College and Research Libraries l I r I r ROSCOE ROUSE Automation Stops Here: A Case for Man-Made Book Collections The following paper was read at the Second International Seminar on Approval and Gathering Plans for Large and Medium-Size Academic Libraries, Kalamazoo, Michigan, October 31,1969. We print it here be- cause i.ts dissenting viewpoint is as timely as it is provocative. THIS IS INTENDED to be a case study but it may be more than that. A brief affair with an automatic book-buying plan proved a disappointing experience for the Oklahoma State University Library, and it is the purpose of this paper tore- late that experience and to consider the reasons why it was unfortunate. The observations made here have no implications or applications for other li- braries. I speak for one library only. The OSU Library experience was a unique one but not an exclusive one; other li- braries have discontinued approval and blanket plans. A brief description of the book selec- tion policy as practiced before the adop- tion of the plan is necessary for an un- derstanding of the situation. The pro- cedure was a very smooth one, it moved without · friction, it was expeditious, and there was little need for conference or discussion between individuals. Each member of the staff involved had his own specific assignment, and he knew what it was; the faculty knew the indi- viduals responsible for selecting in their respective fields and had confidence in them. Mr. Rouse is University Librarian at Ok- lahoma State University, Stillwater. The OSU Library is organized on the divisional plan, and it was the divisional librarians and their staff members (pub- lic services personnel) who were respon- sible for virtually all book selection be- . fore the plan was adopted. These were the people who worked with the stu- dents and faculty and knew their needs. These were the people who helped un- dergraduates with their reference ques- tions, who aided the faculty in becom- ing familiar with holdings in the respec- tive divisions, who serviced the thesis and dissertation collection, who aided graduate students in gathering materials for their theses, who procured materials for them from distant libraries, who knew the holdings of other libraries well enough to direct interlibrary loan re- quests for a good bull' s-eye percentage. Most of these librarians held a graduate or undergraduate degree in the field in which they were working in the library, and all of those involved in the book se- lection process had long tenure, the av- erage being 14.7 years in the OSU Li- brary at the time the approval plan was instituted. A key member of the staff, al- though promoted in rank and salary at regular intervals along with increased responsibility, has held the same position and title for twenty-three years. I 147 148 I College & Research Libraries • May 1970 The book selection routine at OSU was indeed unique. It was tried and found true. There were no complaints of consequence about the acquisition of books and journals, and our files include some letters complimenting the staff on this aspect of their work. The librarians handled about 80 percent of all book se- lection and the faculty the remaining 20 percent. We sometimes heard faculty comments to the effect that new publi- cations were often ordered before they were aware of the need for them. There was a satisfactory relationship between the librarians and the faculty in the building of the book collection. The basic book selection tool for cur- rent titles was LC proof slips. Upon re- ceipt of the new proof slips, the acquisi- tions librarian sorted them into cate- gories for distribution to the divisional librarians. Other selection media were used, of course, such as Publishers Weekly, Choice, mailed advertisements, dealers' catalogs, reprint catalogs, for- eign listings, and specific standard lists such as Books for College Libraries. Staff members felt a direct and personal responsibility for the quality of their re- spective areas and worked very consci- entiously to build them and round them out well. The approval plan agreement was made with a reputable dealer, and the contract specified that the library would be supplied one copy of every mono- graphic U.S. imprint book within cate- gories stipulated as well as all of the li- brary's standing orders. It was the usual kind of arrangement: excluded were general works, juveniles, introductory textbooks, reprints, fiction, medicine, and religion. The staff held the respon- sibility for selecting . newly published works desired in the fields that were ex- cluded and this they did through proof slips and other sources. Despite the satisfactory situation they were enjoying, the library staff was will- ing to relinquish the selection responsi- bility to an outside party so long as they were assured that the job would be done as well, if not better. At the outset I shall admit to the possibility of unfair- ness in an experience of only four months but also point out the fact that this was one month longer than the agent said was needed to have the plan fully operational and going satisfactori- ly. The relationship was indeed of short duration but it was not entered into as an experiment; the contractual agree- ment was a sincere one made on the basis of expected longevity. Full coop- eration was given to the effort by the li- brarians