College and Research Libraries Editor's Note: Over eighty letters were re- ceived in response to the March editorial entitled ··can Academic Librarians Afford College & Research Libraries?" The senti- ments of those who responded were ones of disbelief, shock, and anger. The respon- dents urged that the AC RL publications program be left intact. As several pointed out, the publications remain one of the few tangible benefits of ACRL to many mem- bers. Because of the present restriction on journal pages, w e are publishing only a se- lection of the letters. The sentiments ex- pressed in them are typical of those ex- pressed by others. It was indeed gratifying that so many people took the time and trou- ble to respond. To the Editor: Just a note in answer to your editorial in the May issue of CRL. May I state, honest- ly and frankly, that CRL is the only jour- nal in the library field that has almost con- sistently been worth not only reading, but keeping. I frankly discard American Li- braries, Library Journal, and WLJ after a year, but CRL I keep for five. Were CRL no longer available as part of my ALA dues (maybe I should deduct $5.00 and send it directly to CRL), I doubt if what ALA gives me would any longer be worth it. Robert S. Taylor Director of the Library Center Hampshire College Amherst, Massachusetts To the Editor: Please consider this letter in strong sup- port of your editorial "Can Academic Li- brarians Afford College & Research Li- braries?" My membership in ALA and par- ticularly in the section of ACRL is en- hanced exceedingly by this publication. I consider it the most valuable professional publication which I receive. I sympathize with ALA's effort to reduce Letters operating costs and keep membership dues from escalating. However, this is not the di- rection to go in cutting expenses. Personal- ly, I would question joining ALA those years that I am unable to attend the con- vention if CRL ceases publication. This is a strong statement and it is not meant to mean that other activities of ALA are unim- portant. It means that academic librarians value CRL. I would be extremely unhappy if it ceased publication. To the Editor: Mrs. Ronald C. Turner Acting Head Librarian Whitworth College Spokane, Washington As a recent member of the library pro- fession, I am disturbed to hear that ALA is considering terminating CRL and CRL News-two journals I consider very worth- while--in favor of incorporating them into who knows what. I had to save my pennies to join ALA this year, and I did so only so I could sup- port such publications as yours. If all I re- ceive for my money is some chaotic publi- cation such as American Libraries, I will not be a member next year. I ohn Cosgriff Chemistry-Biology Librarian California Institute of Technology Pasadena To the Editor: I was appalled to read that the ALA is considering the abolition of CRL. The only consolation I have found in paying the ab- surdly high ALA dues has been the exis- tence of CRL. It is by far and away the most important publication for the academ- ic librarian. The present "budget stringencies" of ALA are clearly self-generated. It is prob- ably true that all nonprofit organizations of any size and age (including libraries) tend to be run to serve the interests of manage- /381 382 I College & Research Libraries • September 1971 ment. This is true to the point of being bi- zarre in the ALA. There always seems to be money for an increasingly costly secre- tariat, but never enough for programs of demonstrated utility to the membership. It reminds one of the many poverty programs which have greatly enriched the social sci- ence apparatus but have done little indeed for the poor. I suspect that the vast majority of aca- demic librarians feel as I do and will join in protesting this latest assault on the mem- bership. To the Editor: Robert F. Munn Director of Libraries West Virginia University Morgantown I am astonished and horrified by the COPES and ALA Publishing Board even contemplating the abolition of divisional newsletters. Better to abolish American Li- braries! The primary reason I belong to ALA is to receive CRL and RQ. A drain on the ALA-what utter nonsense. I think you make it clear that we are supporting the ALA and not vice versa and I hope that many other angry librarians write you to offer their support and to make it clear to the ALA that it will receive support from its membership only so long as it continues to serve its divisions because it is the divi- sions which provide the membership with the most directly relevant and useful ser- vices and publications. The ALA is obvious- ly suffering from a rash of bureaucracy and the appropriate remedial action should be taken. To the Editor: (Mrs.) Elizabeth Silvester Head, Reference Department McGill University Montreal, Quebec, Canada Every year when I get my renewal no- tice from the American Library Association, I seriously debate the desirability of paying my dues for the tripe I receive in my issues of American Libraries. Upon reflecting for a few moments, I have always decided that I must pay up again because I want to sup- port the College and Research Library sec- tion. As far as I am concerned, CRL is the only publication from ALA with very much relevance to college librarianship. I would be in favor of college, university, and research library people organizing an association outside the scope of ALA rather than submit to assimilation of our journal into some other publication. I certainly would not hesitate to drop my membership in ALA should we lose our only effective voice. Shannon]. Henderson Associate Librarian Arkansas Polytechnic College Russellville To the Editor: I must admit I'm always surprised at my own naivete about the way in which organi- zations are run. Your editorial in the March CRL astounds me. I have been told many times that the publications of the American Library Association are very expensive to support and are, probably, the major item which continually increases our dues. How- ever, I have always assumed that this meant the publications