College and Research Libraries JOSEPH C. SHIPMAN "Optimum Size" and the Large Science Research Library Techniques for developing optimal size are more meaningful today than they were twenty years ago. Weeding, selective microfilming, and dependence upon neighboring libraries are still useful, but the general framework has radically changed. Exponential curves of growth in research, and in the literature reflecting this research, as well as ap- pearance of entirely new subject areas and proliferation of older dis- ciplines, have produced a new environment. Perhaps the chief em- phasis should shift from the old concern about optimal size to explora- tion of new devices such as establishment of information centers on the Weinberg model and participation by all special libraries in new networks of scientific communication. wHILE "oPTIMUM SIZE" has long been a concern of special libraries and of large library divisions and departments with special missions, it was probably less meaningful and less critical in the past than it is today, in the face of an ever-mounting flood of scientific and technical publication. National resource libraries, university libraries, and other large research libraries now look ahead ten or twenty years, and predict expendi- tures of tens of millions and hundreds of millions of dollars for collections, for services, for automation, and for staff. The collection and storage of scientific and technical literature obviously is nec- essary, but one collection could not pro- vide adequate service, and duplicate col- lections in every scientific and technical institution obviously are not economically feasible. Since such a procedure is out of the question, each technical library, and all scientific and technical libraries Dr. Shipman is Director of the Linda Hall library in Kansas City, Missouri. This paper was read at the 1965 annual meeting of the American Society for Engineering Education in Chicago. 354/ collectively, face unprecedented prob- lems not only in establishing criteria for optimum size, but also in determining what is of value, and what is econom- ically feasible. What is the scale of the problem? Contrary to the popular impression that the spectacular burgeoning of scientific research and of publication is a post-war phenomenon, Derek J. de Solla Price points out1 that the now-familiar ex- ponential curve of growth had its origins in the seventeenth century, and that, since that time, there has been an annual literature growth of 7 per cent com- pounded. This works out to growth by a factor of 10 each half-century, and by a factor of one million in the three cen- turies which separate us this year from the first issue of the Philosophical Trans- actions of the Royal Society. Since, quite obviously, scientists pro- duce science and its literature, it is im- portant to consider some of the facts about these producers. Most of the sci- entists who have ever lived are active 1 D. J. de Solla Price, "The Scientific Foundations of Science Policy," Nature, CCVI (April 17, 1965), 233-38. "Optimum Size'' and the Large Science Research Library I 355 and working today, and indeed, perhaps 50 per cent of them have been produced in the past ten years. This is not a new feature in the scientific scene but has been true throughout the previous cen- turies of modern science. Newton, Fara- day, Darwin, Rutherford, and Einstein could say that most of the scientists pro- duced up until their own times were to be numbered among their working col- leagues, and that most of them were young. For those of us working with the scientific literature this is significant, be- cause the one distinguishing character- istic of all scientists is that they publish. Publication was and is the lifeblood of science. It has been throughout the past three centuries the prime vehicle of com- munication, of dissemination, and ulti- mately, of storage for all scientific knowl- edge. It was more than that, for as W ein- berg has said, "throughout the years the scientific community · has developed an empirical method for establishing scien- tific priorities-deciding what is impor- tant in science, and what is not im- portant. This is the scientific literature. ... The process of self-criticism embod- ied in the literature, though implicit, is nonetheless real and highly significant."2 Science is a cumulative discipline, and therefore, as Newton observed, in all modesty every creative worker stands on the shoulders of the giants who went before. Today, as Gerald J. Holton of Harvard University said recently, it is more likely that the producing scientist sits in conferences side-by-side with the giants on whose shoulders they stand. Perhaps they are not all giants, but their shoulders still support an immense ef- fort in some of the rapidly growing sub- ject areas. The scientific paper has gen- erally been the accepted vehicle of com- munication in this process of cumula- tion, and as Weinberg has suggested, an effective agent in the selection and crit- ical evaluation of the material to be 2 Alvin M. Weinberg, "Criteria for Scientific Choice,'' Minerva, I (Winter 1963), 159-71. communicated, disseminated, and stored. Both of these historical functions seem to be in serious jeopardy today. In the case -of communication, how can the million or more articles appear- ing every year reach more than a tiny fraction of their potential audience? How many people, except for the author, the referees, and the editor, ever read a sig- nificant number of these papers? As to dissemination-which can be defined in this context as controlled or directed distribution of the literature to a particu- lar audience, prepared and able to un- derstand it and use it-how can such dis- semination possibly be effective in view of the proliferation of the literature, the language barriers, and interdisciplinary complications? While some of the com- munication and dissemination function has