who had every reason to be- lieve that this was their acquisitions pro- cedure for the future and evermore. In fairness to the dealer, it must be noted that his service to the region was new but nevertheless· we did not con- tract with him on the basis of expecting poor service for this reason. The lack of organization and the obvious use of un- trained personnel indicated that the company was not ready to take on cus- tomers. The failure can thus be tied in to two basic causes: the good climate that had previously prevailed in book selection at OSU and the lack of good organization on the part of the approval plan jobber. The possibility of future improvement of operations by the dealer was an unknown factor; the satisfactory operation of the system formerly em- ployed by the library staff was a known factor. One major complaint against the ap- proval plan was that the library found it was at times returning 50 percent of the titles sent. Many of these were al- ready in the library, doubtless the result of an overlap in the staff selection pro- cedure and the new plan but it did give the impression that the books received from the dealer were not new imprints. A large number of books were sent which did not classify in the categories I l4 I ~ I I I .... [ specified in the agreement. Some titles duplicated others previously shipped by the plan jobber. Still others should not have been sent because they were not monographs but serials. There was also the inclusion of many older titles, a source of real concern to the librarians. These seemed to classify as remainder stock, titles that were in some instances six to eight months old, many shelf-worn and faded. The dealer admitted that he purposely did not order enough books for all his customers, knowing that all li- braries did not want all books. He would wait until some had been returned be- fore shipping them to other libraries wanting them, which may account in part for the age and worn appearance of some volumes. The staff testified to the shipment of every kind of book in or out of desig- nated categories. Received were text- books, juveniles, reprints, and even some foreign titles. And, of course, there were many books received within categories properly chosen in the agreement which did not qualify as titles needed in the OSU Library. Such titles would not have been selected by the staff under the for- mer procedure and these were returned. The overall quality of books received seemed very poor, especially to librari- ans who had previously been quite dis- criminating in the selection of titles. The instructions given for our library simply were not followed. The librarians were more dissatisfied with the books not sent than with those received. A number of good pertinent ti- tles slipped by the dealer for one reason or another and were not supplied to the library; the stqff learned that they could not place complete dependence upon the plan service, and this loss of confi- dence was the beginning of the end. It was known that the jobber did not have good relations with some publishers. Through all this, the staff was never able to tell a faculty member the status of a Automation Stops Here I 149 book at a given time. Whether or not the dealer would ship a particular title was not known for certain, whereas un- der the former procedure one could tell immediately that the book had been or- dered and its exact status in the order routine. As the librarians became aware of the newly published books that were not sent by the dealer, they felt the need to make selections from the proof slips in the same manner as before. This was the only alternative to haphazard, incom- plete collection building. So it was that the staff found itself back at the old task of selecting books as they had pre- viously done. The all-books plan then became redundant. A ream of corre- spondence between the library and the dealer gives evidence to efforts by both to resolve the highly unsatisfactory situ- ation. Visits to the library were made by company representatives. One basic difficulty in receiving books "unsolicited" through a dealer was in re- gard to bibliographic entry. The Head Cataloger at OSU names this problem as the prime one in the failure of the plan. Prior to using the approval plan, 80 percent of our orders had been made from proof slips and for these no verifi- cation was necessary. The books arrived already identified with main entry es- tablished, whereas books arriving from the approval jobber required verification of authors and titles. Books were re- ceived with multiple order forms pre- pared but the entries were so unreliable that the staff had to ignore them. The OSU Library is one of the ninety- seven cooperating PL 480 libraries in the country and therefore receives a deposi- tory LC card for every book cataloged by the Library of Congress Cataloging Division or one of the cooperating li- braries. Once a book is received in the library it is a relatively easy task, if a proof slip is stapled to the order card in the orders-outstanding file, to find and 150 I College & Research Libraries • May 1970 pull the depository card, type the call number, and make a full set for the cat- alog. There is no bibliographical or en- try problem encountered. Books arriving unordered, on the oth- er hand, must be matched with the cards in the depository