are supported in ad- dition to the dollar amount indicated on the membership form each year. If you aren't getting my $5.00 for CRL, why the hell not? I am, in general, dissatisfied enough with the Association to consider getting out. Cancellation of CRL and other Association publications would probably be the straw that breaks this camel's back. If it does hap- pen, I suggest that you use your present po- sition to try to establish CRL as an inde- pendent publication either self-supporting or supported with a base of funds from re- search libraries and then maintain it as self- supporting after it is underway. Let me know what, if anything, I can do. To the Editor: W. David Laird, ]r. Associate Director for Technical Services University of Utah Salt Lake City Your recent article in CRL astounded us at Knox College. It makes one lose faith in our professional organization, ALA, to learn that the money supposedly allocated to our journal is not really going there. The most important benefit which I re- ceive from my ALA membership at the present time is the divisional publications. I would seriously consider dropping my membership if these journals were consoli- dated, especially if they follow the format of American Libraries. I have ceased read- ing that journal; it reminds me of Colliers when it "updated" its format shortly before it became defunct. I sincerely hope we can continue publishing CRL. To the Editor: Louise A. ] encks Cataloger Knox College Library Galesburg, Illinois Manifold compliments on your editorial in the March issue of CRL! It is beautifully written and the point well made. As a member of ACRL, I am responding to your invitation for comment. It would be doing the library profession a great dis- service to do away with CRL. Those of us in academic libraries rely on CRL, LRTS, and ]OLA for substantive information con- cerning current issues, projects, and re- . search in our field. American Libraries, while interesting and newsy, is far from being the important resource tool that the divisional journals are. Is it possible, if your editorial arouses enough response, that ALA might restruc- ture the budgeting of publications in order to allow self-support? If this does not hap- pen and divisional publications are abol- ished, doesn't it seem reasonable to you that dues should be reduced by five dol- lars? To the Editor: Susan K. Martin Systems Librarian Harvard College Library Cambridge, Massachusetts Keep CRL and CRL News. As a librari- an, I rely on them heavily to keep abreast of contemporary developments in academic librarianship. Your editorial implies that ACRL is sub- sidizing ALA with $47,000. I'm in favor of ACRL having the total $60,000. That is what I understood was happening to the $5.00 designated in my membership for ACRL. Letters /383 Consolidation of journals can be cost-ef- fective, but not less costly, if the proper coverage is continued. Otherwise, someone will be neglected. Bernard C. Rink Librarian Northwestern Michigan College Traverse City To the Editor: In reply to your editorial in the March is- sue of CRL I would like to express my hope that CRL continue as a divisional journal. As director of a college library I .find it ex- tremely worthwhile, which is more than I can honestly say for American Libraries. What sort of . interdivisional publication could so well meet our needs-a combina- tion of LRTS and CRL? I'd prefer to get only CRL and not American Libraries if I had the choice. To the Editor: Ann M. Carper Director of the Library Elizabethtown College Elizabethtown, Pennsylvania About a year ago I wrote someone in re- ply to a request for an expression of opin- ion to the effect that I felt that ALA had long since failed the academic library and that I would like to see ACRL become a separate organization with its own dues and its own convention. I have been a member of ALA since 1937 but your editorial con- vinces me that I should find some other or- ganization until I can be a member of ACRL without paying these exorbitant dues to ALA. I note that ACRL provides about one- third of ALA's membership and I would guess a probably greater percentage of its dues and, in my judgment anyway, gets lit- tle in return. To the Editor: 0. M. Hovde Librarian Luther College Decorah, Iowa This letter is to express my undivided support for the continuation of CRL and CRL News. It is absurd that COPES and the Pub- 384 I College & Research Libraries • September 1971 lishing Board should consider the abolition of divisional publications when they are money makers and also do a tremendous job serving the academic interests. CRL is a fine publication that has reached new peaks under your able leader- ship. You should be commended for a job consistently well done. To the Editor: Peter Spyers-Duran Director of Libraries Florida Atlantic University Boca Raton Your editorial in the March issue of CRL emphasized the unfairness of the ALA pub- lishing board in connection with the pub- lishing of CRL. I feel that CRL and CRL News both should continue to be published. Academic librarians need this forum, especially in view of the fact that most of the ALA pub- lications do not worry about our profession- al problems and at the ALA meetings aca- demic libraries and librarians are rarely mentioned. I hope that sooner or later the unhealthy situation [concerning reorganization plans of the Planning Committee and the Board of Directors] that has developed will be changed, and that academic librarians will have an autonomous or independent or- ganization which will properly represent them. To the Editor: Dr. S. Szilassy Director of the Library University of Tampa Tampa, Florida Your editorial in the March 1971 CRL about the possible demise of CRL is shock- ing! Of all library literature it is the one journal I refer to more than any other when I am looking for facts or experience to back up administrative decisions. It is an indis- pensable vehicle for communication among academic libraries, and I for one protest loudly even at the thought of its not being continued by ALA. · No