shifted to secondary sources-to the abstracts and the current reviews-these secondary publications have themselves become infected with the virus and in many cases are now so voluminous, so frustrating to use, that they too fail-at least in terms of communication. As a re- sult we now see the phenomenon of the mushrooming published symposia, congresses, and conferences, and collo- quia, indicating that scientists may be returning to the old "invisible college" idea, in which personal contact becomes once again the primary device for com- munication within a fairly close-knit group of specialists. The best of these conference and symposium publications -because they have been presented to a knowledgeable group of peers-have the advantage of built-in critical evalua- tion. The ideas presented have been criticized and the criticisms embodied in the published summaries or proceedings. In terms of this sel~ctive and critical evaluation of the literature, the prolifera- tion taking place has generally reduced its reliability and effectiveness. For one thing, the wide variety of sponsorship of current publications-journals and serials published by scientific societies, 3561 College & Research Libraries • September, 1966 or by commercial organizations, or by trade groups, or by government agen- cies, etc., has resulted in a lack of uni- formity or consistency in refereeing, and in critical appraisal. The report litera- ture, now widely dispersed throughout the bibliographic retrieval systems, has sometimes added only bulk-uncritical and nonrefereed papers in great num- bers-to still further complicate the pic- ture. Then, too, there is the Tower of Babel problem. As one nation after another qualifies for membership in the select society of science and technology, new and gener.ally unfamiliar languages ( at least in the West), such as Russian, Japanese, and Chinese, appear ever more frequently and, for example in a field such as chemistry, make up more than 50 per cent of the current publications indexed and abstracted in this subject area. In a special abstracting tool, one issue of which I had occasion to consult recently, contained a total of 373 ref- erences, 148 of them in Russian, 140 in English and 85 references distributed among eleven other languages. In spite of all this, I would agree with Ralph Shaw that "our problem does not appear to be one of too much literature. It would be criminally wasteful if the vastly increased research expenditures of recent years did not result in vast (though not proportional) increases in knowledge, and it would be equally wasteful if the new knowledge thus cre- ated were not made permanently and generally available."3 Our research ex- penditures have doubled on the average every four years since World War II, going from one billion to about eighteen billion dollars in 1965. Nor can we challenge the appearance of the new or even the exotic languages in the sciences. As Dr. Price has em- phasized, the exponential feature of 3 Ralph R. Shaw, " The Function of a Modem Spe- cial Library," R esearch Managem ent, V ( November 1962) , 485-92. growth in the literature involves differ- ing rates of growth for particular sub- jects and particular countries at differ- ent times. There is some evidence to in- dicate that, relatively speaking, Ameri- can, British, and Soviet rates of doubling are slowing down, and that other na- tions are taking up the slack with an in- creased acceleration. The results for the early future are then quite obvious in terms of the resulting language distribu- tion in the literature. One English sci- entist predicted some years ago that by the end of this century more than 50 per cent of the world's scientific litera- ture could be in Chinese. For the large research library there- fore it seems hardly realistic to talk about optimum size techniques without careful consideration of the corrections and adjustments which are called for, and indeed may be already at work within the basic framework of scientific and technical communication. "Optimum size" has of course always been a relative expression, and the de- vices used in its pursuit have recognized differences between libraries in size and in function; differences in mission, as well as the increasing interdisciplinary fluidity which has disrupted many a carefully laid scheme for subject cover- age and control. In the past many devices have been used in determination of optimum size. They have included, among others, the following: 1. The establishment of restrictive limits of subject selection. 2. The use of literature citation counts for determining most-used serial ti- tles, as well as for most-used foreign language titles and for the required and desirable time coverage in the acquisition of backfiles. 3. Dependence upon neighboring li- braries for particular areas of subject strength, and for breadth of coverage. 4. Extensive use of interlibrary loan to tap· more distant library strengths. .. Optimum Size" and the Large Science Research Library I 357 5. Discriminating weeding. 