College and Research Libraries By O L I V E R T . FIELD An Experiment in Catalog Reform A CARD CATALOG is like Mark Twain's weather, about which everybody talks, but does nothing. These are the familiar reasons why people complain about the catalog: it is too difficult to use; it is too big; it does not analyze enough; it appears to hide information instead of exhibiting it clearly. T h e ones to do something about the situation are obviously the catalogers. However, they seldom have the time or the staff for long-term projects and are forced to hope that their running repairs on the catalog will dry up the complaints against it. In addition to being a forlorn hope, this philosophy is plainly dangerous. It is vir- tually an invitation for some superior agency (chief librarian, academic coun- cil, or library board) to issue a ukase that hardly distinguishes between cause or effect and which may worsen a situation it intends to better. T o avoid this, it is the duty of the cata- logers to deal with catalog criticism themselves, since the catalog is their product and their ultimate responsibil- ity. It is merely guided, and in no way governed, by the opinions of public serv- ice librarians. Catalogers who will not meet changing requirements of the cata- log user confirm the popular conception of catalogers as reactionary and unhelp- ful, more interested in dead detail than in living needs. When it became clear that the catalog of the Air University Library was no longer satisfying reader needs, a series of phased reforms were begun. These frank- ly experimental solutions are the subject of this paper. Mr. Field is chief, catalog branch, Air University Library, Maxwell Air Force Base, Alabama. B A C K G R O U N D T h e catalog of the Air University Li- brary was started when the library was established, early in 1946. This was a dictionary catalog, incorporating part of the catalog of the Air Corps Tactical School, forerunner of the university. In December, 1946, the catalog was divided into an author-title catalog and a subject catalog. It was felt that a central subject- authority unit would be a useful adjunct to the catalog branch. This unit was ac- tive from 1947 until 1952. T h e subject- authority unit had the task of establish- ing new subject entries and coordinating those used for books, which were cata- logued with the Dewey classification us- ing Library of Congress subject headings, with those used for report literature, which were classified by a modified ac- cession-number system usually with no applicable Library of Congress headings. So that a new and inexperienced cleri- cal staff might more quickly process the great quantities of materials coming to the catalogers, and so that the catalog in its developmental stage might keep sub- ject mobility, it was decided to use sub- ject guide cards, one for each different subject represented in the catalog, in- stead of typing subject headings on the catalog cards. See also references were typed on the guide cards. Liberal use was made of title entry, analytics for series, and subject headings. This is as it should be in the early years of a library, when the collection is small and what is available must be used to the utmost. T h e growth of the collection has been phenomenal. T h e holdings as of May 1, 1956, were 245,648 books and bound pe- 4 1 4 COLLEGE AND RESEARCH LIBRARIES riodicals, exclusive of multiple dupli- cates, and 527,854 documents. T h i s growth brought with it increasing com- plaints that the catalog was difficult to use. Such complaints are endemic to a large catalog. It is axiomatic that the complexity of a catalog's entries increase with its size. T h e y indicate that the cata- log is going into another phase of its de- velopment, and new plans are in order. A m o n g these are the necessity f o r making firm basic decisions o n the limits of the catalog's function in exhibiting informa- tion available in the library, the p r o m o - tion of such auxiliary aids as bibliogra- phies and vertical files, and the guaran- tee of their competent upkeep. A t this point the chief cataloger stops concern- ing himself with increasing the cataloged holdings and instead considers more pos- itively what he should not catalog of in- coming materials and what materials al- ready cataloged may be withdrawn or have their m o d e of exhibition con- densed. These were the problems considered in the reform of the catalog. T h i s is the order in which they were attacked: the deletion of title entries; the division of the catalog into an author catalog and a subject-title catalog; the removal of the subject guide cards; the removal of see also references; weeding; special treat- ment of U. S. entries; and chronological arrangement of heavily represented sub- jects. D E L E T I O N OF T I T L E ENTRIES T i t l e listing is the first expendable function of the catalog which comes to mind when the time comes to change the catalog f r o m its medium-size collection service to large collection service. T h i s p r o b l e m was faced in 1953 by a previous chief cataloger at the Air University when he issued a very sound and care- fully worked out m e m o r a n d u m o n title deletion, together with a list of over a hundred filing words generally to be de- leted as titles.1 T i t l e deletion is a hazardous expedi- ent. Based o n the assumption that, with certain exceptions, titles beginning with subject headings may be safely deleted, together with certain frequently used title-starters such as How, Introduction, Handbook, and titles that represent liter- ary forms, such as Essays, Poems, Plays, it soon runs into difficulties. How to Build Modern Furniture is quite a dif- ferent recovery p r o b l e m f r o m How to Get It from the Government or How to Help Your Husband Get Ahead in His Social and Business Life. T h e r e is no as- surance that the reference librarian, much less the reader, w h o discovers the first title under FURNITURE will automati- cally turn to u. s.