better illustration of the value of CRL is the latest (March 1971) issue. Four of the five articles are so cogent to the problems that are on my desk right now. I hope, for goodness sakes (and for the sake of academic librarianship) , that you get an overwhehning response to your edi- torial. Whereas I have been somewhat neu- tral on the prospect of ACRL going its own independent way, the proposal to stop pub- lishing C RL would place me solidly in the camp of independence. We simply must keep our journal going! Ralph H. Hopp University Librarian University of Minnesota Wilson Library, Minneapolis HAYES AND MASON ON AUTOMATION ... August 29 , 1967 Dr. Robert M. Hayes, Director Institute of Library Research University of California Los Angeles, California Dear Bob: I am much interested in your criteria of evaluating university collections and your numerical application of it to our campus statistics. There are some things about the criteria that I do not completely under- stand. I. In your list of Nominal Values the fourth from the bottom indicates "Res. Fa- cility." Does this mean Research Facility, and if so, what constitutes a research facil- ity? 2. You state that the criteria are addi- tive, not duplicative, and I am not quite sure what you mean by these two terms. 3. In your list of N aminal Values, you specify both titles and volumes in different categories. Do you have a formula that you apply to the number of titles to convert them to volumes? Would 1.25 X volumes be reasonable? My thinking is prompted by the fact that many universities do not have separate statistics for the number of titles they own, but nearly everyone has an idea of the number of volumes they own. 4. In your application of the criteria to Hofstra you had a large number for Histori- cal Growth. Since this factor is not included in your N aminal Values, I wonder how you compute it. In the kind of evaluation our committee is talking about it will be a siz- able factor in computations. I should be glad to have your usually penetrating com- ments on these points. Bob Blackburn in his comment on the EFL draft of the Position paper urged the provision of additional space in the com- puter complex beyond that which you de- scribed for the purpose of housing comput- er equipment for conversion from one stage to another, and for standby computer equipment. He indicated that librarians who had used computer techniques for some time have been concerned with the need for standby equipment to carry on procedures during down time. I wonder if you have run into this need to date? If to the already high cost of computer equip- ment we must -add additional high costs for standby equipment, it certainly will slow down the rate of application of the com- puter to library techniques even more than now. I wonder if joint-use, standby equip- ment is not possible at least within limited areas? I should be happy to have your com- ments on this problem. All of us watch with great interest the development of your institute and look with considerable envy at the university system that is wealthy enough to launch you. Cordially, Ellsworth Mas on Director of Library Se1·vices Hofstra Univ ersity Hempstead, Long Island New York September 1, 1967 Dear Ellsworth, Your comments requesting amplification of the criteria I sent to you are all well tak- en, and each epitomizes at least one of the problems one faces in using such criteria. Let me handle each in tum. 1. "Res. Facility" does indeed stand for R esearch Facility and is intended to cover all administrative entities established for "organized research" (as contrasted with individual faculty research) , viz., those "in- stitutes," "laboratories," "centers," etc., es- tablished to administer grants and contracts in specific subject areas. There are probably 100 such institutes in the University of Cal- ifornia today (the Institute of Government and Public Affairs, the Brain Research In- stitute, the Institute of Library Research, the Law-Science Research Center, the Mri- can Studies Center, etc.) . They place an immense burden on library resources. The problem is how to measure it. Alternatives Letters / 385 would be "Number of Grants and Con- tracts" or "Number of nonfaculty research staff," etc. 2. "Additive, not duplicative" means the following: If I have a doctoral program, I want 12,000 volumes no matter how many students or how many faculty I may have. Thus, if wi is the number of volumes for factor i, and ni is the size of academic pro- gram for that factor, the total number of volumes would be: N = n1w 1 + n2w 2 + .... + n 1{wk. That is what additive means. Duplicative would mean that there was overlap among criteria and, for example: N = n 1w1 or N = n2w2 or N = nkwk (i.e., the total number of volumes might be expressed as, for example, 100 volumes per student, or 50,000 per doctoral program, etc.) 3. Replace all use of "titles" by "vol- umes." The version I sent you apparently was copied from an early copy when I had not yet resolved the problem you raise, viz. , libraries can't tell how many titles they own. Don't use a conversion of any kind from titles to volumes. Simply replace. The problem of course is more than simply the difficulty in knowing a number; it's one of appropriate measures for different purposes. (Parenthetically, "volumes" is intended to cover all bound volumes, including bound serials. I have no present means of accom- modating microforms, unbound serials, newspapers, etc.) 4. Historical growth is the result of that apparently innocuous "2 percent per year" annual growth. For some reason people never seem to appreciate the real effect of "exponential growth" (which 2 percent per year represents) . First, let me comment as to its purpose and what it is intended to do. Even if a campus were to have a complete- ly stable academic program-no growth in students, in faculty, in number of degrees, etc.