file. This be- comes almost a professional task unless one has a clerical person who has had good experience with corporate entries and other bibliographical intricacies. Books and depository cards did not, of course, arrive at the same time, the cards almost always arriving much later. The books would therefore wait in the cataloging department until the Library of Congress had prepared and distrib- uted cataloging copy for them. These books had to be temporarily controlled unless they were treated simply as not having been received. Cards had to be checked against many shelves of books, or books had to be checked against many drawers of cards over and over again until the docking in space was complete. In other words, each time a shipment of books was received, the books had to be checked with the de- pository catalog; each time a shipment of LC depository cards was received, these had to be checked with the books awaiting LC copy. A given book might be checked against the depository file a dozen times or more before it was matched with the proper card. In some instances cards never appeared and the searching continued for an extended pe- riod, in which case the cataloger would eventually prepare original copy for the book. In short, the processing staff found itself in trouble from the begin- ning with no let-up seen after four months. The books were not getting on the shelves any faster and additional burdensome tasks were found necessary under the new system. The library was not buying any more titles than before but the processing work was much heav- ier. The division librarians, who hold the greatest responsibility for the selection of titles, maintained that the greatest shortcoming of the plan was the narrow bibliographic base upon which the agent operated. The public services librarians found it necessary to search the proof slips anyway, because so many good works were overlooked by the dealer. The agent's staff (or computer) did not send everything in the fields shown in our profile. An example given was con- cerned with the laser beam. We were in need of everything, literally, published on the subject as a physics graduate stu- dent at our institution designed the in- strument that sent the laser beam to the moon and back last July. The library found that it could depend upon r eceiv- ing through the plan only a small part of the material needed because the jobber did not furnish materials from a number of U.S. publishers or from numerous so- cieties, institutions, and associations which issue scholarly publications. The same librarian who gave the laser beam example said he found it much faster and more satisfactory to choose books from LC proof slips than by using the books themselves. He felt that there was too much time involved in reading tables of contents, prefaces and such, whereas the LC card with its call num- ber, subject headings, and full title gave all that was needed to make a decision, in most cases, especially in the sciences. The conclusion was reached that only the OSU faculty and library staff knew best which editions the library should have, which publishers were best for specific titles, which editors it preferred. Oklahoma State librarians were better and more currently informed about their degree programs, departmental projects and studies, thesis topics, and specialties of the faculty. There is no time lag in al- tering the profile when the job is done by a well-advised librarian right at the source of information. 1 1 Disenchanted with the whole idea, the OSU Library staff, almost to a man, was pleading to return to the former method of selecting and ordering books. The approval plan was cancelled and the proof slip routines reinstated. When this was accomplished, the staff found that they were still trying to extricate the library from the red tape of the plan a year and a half later. Today the book selection procedure is as good as it ever was. The librarians and the faculty are content with this routine, and they are of the opinion that the development of the collection in the various fields and disciplines is as good as can be expected for a library with a limited budget. The writer recently received a long- distance call from a million-volume li- brary in the far West. The caller said he needed advice about his acquisitions program. The library had been on an all-books plan the year before but the supplier did not furnish materials as promised. The librarian was unable to answer the faculty's questions as to whether certain titles were coming, and a large percentage of new titles never reached the library. There was much confusion regarding the serials that the dealer should supply and those that would come through the library's own standing orders; it was obvious that the supplier could not differentiate between serials and monographs. The librarian said too much time was wasted in re- viewing the books that were received. It was therefore decided to discontinue the plan, and the staff returned to their for- mer method of selection with faculty consultation. The librarian said they soon found themselves in trouble again and were desperately seeking a solution. Further inquiry brought out evidence that the present dilemma stems from a staff shortage. In this case the all-books plan had been turned to as a panacea, which it was not, and a return to the old