—POL. & GOVT.—1945- for the second and SUCCESS and WIVES f o r the third. N o r does the answer lie in omitting f r o m the catalog all title entries that ad- mit of easy subject approach, but includ- ing all that d o not. Under these circum- stances, a reader finding the title How to Get It from the Government but not How to Build Modern Furniture assumes that the latter title is not in the library collection since it is not listed. Of all the unfortunate things readers will choose to remember, the worst is the un-truism: " T h e catalog lists all the books in the library by author, subject and title." T h e r e is no real solution to the title problem, short of including all o r exclud- ing all. Any middle road opens the way to varying interpretations. W h a t was fi- nally done at the Air University Library was to h o l d fast to the practice of dele- tion, inserting guide cards throughout the catalog bearing the sometimes de- leted word and a warning: H O W Titles beginning with this word are not generally included in the card catalog. If you do not find your title, look for it under 1 U . S . A i r U n i v e r s i t y . Library. Catalog B r a n c h . "Cataloging and Classification N o t e s and D e c i s i o n s 17-53 ( R e v i s e d ) A u g u s t 14, 1953." U n p u b l i s h e d . SEPTEMBER, 1956 4 1 5 subject, or ask a librarian to help you. T H E C A T A L O G IS DIVIDED A G A I N T h e other thing that was done to make the catalog more useful by giv- ing the reader some clue to omitted titles was to divide the catalog into an author section and a subject-title section. In libraries where the dictionary ar- rangement is not used, the most common plan is to arrange the catalog into an author-title section and a subject sec- tion.2 This arrangement deals neatly with the relatively minor problem of authored and non-authored entries, but assumes that readers will run down the smallest item of information, no matter how carefully it is hidden. This concept has caused the catalog to become increas- ingly a tool created by catalogers for each other. If it is true, as catalogers so often say, that even reference librarians do not know how to use the catalog, the fault is as much that of the catalogers as of the reference librarians, since, while it is true the latter must know their sub- ject headings to use the catalog, it is also true that the former must arrange it for most efficient use. While both are rauc- ously laying down the qualifications of their game of card-chase, the poor reader has quietly disqualified himself and left the field. While the reader may give up soon in a dictionary catalog, he has scarcely any chance at all in the author-title and sub- ject catalog. This arrangement ignores two facts. T h e first is that titles support subjects and not authors, save incidental- ly. T h e second point is that readers are as likely to think of titles as of subjects. Anyone who has served a term in a ref- erence department knows that a perti- nent title comes to mind at least as often as the sometimes elusive subject assigned to it. Many distinctive titles are quasi- 2 V . J. B u r c h , " D i v i d e d C a t a l o g : D u k e University- Library Catalog F a c e s the F u t u r e , " CRL, I I I ( 1 9 4 2 ) , 219-23; A . F. W o o d , " C a l i f o r n i a D i v i d e s I t s Catalog," Library Journal, L X I I I ( 1 9 3 8 ) , 723-26. subjects in that they embody the subject matter of the book in a vulgate form, often completely different from the offi- cial subject entry for it. These quasi- subjects constitute a mantrap in a cata- log divided by author and title and by subject. Even the most astute reader runs the danger of assuming that they repre- sent the total holdings of the library in his field of interest. This arrangement of the catalog makes no provision for lead- ing the reader into his subject field, since the cross references from his quasi- subject to its catalog form are not in the author-title catalog at all, but in the sub- ject catalog, and the reader may never look there. T h e arrangement by author catalog and by subject-title catalog, on the other hand, obviates completely any such cul- de-sac, since it drives the reader from his quasi-subject to the "official" subject by means of an inescapable cross reference. An example will show this. If a reader interested in the subject of federal aid to education finds a book called Federal Aid for Education in the author-title catalog, he is likely to stop there. In a subject-title catalog, this title would be deleted, as it stands close to the refer- ence FEDERAL AID T O EDUCATION See EDU- CATION AND STATE—u. s. However, it would be a most unrealistic cataloger who would delete the title Federal Aid for Education from the author-title cata- log, expecting the reader to continue his search in the subject catalog under FED- ERAL AID T O EDUCATION and thence to EDUCATION