-the library would still need to grow simply to keep up with the publications in the fields of present academic interest on campus. New bound volumes of the rele- vant serials must be added; new books in the fleld are written; old books must be re- placed; etc. What is that growth in stable fields? I estimate that, on the average, it's 386 I College & Research Libraries • September 1971 about 2 percent per year. Second, what is the effect of 2 percent per year? Well, as- sume you have 100,000 volumes and no growth in academic program. Then presum- ably only the 2 percent growth would be operative. In ten years, you would add 23,000 volumes and in twenty years, nearly 50,000 volumes. Of course, when this growth rate is added to a growth in aca- demic programs, the effects are much great- er. 5. Your final comment, relating to Bob Blackburn's suggestion that standby equip- ment be included in planning is an impor- tant one. On the one hand, libraries are operational agencies and cannot afford to be brought to a screeching halt because equipment is nonoperative. On the other hand, the economics of equipment in li- braries is so marginal anyway that the add- ed burden of nonproductive standby equip- ment would virtually eliminate it from eco- nomic consideration. My answer is a rela- tively unpopular one, but I am convinced that it is the only viable one. The library system of machine utilization must be de- signed to include the facility for machine independent operation as a normal part of the day-to-day procedures. In this way, the library can continue functions in pretty much its normal way even if the machin- ery is down. As far as "joint-use, standby" facilities are concerned, I am very dubious, unless they are also used as an integral part of day-to-day procedures. However, there is always the necessity of identifying com- patible installations in use in the near vi- cinity which are willing to provide "second- shift" time to you. That's a different matter from "joint-use standby," however, at least as I would interpret your meaning. D ear Bob: Sincerely yours, Bob December 29, 1967 We seem to be at ·the point of getting an IBM 1790 as a gift, used, but still quite good. I w onder if ownership of a computer changes radically their economics of com- puterizing library operations. If it would, the problem of replacement would remain, and I wonder how long is the effective life of such a computer. Many thanks for any information you may be able to give us. Dear Ellsworth: Sincerely yours, Ellsworth January 5, 1968 What a pleasure to hear from you! And particularly with the news that you are get- ting an IBM computer as a gift. (You list IBM 790 with an inked "one" to produce IBM 1790, and I'm not sure what it is, since I don't recognize either number. Is it an IBM 7090?) I assume that you are re- ferring to the university, when you say "we," and not the library itself. Now you have posed two questions. 1. Does the availability of the computer (for free, or effectively so) radically change the economics of computerizing library op- erations? 2. How soon might the library be faced with the necessity of changing its operation again, if and when the computer is changed? First, I would suspect that the availabili- ty of the computer would have a negligible effect upon the economics of computeriza- tion of libraries. There are several reasons for this opinion: ( 1) The computer costs for library clerical operations are probably small anyway and even reducing them to zero won't have a determining effect. ( 2) The computer itself usually represents only a part of the operating expense of a com- puter installation (operating personnel, p e- ripheral equipment, etc., all would consti- tute continuing costs). ( 3) The costs for "system development" (i.e. , systems analy- sis and evaluation, programming, conver- sion, check-out, etc.) represent the over- whelming factor in the library's decision; these are costs incurred independent of whether the equipment itself is free. ( 4) Another issue in the decision is not an eco- nomic one as such, although it has very significant economic consequences-viz. , what is the basis for availability of the com- puter? Remember, the library is an opera- tional agency and must be guaranteed scheduled, ready, and continuing access. ( 5) Which brings me to the economic im- plications of the noncomputer issues in computerization. Mechanization of library clerical processes will involve significant changes in library processes a~d metho~s of operation for the library staff Itself. Therr costs become the dominating economic is- sue. They are dependent on the fact that the computer is free only to the extent that one may be able to put more of the burden on the computer than one would normally be able to afford. All of which says that the decision to "computerize library operations" is aff~~ted by free computer time only in a negligible way. Second, the rate of obsolescence of com- puters is something fascinating to behold, and I am not at all convinced that it is re- alistic. Each new generation of computers has been more capable and has provided more "computing power for the dollar" than the previous ones (and by orders of magn~­ tude, not just by minor amounts) . But It has also created a great number of prob- lems in conversion to the new computer. For those like the library or the universi- ty's admurlstration, with an operational use of the computer, the likelihood is that 0e problems in converting to a new generatiOn of computers are greater than the hypo- thetical improvement in efficiency would warrant. (Recall that the computer itself represents only a small issue in comparison with other costs.) As a result, for them ob- solescence is of minor importance and the possibility of "using a better computer" is less likely to affect the decision to ~han~e. Unfortunately, however, in the university the bulk of computer utilization is not op- erational but ad hoc. For such use, the ex- pansion in capability and co~putin? power per dollar weighs very heavily. Smce th.e ad hoc users are likely to control the deci- sion as to whether to change a computer, I would anticipate a rapid rate of obsoles- cence. To say it another way-the effective life