a low probability of continuing use. Clearly, for economy of research re- sources, the probability of use should be increased-by making the resources available to the largest community of researchers; by increasing the continuity of research interest; or by rationalization decisions which channel the develop- ment of these resources in the first in- stance. THE RESEARCH INFORMATION CYCLE Figure 1 illustrates a well-known phe- nomenon which we shall here call the research information cycle. The stages by which research conclusions are dis- seminated proceed by increasing for- malization: from the stage of personal communication to a known colleague, through oral group communication ( lec- tures to meetings), to what the scientist calls the "pfimary literature." This is in turn extracted and compiled, becomes standard fact, and is superseded (be- comes obsolescent) by new research. 3 • 4 The library is the formal repository for information in the generic sense. If the information is numeric or un- published or ephemeral, the repository may be called a documentation or in- formation analysis center, or data bank, but the substitution or the interpolation of an information analyst-a surrogate researcher-does not affect the form of the cycle. A library (or its kin) is cen- tral to the cycle. This is an open cycle. By the time in- formation has gone through the succes- sive stages of publication, the researcher himself is no longer interested in it. In- formation is of the greatest use to him when provided through the shortest path. No wonder that the largest reli- ance is on the "invisible college" -per- sonal contact accounting for the largest fraction of information supply. Because this is an open cycle, one must ask whether rationalization of re- search is a means of closing or recurving the cycle to reenter a prior stage. Ra- tionalization is shown in Figure 1 as im- pinging upon the researcher. Proceeding around the cycle through the stages of formal publication, rationalization could take the form of controlling pub- lication or distribution; of evaluating literature resources, both those in the li- brary and those not available; and of choosing areas to be included or exclud- ed when setting the scope of research interests. The library itself cannot make those rationalization decisions. The demand upon it is potentially limitless. By defi- nition it is a service; only if the demand • University Libraries I 449 Results- Topic and interests ( background provides Informal Communication / ""' ~n ~ (writing) Primary Publication (Journals, Patents, Reports) extraction RESEARCHER information LIBRARY \ \ (Personal or ~ Institutional) ~ Secondary Publication " RATIONALIZATION? "' Obsolescence __. Tertiary Publication (Textbooks, Treatises) (Abstracts, Indexes, Monographs ) Figure 1 The Research Information Cycle: Generation and Recycling of Information (the Publication Cycle) is given limits can the library attempt to satisfy all users. The relationship between researcher and library is further shown in Figure 2. Th~ horizontal diameter (provision of information from library to re- searcher) has already appeared in Fig- ure 1, the information cycle. In Figure 2, the generation of information de- mand by the interests of the individual can be seen to result in increased com- petition for finance, and in a budget de- mand by the library upon the institu- tion's administration. 5 The rationaliza- tion decision which will limit these de- mand flows must therefore be made by the individual, before he decides upon his research, or by the institution, either before or after the individual has com- mitted his personal resources. To reiterate, the library cannot make these rationalization decisions. As a ser- vice agency, it can only satisfy demands, or explain why it fails to satisfy them. The explanations are presumptive, and ·the researcher should accept them and temper his demands accordingly, or seek elsewhere to influence budget decisions. TEACHING, RESEARCH, AND MOBILITY We accept the reality that research and teaching demands upon the library . 450 I College & Research Libraries • Nove1nber 1972 Topic of Specialization 0""' / .. untapped Research Information Administrative Unit (Department, Institute) internal rationalization resources Institutional Administration -LIBRARY Figure 2 The Research Information Cycle and the University Library: the Demand Cycle are so closely linked that one is unable to distinguish between acquisitions made for the one purpose and for the other. A preliminary study at our insti- tution (not yet available for release) tends to confirm this; the opinion of teaching staff is that their research makes a relatively small incremental de- mand upon the library, over that of their (graduate) teaching. Because an individual moves not only among several institutions in a typical career, but more importantly from proj- ect to project, the probable consequent use of particular specialized library re- sources may wane in a year, or in a few years. He may commit the institution di- rectly to large library expenditures, and indirectly to equally high salary and ser- vice costs, while engaged in closely-re- lated teaching and research, and depart, metaphorically or bodily, for other pas- tures after a few years. This cost can be looked upon in two ways: as a necessary capital expenditure associated with the decision to engage in some new research program, and thereafter to be treated as a sunk cost; or as a capital investment (in the indi- vidual) for which a return must be computed. The return to the institution and to society upon such investment is known to be small. We may hypothesize that in many instances the investment is never fully repaid. The larger part of the return is to the researcher in the form of prestige and salary increases. We prefer to treat expenditure on li- brary resources as investments rather than sunk costs, because they involve continuing service (housekeeping) costs to the institution. We ask then how the return on such investments can be in- creased. Clearly, by increasing continu- ity of interest in them, increasing the number of potential users, or by locat- ing the resources where either of the first two factors can operate. We see no reason not to treat the in- formation cycle as operative at regional or national levels. The implication is that the researcher and the institution must consider the location of library re- sources at these levels before making a decision to engage in particular research. Institutional administrations should en- courage researchers to undertake their work in the places where resources exist, possibly unused. Alternatively, we rec- ommend that, so far as possible within the framework of institutional and pro- vincial rights, library resource strengths be relocated to suit the needs of re- search, rather than unnecessarily dupli- cated. There are costs and technical problems associated with such activity, but we believe them to be less than the cost of unrationalized and uncoordinat- ed activity. DECISION-MAKING IN THE INSTITUTION Two types of impetus for research can be identified: individual initiative, and