manual system was a nightmare of an- Automation Stops Here I 151 other kind-for reasons that were easily identified. The Council on Library Resources is supporting a two-year study of an ex- perimental model engineering library at MIT incorporating "new technological developments." The physical remodeling alone for the project, which will be car- ried out under INTREX, has cost $2 million, and the council made grants of more than a million dollars to MIT to support INTREX. The library will in- corporate such software as the text ac- cess system and the augmented catalog, which is a computer-based bibliographic mechanism utilizing the cathode ray tube to rapidly and interactively search a remotely stored catalog in which each document is cataloged in great depth. The text access system can be used to retrieve those documents from a remote- ly stored microform file. Programmed teaching machines will also be a part of the library system. We can expect to see a computer age library and retrieval sys- tem emerge from this kind of invest- ment and experimentation. I was interested to know if this, the nation's most forward-looking library (the adjectives are mine), acquires its books through a blanket or approval plan. In my communications with them I almost felt like apologizing for even suggesting a manual procedure in their operations. In correspondence and by telephone conversation with the librari- an I received a response which was in good humor but very positively and em- phatically stated: the library now em- ploys and expects to continue to employ the manual, individual, and personal form of book selection, all done by mem- bers of the library staff and faculty. No approval or blanket plan is foreseen in their library. About 95 percent of the ti- tles are selected by librarians and the remainder by the faculty. This highly mechanized library is quite satisfied with this arrangement, and there seem 152 I College & Research Libraries • May 1970 to be no plans to change it. The Stanford University library uti- lizes about seventeen various blanket and approval plans. Still the library em- ploys a number of librarians who are book selection specialists, a staff which makes up, to use their wording, ua net- work of acquisitional interests." Their specialists use Publishers Weekly and proof slips for selection purposes. How else would a great library system have full coverage from societies and associa- tions, private presses, little-known pub- lishers, U.S. and foreign governmental agencies, vanity presses, the U.N., pub- lications from underdeveloped countries and near-print materials? Such a vast se- lection and acquisitions program could not today be successfully handled in toto by a commercial firm, even seventeen of them. David 0. Lane, in preparing his pa- per "Approval and Blanket-Order Ac- quisitions Plans," queried sixty-six me- dium-size academic libraries and re- ceived forty-six replies. 1 Thirty-one of those replying used approval plans (three of these were dissatisfied, two undecided). Thirty-eight of those reply- ing used blanket -order plans (five of these were dissatisfied, four undecided). Some of the reasons given for the dis- satisfaction by those who expressed it were as follows: serials present a prob- lem, duplicates were received, too much junk was received, too limited, takes too much time, pertinent books not re- ceived, late receipt, guidelines not fol- lowed, and problems in billing and in- voicing. In regard to Lane's inquiry concern- ing the satisfaction of the faculty with the plans, only thirty-nine librarians out of the forty-six who responded gave an answer and twenty-eight of these re- plied in the affirmative. It is interesting to note that his research showed that the median percentage of current im- prints added by the operation of the ap- proval/blanket order plans in these li- braries was 28 percent. The largest num- ber of those on the plans indicated their interest in retaining them and most ex- pected to expand to other plans. · Three libraries expressed their intention to do away with blanket orders. In the Summer 1969 issue 'of Library Resources & Technical Services, Ian Thorn wrote of the added work involved with blanket and approval plans. 2 "This method of procurement," he wrote, "oth- er things being equal, does not result in 'less work' for the acquisitions people. On balance, the acquisitions department will require more man-hours to process a given number of titles received on blanket order than it would if these same titles were ordered conventional- ly." He modified his use of the word "acquisition" to exclude the selection of books, meaning procurement only. He further says, "While it eliminates some operations, however, blanket ordering creates others." Margit Kraft in her paper, "An Argu- ment for Selectivity in the Acquisition of Materials for Research Libraries," makes the point that the machine will undoubtedly handle quantity for us bet- ter but that it does not differentiate be- tween quantity and quality, noting that this requires human intellect.3 She points out the fact that one feels like a heretic even to question the arguments put forth by virtually all U.S. academic libraries for the building of giant book collections and goes on to say that the urge to preserve an object assumes it has value. Her paper is a sound and sol- id treatise