of a computer is virtually unlimit~d \ al- though there will in time be a detenoratwn in its performance reliability), and there- fore computers don't become worn out. ~d they don't become obsolete very rap1dly. But they do become obsolescent, in the sense that something better can replace them. The problem you will be faced with is that you are probably going to be depen- dent upon someone's else decision as to when the computer needs to be replaced. Letters /387 I've probably said more than you wanted to hear, but I'll be interested in learning how you proceed. Dear Bob: Sincerely yours, Bob January 17, 1968 Your letter of January 5 provided a total- ly complete and concise answer to my in- quiry, the likes of which I would be grate- ful to receive from everyone of whom I ask questions. The machine is an IBM 7090. Let me place this inquiry: How does the library world begin to move toward stan- dardized programs to computerize library methods that will provide the basic results to anyone willing to accept the program package? Systems development seems to in- volve tailor-made analysis in minute detail of the progression of methods presently in use in a library, including some improve- ments, and then programming the comput- er to perform them. Supposing we were to forget about in- terim methods and not care how they were done so long as terminal actions result- that is a book would reach the shelves with cards-~ockets-labels, cards, would end up in the catalog (or entries in a print-out cat- alog), etc. If one library would program to achieve these results, why could not any other li- brary with the same computer accept the same program to achieve the same ends, disregarding the middle? The variables in terms of ends are not great, whereas the variations in programs seem to be total. Given this cost relief, I should imagine that computerization of library methods would be possible for many more libraries than now use them. Dear Ellsworth: Sincerely yours, Ellsworth January 24, 1968 Your letter of 17 January 1968 raises what has been perhaps the most frustrating issue in my work over the past five years, and more. In principle, there seems to be little doubt that a "packaged program" will widely serve the library community. The 388 I College & Research Libraries • September 1971 frustration for me comes from my own in- ability to bring it to reality. Why? There seems to be a number of hurdles to be overcome. 1. Packaged programs have only recently become recognized as useful in the comput- ing community itself. There is little glamour in their development, and as a result, most of the really good programmers have con- centrated on the development of "program- ming languages" (such as Fortran, COBOL, and PL/ 1) and "operating sys- tems" (which manage the computer itself, particularly when it must handle a variety of programs and a number of users) . 2. The Computer Configuration is an overriding consideration in the actual crea- tion of a packaged program. Therefore, since installations >available to individual li- braries differ radically, it is not clear that a large number of libraries will really be able to use a packaged program developed for a particular machine configuration. 3. The Changes in computer configura- tion make obsolete any packaged program which has been operational. This effect has been particularly devastating over the last two years, with the change-over from IBM 1401 and 1410 to IBM 360/30 and 360/40 and from IBM 7090 to IBM 360/ 50. Whereas there were well-proven programs operational on the earlier machines, they suddenly became nearly worthless with the new ones. During the past two years, this effect has been amplified as the 360 instal- lations themselves have undergone a succes- sion of changes-in both hardware and "operating system." 4. The operating procedures, as I point- ed out in my earlier letter, represent the really significant issue as far as the library itself is concerned. Unfortunately, it is al- most impossible to separate the design of the computer program from the design of the library's operating procedures. This is why computer people invoke the magic of "total system design" in which the program and the procedures are tied together. One could conceive of a package which includ- ed standardized procedures as well as stan- dardized computer operations, but that's where the rub comes. The effects on the li- brary itself are now the predominant issue. To illustrate: a package for serial control depends upon the procedure for serial check-in; This differs so radically from li- brary to library that it is virtually impossi- ble to standardize. 5. The re are legitimate differences among libraries and their operating procedures- size, policies on service, type of institution, etc.-which appear to preclude any stan- dardization of procedure. As a result, one must think in terms of a set of packages, or perhaps a set of "modules," which provide subcapabilities and can be put together in different combinations to form the desired set of packages. But this multiplies the task of development-either by the number of different packages or by the greater work of defining appropriate modules. 6. The "Not-Invented-Here" Syndrome seems to be ever present, and minor differ- ences, to which presumably one could ad- just, are used as reasons for separate devel- opment. 