AND STATE U. S. T o come to the conclusion that the author and subject-title arrangement is the best one for a given situation is one thing. T o clear the way for this transfor- mation in a heavily used catalog of 655 trays is another. T h e first step in the shift was to draw up a list of the title entries to be transferred from the author- title catalog to the subject-title catalog. These were: all titles, all non-author series, all non-author serials, all title 4 1 6 COLLEGE AND RESEARCH LIBRARIES analytics. T h e list gave examples and outlined procedure. Next, a 60-tray unit was added to the catalog to obviate heavy shifting during the change-over, and space was made where the need f o r it was anticipated. Finally, the task of shift- ing the catalog cards was done by the readers service librarians. T h e y made time for it in their regular schedules and completed the task in eighteen days. W i t h o u t the help of the readers service librarians, the task w o u l d have been quite impossible. SUBJECT G U I D E CARDS T h e catalog was n o w ready for the next phase of change, the removal of the subject guide cards. N o one w h o has had to set u p a library catalog in an institution where consider- able growth is expected w o u l d question the decision to use guide cards instead of typed-on headings, given the need for speed and faced with the problems that numerous, changing subject headings present. However, the fact remains that the only continuing g o o d of the guide cards is that they require slightly less clerical attention, and, on the g o o d housekeeping side, they appear to keep catalog cards looking cleaner. Every- thing else is against them. Subject guide cards complicate the cataloging process by requiring that spe- cial measures be taken to make sure they are made or not made, as the need may be. T h e y slow d o w n filing because they hide cards filed above the rod, and cause them to be overlooked by revisers. W h e r e a run of subject guide cards with the same main subject but with dif- fering subdivisions has only one author card behind each of them, the question arises whether subdivisions of less repre- sented subjects are necessary o n guide cards, or, indeed, if they are necessary at all. T h i s latter query is a reasonable one, but disquieting to a cataloger if it comes f r o m the administration. A cata- loger is trained to base present practices o n future expectations, not to limit them to the present situation. T h e r e is future trouble implicit in the use, say, of the undivided heading ICELAND f o r a handful of books o n such separate subjects as C O M M E R C E , DEFENSES, DESCRIPTION AND TRAVEL, E C O N O M I C CONDITIONS, E C O N O M I C POLICY, HISTORY, a n d POLITICS AND GOV- E R N M E N T . O n e answer to this p r o b l e m is to leave subject subdivisions off guide cards, in the h o p e that readers will mine them out of the tracings at the b o t t o m of the cards. However, the visibility hazards oc- casioned by this solution lead only to gross misfiling and reader dissatisfaction. Intended to make f o r clarity and order in the catalog, subject guide cards tend rather to present a busy, cluttered ap- pearance that dismays the reader rather than aiding him. Furthermore, their sav- ing in clerical help does not outweigh their nuisance value, since cards must g o into typewriters for call numbers in any case, and the subject heading can then be easily added. T o revise typing of headings is n o time-consuming task if clerical personnel are basically able. T h e mobility advantage of the guide cards decreases as catalogers' knowl- edge of the special subject fields in which they are working matures, and as neces- sity teaches the chief cataloger restraint. Since he does not have the staff the L . C. Joneses have, he must curb his desire to keep u p with them. T h e worst feature of the guide cards is their fiendish fertility. T h i s is demon- strated by the f o l l o w i n g tables: Subject Catalog Status on 30th September 1954 Card trays in subject area 360 Number of subject cards 243,100 Number of guide cards 43,000 Breakdown of a Typical Subject Catalog Drawer on 30th September 1954 Cards in tray 631 Number of guide cards 275 Number of subject guide cards controlling less than 3 cards 180 Number of see reference cards 31 SEPTEMBER, 1956 4 1 7 Whether one looks at the lesser total figure and says that every sixth card in the subject area is a guide card, or at the larger sample figure and says that every third card is a guide card, one has an uncomfortable feeling that the guide cards will eventually take over the cata- log. There is only small comfort in the thought that the number of additional new subjects will decrease as the catalog grows. Actually, the average monthly ad- dition of new subject guides over the period 1946-1954 was 398 per month, and the average monthly addition of new subject guides in 1954 was 281. T H E G U I D E CARDS G o Plainly, the next project was to re- move the guide cards. However, the problems involved in this were not sim- ple ones: If the guide cards were re- moved, who would be found to type on a quarter of a million subject headings? A typist can type 95 headings an hour, so the task would require as a very mini- mum 65 work weeks of a typist's time. Obviously a method other than typing was required. Hand-set rubber stamps were finally chosen as the solution. As the first step, it was decided