which will give most of us pause regarding our use of all-books- current plans as we acquire, process, and preserve at great cost the good, bad, and indifferent. Gordon Williams published a study in 1966 in which he refers to a source which asserts that the technology library at Northwestern University could be re- j ~ I 1 I I J I ~ J I ~ ~ I I duced by 75 percent and still satisfy 99 percent of its present users, and the gen- eral library could be reduced by 60 per- cent and satisfy 99 percent of its users.4 At the Symposium on Approval Order Plans sponsored by the Pacific North- west Library Association in 1967, the fact was brought out that United States libraries acquired twice as many books in 1966 as they did in 1960. The rate of increase between 1965 and 1966 was 29 percent. About fifty new institutions are established each .year and by 1975 aca- demic libraries will be spending $300 million a year for materials. Perry D. Morrison said at this conference, "We hope that the computer's tail will not wag the intellectual dog."5 He points out three advantages to using an all-books plan and seven disadvantages, includ- ing the fact that the automatic plan builds an uncritical collection, and he re- marks that one becomes too dependent upon a single supplier and "subject to the tyranny of his computer." The writ- er, a faculty member himself, said he did not feel that his interests were being served if it were all to become automatic and superficial. ... At the same conference, LeRoy C. Merritt presented a paper entitled "Are We Selecting or Collecting?" in which he said, "My contention is that the qual- ity of the collection produced, not the promised increase in efficiency or "order- ing procedures, is the true issue."6 When the Michigan State University library left the divisional plan of opera- tion, Dr. Richard Chapin lamented the loss of the advantages that plan offered in the development of the book collec- tion.7 He said he found it necessary to redefine their efforts for resource devel- opment, and specific discipline assign- ments were passed out to members of the staff. A book selection department was created in the library. A recent issue of College & Research Libraries includes a paper titled "Book Automation Stops Here I 153 Selection in Academic Libraries: A New Approach," by J. G. Schad and Ruth L. Adams. 8 These writers advocate a work- ing combination of faculty and librari- ans to build the most satisfactory and relevant book collection. There is· no ref- erence whatever to any kind of approv- al or blanket or all-books plan. A complex operation can be auto- mated if it is consistent and standard- ized. Book selection is neither. The in- formation How emanating from large numbers of books as they are issued forth is made up of many unique and varying parts with shades of difference that may be extremely important to a particular library situation. A machine cannot deal properly with this kind of~ fluctuating subject matter, with linguis- tic and semantic materials; at least the machines in use today cannot. A book is the result of the thinking process of a man, and a machine will treat words just as though they were static, inflexible, sterile, categorical bits; the human mind extracts much more than this from the printed page. It is almost as though one were attempting to put a man's thinking process into a machine for recall as needed. We may try, but I hardly think we can be successful, really successful, in building a good, really good, library collection by automatic book selection alone. We can, I think, do this very well with a combination of machine and hu- man intelligence. The OSU Library approval plan ex- perience was an unfortunate one, but it did occur at one point in time under a specific set of circumstances. Today it might well be that those circumstances do not exist and the same set of pro b- lems would not arise. The OSU Library may very well one day come back into the fold and employ some kind of gath- ering plan, but if we do, I think we shall still use the human touch to tailor a book collection to fit our own particular needs. 154 I College & Research Libraries • May 1970 REFERENCES 1. David 0. Lane, "Approval and Blanket Order Acquisitions Plan" (A paper read before the Institute on Acquisitions Pro- cedures in Academic Libraries, Univ. of California, San Diego, Aug. 25- Sept. 5, 1969). 2. Ian W. Thorn, "Some Administrative Aspects of Blanket Ordering," Library Resources & Technical Services 13:338- 42 (Summer 1969) . 3. Margit Kraft, "An Argument for Selec- tivity in the Acquisition of Materials for Research Libraries," Library Quarterly 37:284-95 (July 1967). 4. Gordon Williams, "Academic Librarian- ship: The State of the Art," Library ]ournal91:2417 (16 May 1966). 5. Perry D. Morrison, "A Symposium on Approval Order Plans and the Book Se- lection Responsibilities of Librarians," Library Resources & Technical Services 12:133-39 (Spring 1968). 6. LeRoy C. Merritt, "Are We Selecting or Collecting?" Library Resources & Tech- nical Services 12:140-42 (Spring 1968). 7. Richard E. Chapin and Ralph E. Mc- Coy, "The Emerging Institutions: Mich- igan State University and Southern Il- linois University," Library Trends 15: 266-85 (Oct. 1966) . 8. Jasper G. Schad and Ruth L. Adams, "Book Selection in Academic Libraries: A New Approach," CRL 30:437-42 (Sept. 1969).