7. The need for Library Systems Analy- sis is present anyway because of the cost considerations in the library, which are much larger than simply those of mechani- zation. So perhaps the NIH syndrome is not bad, anyway. Despite all of this, I am personally con- vinced that packaged programs are the an- swer, and we are continuing to pursue the analysis of what they should do and of how they should be developed. I look forward to your own reactions. Sincere ly yours, Bob MoRE oN MASON ... To the Editor: I find Ellsworth Mason's paper, "The Great Gas Bubble Prick't," in the May 1971 issue of CRL an unfortunate addition to the library literature. An analogy became unmistakable to me while reading the paper, namely the picture of a youngster with hands and face smeared with chocolate frosting looking innocently at his mother, pointing to his brother's face at the kitchen window, and saying, "But mommy, he made me do it." I find credibil- ity difficult for Mason's statements that li- brarians exhibit "command and critical bril- liance" in daily library operations and apply "intellect and managerial methods" in li- brary practice, if I am asked simultaneously to believe that librarians have been "lem- mings" with respect to library automation. Though individuals do not behave with 100 percent consistency, the amount of behav- ioral difference given by Mason suggests that one of the descriptions is closer to the usual pattern with his evidence pointing strongly to the latter. Mason forgets, in his illustration of the delivery of a Continental automobile that, if the purchaser doesn't know how to drive, mere delivery won't permit the purchaser to use the hardware for transport purposes (unless he has a chauffeur). Continuing the analogy of the automobile, use of it involves greater expense and greater per~onal and environmental hazards than use of the horse and buggy. It also provides a different mode of transportation. Its wide acceptance appears to suggest that pros and cons have been weighed in favor while, simultane- ously, efforts are made to reduce negative aspects of its use. Though Mason furnishes few specific facts to support his generalizations, knowl- edge of library automation experience per- mits me to accept what he has said as phe- nomena that can and have occurred. How- ever, the appearance of scholarship con- veyed by footnotes that are woefully lack- ing in authoritativeness I find deplorable. I would have expected my students to sup- port the "Truths" from data in the automa- tion literature. Mason sets a poor example, both for students and his peers. The salutary aspect of Mason's paper is that it enables librarians and library educa- tors to have a better understanding of the knowledge, skills, and managerial capabili- ties that are needed by members of the pro- fession. If a whipping boy must be found for librarians' dilemmas, he is less likely the computer and more likely the growing de- sire of librarians to streamline their opera- tions and provide active rather than passive information services. Judgmental errors made in library automation projects are symptomatic growing pains. They can be learned from if analyzed maturely. Rowena Swanson Professor Graduate School of Librarianship University of Denver Denver, Colorado Letters / 389 To the Editor: The most perceptive and entertaining pa- per that I have had the pleasure of reading in a long time was that of Ellsworth Mason in the May issue of CRL. Someone to point out the cost problems of computer use and the pitfalls of automation has long been needed. Let us hope that Mason's warning cause those who have not yet ventured into the computer water to have second thoughts and those who have already been scalded to re-evaluate their costs before they pour more money down the drain. To the Editor: Richard A. Davis Assistant Librarian The John Crerar Library Chicago, Illinois Thank you for the bitter, literate, inci- sive, derisive, funny, and very, very good article by Ellsworth Mason, "The Great Gas Bubble Prick't," in the May 1971 is- sue. To the Editor: Clyde King James M. Milne Libra,ry State University College Oneonta, New York The Gentleman of Quality is to be com- mended for his virtuoso performance with the English language. Buried deep within his florid rhetoric there are even little dabs of truth here and there. After all, a fair- minded evaluation of the computer must admit to some failures, but let us not con- fuse wishful thinking with fact. The effect of ow· Gentleman's tirade against computers has not been to puncture a bubble; rather, he has built a balloon, filled it with hot air, and gone on a trip of fancy-such a contrast with a related ar- ticle by Ellsworth Mason in the May 15 is- sue of the Library Journal! That a1iicle, ti- tled "Along the Academic Way," is a high- ly rational, carefully considered statement of several major problems facing academic libraries (automation being only one of these problems). But "The Great Gas Bub- ble ... " is an emotional outburst unbecom- ing to any man, least of all a Gentleman, especially one of Quality. There are good reasons to question the 390 I College & Research Libraries • September 1971 automation of a library. After all, what is to b e gained by mechanizing an existing process without examining the underlying assumptions? For example, if one simply automates a 3x5 card the result is still a 3x5 card. Successful automation is preceded by a rigorous questioning of all assump- tions, procedures, and methods. Only then is the path clear to think about automating. The computer is a tool. As with all tools, a requisite for productive use of the com- puter is that one fully comprehend what it can do and what it cannot do. A simple tool is limited in its applications and re- quires scant direction to be effective. A complex tool can be used for a greater number of applications but requires a great- er number of directions. The computer, an extremely complex tool, can be applied to- ward an almost infinite variety of applica- tions because it can be programmed in as many ways. To suggest that the computer should be programmed in advance of man- ufacture is to miss the point of the comput- er altogether. So-called computer failures do occur, but they are by no means universal. Most are human failures: witness the saying "Gar- bage in, garbage out," or the Programmer's Lament, "The computer always does what I tell it to do-damn it!" The real culprits are the inept programmers hired by the vague employers who can't (or won't) say what they want the computer to do. The economics of automation is a com- plex picture. Only rarely can manual and automated processes or their costs be di- tectly compared, for very often the magni- tude of leverage and range of services are substantially