that any subject represented 25 or more times in the catalog should have its subject stamped on, and any subject represented less than 25 times should have its subject typed on. T h e catalog branch prepared a list of subject headings represented 25 or more times in the catalog, and on this list starred those represented 100 or more times, for consideration for chronological arrangement. This list contained 1,565 entries in all. T h e director's office staff set up the stamps and the readers service librarians again helped, this time by do- ing the stamping. Stamping was com- pleted in two months' time. T h e typists began typing cards for the less repre- sented subjects concurrently with the stamping operation, so that the catalog might present its changed appearance as soon as possible. While no one would say the catalog now presents the highly groomed appearance given by tray after tray of meticulously typed headings, it can be reported that stamping has re- duced by over one half the time required to get rid of the guide cards. T h e question may arise why an outside service should work on a project that would appear to be the responsibility of an inside service. One cannot think the readers service people welcomed this task—who would, being in his right mind?—but they did need to know what was taking place in the catalog so that they could use it most efficiently while it was undergoing change. It was impor- tant, too, that they should see how the old construction of the catalog was im- pairing its present usefulness. Needless to say, they turned up some cataloging howlers that needed attention, and throughout the project they asked pointed and thoughtful questions con- cerning the value of present practices. Credit for the accomplishment of this task is due the readers service librarians, but even greater credit is due the cata- logers who for a decade worked to create so basically sound a fabric that it could withstand being rent apart and reorgan- ized. It is not every catalog that is worth reorganizing. Cases are known to exist in which a completely new start is the best solution to the problem. Thanks to the work of the present and former staff, the catalog has assumed its new look, with an average of six guide cards in trays which used to hold as many as 275, and redundant title entries have disap- peared. T H E See Also REFERENCES Under the old guide card plan, see also references were typed on the subject guide cards. As cards were stamped or typed during the catalog's reformation all guide cards were turned over to the catalog branch, where those with see also's were retained, and those with no see also's were destroyed. 4 1 8 COLLEGE AND RESEARCH LIBRARIES Removed from the catalog were 4,587 guide cards with see also references, and 32,750 "plain" guide cards, exclusive of those in the U. S. subject area. Under these guide cards were 105,570 subject entries. T h e decision to be made con- cerning these guide cards is whether they should be returned to the catalog wholly, in part, or not at all. T h e final decision depends on the opinions voiced by the readers service librarians. So far they have said nothing, one way or the other.3 If this silence continues, it is likely that only those see also references that refer from subjects originated locally will be returned to the catalog. See also's originated by the Library of Congress would under this plan be removed per- manently and the Library of Congress printed subject lists be placed near the catalog for perusal by those interested. It is known that the maintenance of see also's from ABBOTS see also CHRISTIAN BI- OGRAPHY to ZUNI INDIANS see also PUEBLO INDIANS takes the full time service of one cataloger contributed piecemeal by the whole staff. This is too great a price to pay for the upkeep of Abbots or of Zunis or for what lies in between. Such changes as these cannot be under- taken without staff-wide publicity. This was done by means of A Manual of Prin- ciples on Limited Cataloging for the Air University, issued in its fourth draft form in May, 1956. Drawn up after consulta- tion with the readers service librarians, this manual incorporates their expressed needs and the local catalog practices. F U T U R E PLANS Other plans for increasing the useful- ness of the catalog are still under study. In the probable order of their considera- tion these are: precataloging selection, the treatment of retrospective materials and the simplification of entry, particu- larly for U. S. government bodies. 3 T h i s appears to be f o l l o w i n g the experience of the U n i v e r s i t y of California catalog, which contains se- lected see also's from main subjects to identical place or language subdivisions, but does not use see also's per se. I n s t e a d , the Library of Congress subject head- ing lists are made available at the subject catalog. No matter how it is handled, catalog- ing is a time-consuming process which should be applied only to materials pos- sessing more than temporary value. It is felt that much of the confusing bulk in the catalog will disappear when the catalog stops being the only and inevi- table terminus for guidance to acquired materials. Complete plans remain to be made in other methods of disposing of ephemeral material, and among these are the promotion of the use of indexes, ver- tical files and wastebaskets. It is in the first two that readers and public service librarians should look for