the research-oriented group. The group may be based on a teaching pro- gram, a primary research purpose, or an administrative unit such as a teaching department or school. The group may have a direct impact on the library if their decision is to University Libraries I 451 mount a program which will require li- brary resources and services. The impact rna y be indirect if the decisions made delimit individual interests, involve re- cruitment or the selection of the re- searchers themselves. How are these ad- ministrative decisions and interests gen- erated? How are they sanctioned by the institution, and at what level? Frequent- ly they are "sparked" by a strong indi- vidual, around whom a group clusters, from which a formal program proposal may eventually evolve. It has been our experience that a cur- ricular decision made within a teaching department is often determined without adequate consideration both of avail- able library resources, and of cost and lead time required to develop needed re- sources. Program decisions, at least at the graduate level in Ontario, now in- volve formal communication with the library, and a lead time of hearly two years for the whole process. This is a de- cided help. The pure research decision at the group level may not consider library re- sources in advance at all; yet it is likely to have the largest cost because the Ill- formational resources needed are pri- mary, sometimes unique, costly, exten- sive, and always highly intensive. The well-known Bradford-Zip£ Law of Dis- tribution of Informational Resources in a subject demonstrates that to be ex- haustive in even a narrow topic requires a very large coverage of the literature. Has the individual a "right" to do re- search? We would point out that mem- bers of an institution agree to operate within preestablished regulations; they agree at least to certain limits to other "rights." These limitations in no way affect academic freedom to inquire as the individual wishes. Rather, there are economic realities which may limit the depth to which a research project may be pursued, indeed sometimes preclud- ing it altogether. As part of the recruiting process by 452 I College & Research Libraries • November 1972 which an individual comes to an institu- tion, the candidate has an obligation to explore its existing resources, as well as to make his interests known. The ad- ministrator who does the recruiting has an obligation to test the "fit" or "match" between the institution's needs and the individual's potential-and also between the individual's research inter- est and the institution's potential. The crucial question should be: what obliga- tions, immediate and continuing, must the institution take on; what must be promised him, if he comes here? The individual should be introduced to un- explored resources in his general area, with the possibility that his research in- terests may be enlivened. The library often assists a graduate student in find- ing a thesis topic; how often has this oc- curred in the case of a staff researcher? After the "match" is accepted, the in- stitution has at least a limited obligation to provide resources for the individual's research. The department (or other unit) must ensure that this provision is made-not by coercion of, but by co- operating with the library to explain its needs to the administration. If promises are . made they must be kept, but if the library has not made the promise, it should be under no stringent obligation to shift its budget and make such pro- vision. For his part, the individual should consider whether his "right" to do re- search yields precedence to the obliga- tions of the program · in which he par- ticipates, or the limits of scope of the institution he has agreed to join. When a program-level decision for research has been made, the institution (since it has a larger power of review over the decision) has the larger responsibility to support it. Vertical divisions between parts of the university-departments, institutes, faculties, schools-cause a lack of in- ternal rationalization. Such hiatus of structure means that the university tends to add programs, projects, courses, or research interests without its own or- ganic wholeness in clear view. Individu- al units support this autonomy of aims, supporting each others' rights to deter- mination, rather than the integrity of. the whole. This is a core problem of rationalization: it must be internal as well as external; within the institution, as well as among institutions. Rational- ization must also exist interinstitution- ally within any given discipline. SUMMARY AND RECOMMENDATIONS We would summarize points which we have tried to make evident: 1. Failing rationalization, university li- braries face limitless demands. 2. Libraries cannot make rationaliza- tion decisions since they are service agencies. 3. The researcher generates a demand, but there is no overall pattern to the sum of demands. 4. Common to the cycle of the genera- tion and dissemination of informa- tion, and the cycle of research de- mands within an institution, is the library-researcher axis-that of the provision of information. 5. Individuals are mobile both physi- cally and in research interests. 6. Library resources should be consid- ered investments, rather than sunk costs, and an effort made to increase the return by encouraging use. 7. Return on library resources can be increased by: (a) increasing continuity of interest (probably at the program level); (b) increasing the number of poten- tial users, by regional availa- bility; ( c) locating the resources for best use, by transfer if need be; (d) encouraging researchers, espe- cially recruits, to explore avail- able but underused resources. 8. Rationalization decisions are neces- sary within an institution as well as among institutions; among the vari- ous groups in a university, as well as among like groups in various uni- versities. 9. Rationalization decisions are pos- sible particularly: ( a) in the recruitment process for researchers; (b) at the program level. 10. Program and research decisions in- volve a library factor, with cost and lead-time components, for which a positive mechanism should be in- cluded in the decision procedure. University Libraries I 453 11. The individual's "right" to do re- search may be constrained by cost problems. 12. The faculty recruit and the admin- istrator have an obligation to test the match of interests and potential, and of available resources and budget; and having accepted the match, reconcile themselves to such limits. 13. The institution, having accepted an individual, has within cost limits an obligation to support his interests, including research. REFERENCES 1. Lee Ash, Yale's Selective Book Retirement Program. Hamden, Conn.: Archon Books, 1963. 2. Richard W. Trueswell, "Two Characteristics of Circulation and Their Effect on the Im- plementation of Mechanized Circulation Con- trol Systems." CRL 25, no. 4 (July 1964), p. 285- 91. 3. S. C. Bradford, Documentation. London: Crosby Lockwood, 1948. 4. G. K. Zipf, Human Behavior and the Principle of Least Effort . Cambridge, Mass.: Addison- Wesley, 1949.