altered. A bona fide cost com- parison not only examines the before and after costs (along with before and after services) but also the cost and effect of du- plicating the automated services (including all of the by-products) by hand. Of course, such comparisons are properly made in ad- vance as well as after the fact. There is important work to be done in the automation of libraries and information services. One reason that the results have been sometimes disappointing is that we are dealing with words-language. A prin- cipal key to future work lies in the field of linguistics. As we improve our understand- ing of how words work, we can build more efficient, more effective, and more rapid means of controlling information. Larry Auld Head, Technical Services Oregon State University Library Corvallis To the Editor, When I was told some weeks ago that a comment which questioned the effectiveness of library automation had appeared on the cover of the Library I ournal I was pleased. The computer has been an essential ele- ment of my teaching, writing, research, and other work for nearly twenty-five years, and I have been appalled by the intellectual corruption and the waste of funds that I have seen in ill-conceived and dismally mis- managed automation projects, in a variety of fields; and by the drivel that has been promulgated as so-called computer science. I have felt, however, that reasoned criticism of such matters was regarded as bad form, or perhaps just noncomformist and therefore irrelevant (like criticizing the Vietnam war until a few years ago). "Computerization regardless" was the accepted dogma, and as such could not be assailed simply by reasoned argument. I was pleas ed, therefore, to hear that an influential journal had given prominence to a questioning of this dogma by a librarian -Mr. Mason-and even more pleased to hear that he had published an extensive critical aliicle in College & Research Li- braries. I have read the article several times. I think that many of Mason's specillc criti- cisms of the absurdity of individual projects, the abandon with which they were launched, and the irrationality with which they are being escalated may well be valid. I think that a great service has been done if his article leads to reasoned critical analysis of present projects by more people, and to reasoned analysis of future plans. But allow- ing the possibility that some or perhaps all of the projects that Mason visited are as bad as his scatological allusions suggest, I am bothered by its overall tone. It is the recourse to dogma in professional matters that alarms me far more than the dogma's specific content, when I hear un- sound computational projects "justified" on the grounds that "computerization is es- sentially good," just as when I hear un- sound elementary math textbooks "justified" on the grounds that "the child must be taught at the earliest age that sets are the theme that unifies all mathematics." It would be oversimplistic to blame the ills of professional life on a conformist re- quirement for administrators to pander to current dogma in order to maintain credi- bility, or to suggest that this just happens in the U.S.A. today. I think it is a serious problem , however, and the dangers inherent in the party line approach are in no way changed or mitigated when some of the cliches happen to get reversed. I am wor- ried that Mason's article may have just this impact, providing a pseudo axiom "com- puterization is inhe1·ently nonbeneficial" to axe and to block work of potential merit as indiscriminately as patent absurdity has been launched and adulated in the past. There are several questionable points of technical detail in Mason's article that could contribute to such a switch, and I would like therefore to comment on these in my role as a technician, particularly since he advocates the use of reason and decries re- course to dogma. Mason states that "the computer is not subject to reasonable surveillance at any level of operation." He says this is "a fact" (footnote 5), and explains that it is "in- herent in the occult nature of the com- puter." Since I associate the word "occult" with the supernatural, I was just a little wonied that Mason had succumbed to the belief that spindling, folding, and otherwise mutilating IBM cards is discouraged in def- erence to the laws of sympathetic magic. I do take exception his remark, however, even with the use of "'occult" to mean hid- den, concealed, secret, communicated only to the initiated, not appreh ensible to the mind, recondite, mysterious , unexplained, etc. The behavior of the computer is deter- minististiQ-inexorably, relentlessly, inhu- manly deterministic-and it is documented extensively and, at times, quite intelligibly too. I believe the computer can be immense- ly beneficial, that reasonable surveillance at every level of operation, though often diffi- cult, is possible, that many of the people who work with computers do metaphorically get away with murder, and that they are largely responsible for the myth that Mason is citing as fact. I think that society will really benefit from the computer when ad- ministrators assume the necessary control, Letters / 391 as I and others have discussed at various times in the past. 4 Proclaiming this to be impossible is a profound disservice to those who are trying to raise the standards of computer use and education. Mason is convinced that "the high costs of computerization make it unfeasible for library operations." He quotes Veaner: "The old idea that an automated system could be operated at a new lower cost than a manual system is dead, indeed." I cannot accept the failure of appallingly expensive efforts to automate some library work as proof that no library work may ever be automated in- expensively and advantageously, since I have seen modest efforts to automate other types of work succeed alongside grandiose disasters. I hope that future proposals for library automation will be reviewed in a way that allows inexpensive studies to be made which note the causes of past fiascos