6. Selective microfilming, and the syste- matic purchase of available micro- forms. All of these approaches were and still are practicable and manageable, de- pending upon the particular library's size, function, objectives, and support. For the special library, the problem ap- parently is still much the same as it was. To quote Ralph Shaw again, .. the need in terms of storage and obtaining mate- rials is simply that of housing those things used frequently enough to justify their space . . . and developing a com- munication system which would produce the desired material from any source in which it was housed as promptly as it was needed ... the' job of the special library is to support the research pro- gram for which it is maintained-no more, no less."4 The catch is in the phrase "as prompt- ly as it was .... " Here we s?ift back ~o the large research library, With all of 1ts unsolved problems and complications re- garding .. optimum size." Shaw feels that it should not be difficult to .. develop a network of intelligence services starting with our great research libraries to pro- vide a reservoir of rna terials; a series of information centers, with bibliographical and substantive competence to bring to- gether the significant materials, and eliminate redundancy in broad areas of science and technology .... " Perhaps it is not so difficult .as we think, but at a first look it would seem to require a totally new approach to the literature by ~he scientist himself, rather than a solutiOn by the traditional custodian of the litera- ture-the librarian and the documental- ist. The potential network of great re- search libraries is available, providing the reservoir of materials and the neces- sary bibliographical competence. ~he information centers are already growmg in number and in effectiveness and are • Ibid. , p. 488-89. bringing to bear some of the substantive competence which is the crying need in the current literature situation. It is but a rare beginning, and in spite of the challenges raised by Weinberg, Wige- ner, Bernal, and many others the work- ing scientist generally has not conc~r~~d himself with the primary responsibility which would seem to be his in the total process of evaluation and control of sci- entific communication. In looking for the patterns which are necessary for adequate control and man- agement of this communication, our problem is not primarily one of a n~ed for new machines. We have the machme capability, and it will play a vital role, but it seems to me, an ancillary role. Without going into the specific ques- tions of this role, it seems to me that computers will make it possible to do all the jobs we need to do, but they will not themselves do the job. Weinberg has emphasized the impor- tance of one promising development- the emergence of the information center which far transcends our earlier, more limited conceptions of such centers. The new centers will be manned by scientists and by librarians, all dealing with in- formation problems in the broadest sense. Scientists on . various levels of the Wigener social hierarchy will try to learn about everything that is published in special fields. They will scrutinize, codify, evaluate, compile, review, and synthesize. Their work will result in published abstracts, bibliographies, com- pilations of data, compilation of informa- tion about instruments, equipment, and techniques. Eventually, on other levels, they will extract, relate and generalize, far more systematically than is done to- day. The center will handle the difficult questions dealing with its special area of competence-questions which will be referred to it by national, regional, state, and local referral centers-probably lo- cated in the libraries representing these (Continued on page 392) 3921 College & Research Libraries • September, 1966 Stockpiling of extra copies of noncopy- righted journal articles which have been photocopied once, or more likely more than once, might be studied. This exami- nation should be from the viewpoint of considering storage costs for items which OPTIMUM SIZE .•. ( Continued from page 357) areas. These libraries will depend upon the centers for their published output as an important device for aiding in the evaluation, digestion, and manage- able assimilation of the literature for which they are responsible. In such a development, all libraries could approach the problems of opti- mum size and of division of subject re- sponsibility confidently and rationally. There is considerable evidence that the great research libraries-government, university, and private-are already mak- ing tentative shifts to prepare themselves for their proper roles in the network of collections and services which must eventually evolve if we are to solve our problems. The nationallibr.aries of medi- cine, of agriculture, and other remarka- ble concentrations of subject strengths in various government departmental librar- ies are becoming accepted as true na- tional resource centers, as indeed they have long tacitly been. There is much talk about the establishment of regional branches throughout the country. The Department of Agriculture has had so- called branches for many years, but not on the scale suggested here. The Library of Congress, which established a science and technology division only after World War II, and of course has tremendous resources and capabilities, is a central may never be asked for again, as well as production costs, and relating them to the present practice of photocopying only on demand, albeit if sometimes re- peatedly. •• point in this shift and will spearhead many of the costly e~periments and pro- grams necessary to achieve it. University libraries, for the first time, are separating their research functions in separate buildings, foreshadowing com- ing changes in function. At least one of them, Yale University, is separating its science research materials and will con- centrate them in a p.ew building. The former Midwest Inter-Library Center has changed its name, and, from all indications, its future role in the re- search library picture in the nation. John Crerar and the Linda Hall libraries are becoming increasingly national re- sources, rather than limited local or re- gional library features. In Great Britain, the national lending library of science is a well established, working entity, and plans for a national science reference li- brary are well along. In conclusion, it might be said that «optimum size" techniques may need to be related to a situation in the scien- tific literature and in scientific libraries, which show every indication of radical and extensive changes ahead. These changes will come very quickly, for the pace of achievements in new knowledge and in new literature, which demands these changes, give us very little time. ••