current affairs briefs, topical serials and other ready- reference materials whose interest is lim- ited in time, and into the last that more gift material can often go. This does not imply that the library's present acquisitions policies are unsatis- factory. Anyone familiar with these, and with methods by which they are carried out can only have the greatest profession- al admiration for them.4 However, cata- logers also have a logical responsibility for participation in selection. When cata- logers take this part, their mental atti- tude is improved, because they need no longer wonder, as they sometimes must, if they are engaged in nonessential work. There are at present 511 subjects in the catalog represented 100 to 800 times. T h e approach to cataloged materials would be made easier if retrospective materials were withdrawn, possibly to an historical catalog. Such an operation would be selective, and not a blanket re- moval by date. T h e remaining cards, ar- ranged chronologically by imprint, the latest first, would have a form reference to the historical catalog. An advantage of date-arrangement of the remaining cards would be the ease with which sub- jects could be kept up to date. It is well known that the U. S. area in any catalog is rough terrain for the com- 4 Mary L o f t o n Simpson, " A n E x p e r i m e n t in Acqui- sitions with the Lamont Library L i s t , " CRL, X V ( 1 9 5 4 ) , 430-33. (Continued on page 430) SEPTEMBER, 1956 4 1 9 edited, manufactured and distributed to some 5,000 ACRL members last year for a total cost, including headquarters salaries, of $4,700. The key to ACRL's financial picture lies, of course, in the main budget, which in- cludes the net cost of the journal. ACRL began small in number of members and in income, and grew slowly during the first years. It had cash balances of $8,500 in 1949, $11,500 in 1950, $14,000 in 1952, and a high of $16,800 in 1953. As this balance increased in size it became apparent to the Board of Directors that a broadened program was possible and desirable. ACRL is a service organization and takes no satisfaction in amassing a large bank account. Therefore in the fall of 1953 the ACRL staff was in- creased from 2.5 to 4 with the expectation that this staff cost would result in severe deficits for several years and might be more than the division could finance indefinitely. The divisional year-end balance decreased to $13,350 in 1954, $8,600 in 1955 and will drop to approximately $6,800 in 1956, ac- cording to official May estimates. Of course the ACRL program expanded in many ways other than staff in the past three years. It is a particular satisfaction to report that income will nearly equal expenditure this past year because very large sums have been spent on the Organization Manual and the foundation grants program, which will be self-supporting in 1956/57. The year now ending is the last during which ACRL will receive support under the ALA formula which has been in operation since 1950. Likewise this is the last year for operation under a divisional budget drawn up by the Board of Directors. During the next few months ACRL will turn over to ALA its current balance of approximately $6,800 as well as Monograph assets esti- mated at $2,750. In short, our current financial picture is sound. Our program as well as our costs were nearly doubled in 1953, but the divi- sion has been able to build its income suf- ficiently, without any additional or special aid from ALA, to the point of a nearly bal- anced budget, and to maintain a reserve for emergencies and special projects.—Arthur T. Hamlin. A n Experiment in Catalog Reform (Continued from page 419) m o n reader. Its complexities are multi- plied in a collection that is both govern- ment sponsored and heavy with report literature. A t present, the U. S. is being removed f r o m all official U. S. corporate entries, and the cards are being refiled under the next w o r d in their heading. For example, u. s. DEPT. O F STATE is be- ing changed to DEPT. O F STATE. A n even more drastically direct entry is planned for A i r Force headings which will change u . s. A I R F O R C E , STRATEGIC A I R C O M M A N D , for example, to STRATEGIC A I R C O M M A N D . T h i s treatment is very concisely de- scribed by C r o x t o n as "Entry . . . by the smallest significant c o m p o n e n t . " 5 Entry similar to this has been used successfully for some years in the Air University Pe- riodical Index. T h e s e are the future plans for the cata- log. T h e y call f o r a great deal more thinking as their full implications are not k n o w n nor their pitfalls completely explored. A catalog as big as that of the A i r University will make too great a noise if it collapses under ill-considered change. W i t h us, the needs of the cata- log as well as those of the p u b l i c it serves are under continuous scrutiny, and for the rest there is inspiration in the motto of the A i r University itself: Proficimus more irretenti. T h i s writer prefers to translate it as: W e are of service, not being held fast by rules. 5 F . E . C r o x t o n , " I d e n t i f i c a t i o n of T e c h n i c a l Re- p o r t s . " I n The Production and Use of Technical Re- ports, ed. b y B e r n a r d M . F r y a n d R e v . J a m e s J . K o r t e n d i c k ( W a s h i n g t o n , D. C . : Catholic U n i v e r s i t y of A m e r i c a P r e s s , 1 9 5 5 ) , p. 127. 430 COLLEGE AND RESEARCH LIBRARIES