and avoid their repetition. It would be sad if such efforts were blocked by the accept- ance of a myth that all computer studies require vast funds, or a myth that automa- tion is inherently and inevitably more costly if not downright disastrous. Mason believes that computerization "will become increasingly expensive in the future." He states "a computer operation is· incapable of becoming stabilized" and speaks of "the agonies, dislocations and set- backs involved" in a change of computer generations, "with no assurance that the same level of result can be achieved." Later he speaks of "the agonies of programming, reprogramming ... deception by computer experts .... " Mason states "computer ex- perts laughed when I suggested economy as a motive for adopting the computer." He states as "absolutely false" the possibility of "economies in future programming by having programs convertible to later genera- tion computers." He says "all the library computerators I questioned agree that trans- ferability of programs is completely un- 0 See, for example, th e introduction in the au - thor's book Computer Programming :in English ( Har- court Brace & World, 1969) and his chapter, "Com- puter Hardware and Software for Librarians," in the Proceedings of the 197 0 Conference on Collaborative Library Systems D evelopm ent ( to be published by the MIT Press later this year). The supervision of pro- gramming, the achievement of flexibility, the dangers of overselling new t echnologies, and many of the problems of computer typesetting that were ignored and which h elped get the field a bad name a re dis- cussed in the author's book, Computer Typesetting- Experiments and Prospects (MIT Press, 1965). 392 I College & Research Libraries • September 1971 feasible at present and in the future." I disagree categorically with every one of these statements, and I connect them with attitudes toward computing that are par- ticularly rife around some university in- stallations. Over the years I have moved programs from IBM 704 to 709 to 7090 to 7094, from a PDP6 to an RCA Spectra 70- 45 back to an IBM 7094 and then to a 360-50; in the last two years I have been running programs in parallel in installations of several different models of IBM 360; I have moved programs back and forth across the Atlantic, and onto XDS and CDC and UNIVAC machines, and even onto some English computers; I have carried on while central processing units were changed and operating systems upgraded; and this has been completely nontraumatic, at least in the commercially operated service bureaux in which the programs have been processed. I wish I could say the same of all the uni- versity installations with which I have dealt. It is possible for a crew of systems pro- grammers to keep an installation in a state of constant upheaval quite unnecessarily, and in particular without the slightest change of hardware or software by the manufacturer. I have seen computer center staff force users out of compatibility with other installations in matters that are com- pletely standard for reasons that seem to range from dowm·ight incompetence, to an arrogant desire to exert control over other people's work, to regarding the computer as a toy for their personal amusement and a vehicle for practical jokes that verge on the malicious. A bad workman often blames his tools, and "genius type" programmers in applications groups at times do their bit also to contribute to the agonies that Mason describes. Mason confuses sharing a computer (I share the local public library with other readers) with time-sharing (but we do not try reading the same book concurrently). The cost of on-line consoles is quite irrele- vant to the cost reduction of batch proc- essed work on a powerful computer that other people use for batch processing as well. As regards Mason's comments on ris- ing personnel costs, it is true that pro- grammers are included in the present up- ward drift of salaries, and it is true that a powedul machine requires a lot of work to justify its presence, and in consequence a large number of user personnel may be running jobs on it. It is the programming effort per application, however, that should be considered, and the ease of writing and debugging programs has been increased considerably by recent hardware and soft- ware developments which have permitted a considerable increase in the cost effective- ness of the time of applications program- mers. Mason refers to the Emperor's New Clothes at the end of his article. I tell my students every semester to bear it in mind whatever the cloth is supposed to be. The plot of the Alchemist has been replayed quite a few times on the computational scene, too, and category J in Stith Thomp- son's classification of folktale motifs (from "Absurd disregard of facts" to "The easy problem made hard") includes prototypes of several computer situations that Mason mentions , and more beside. But they could apply to unreasoned indiscriminate axing also. I think that Mason may have been a little harsh on university administrators and the computer industry. Perhaps I am somewhat naive, but I think that the wish "to do good" plays a major role in many administrative decisions by people who may be pressured and given misinformation. As far as the pressuring itself is concerned, the age of the grey flannel suit in the ivory tower may be on the wane, but the salesmen employed by the manufacturers would have been for nought without the husksters on the faculty. Using a computer may dramatize the cost of charlatanism, but is not a prerequisite. To what extent are situations Mason de- scribes so eloquently being recast now with other "new technologies" as backdrops, and for that matter has professional and public life been free of faddism hitherto? Mason's article will doubtless bring li- brary automation under wider scrutiny, and may bring wider recognition to the projects that deserve it, as well as a curtailment of same causes of waste. Dr. Michael P. Barnett Director of Research and Development The H. W. Wilson Company Bronx, New Y ark