College and Research Libraries B y M A R Y D A R R A H H E R R I C K The Development of a Classified Catalog for a University Library' Miss Herrick is associate librarian, Chen- ery Library, BQston University. TH E C H E N E R Y L i b r a r y is t h e c e n t r a l unit of the Boston University L i b r a r y system. I t serves the College of L i b e r a l A r t s and the G r a d u a t e School primarily but is used regularly by students from the other six colleges on the campus and by some off- campus groups. T h e subject catalog for this collection is a classified catalog based upon the L i b r a r y of Congress system. Although catalogs in classified form are historically ancient and exist today in a f e w special libraries or for some special collections, it is believed that the Chenery L i b r a r y is unique among university libraries in this country in using this medium as the only subject catalog ap- proach to its central collection. It also serves, to a limited extent, as a union sub- ject catalog since six separate college collec- tions and one bureau library are being cata- loged in the central unit. T h e s e holdings have, or w i l l have, principal entries in the classified catalog. ( T h e r e is no union sub- ject representation at present f o r the nine other college libraries in the system.) Although no other university may at this * Miss Herrick's thoughtful and expert paper should answer many questions which have been asked concern- ing the classified catalog at Boston University. Admit- tedly the decision to inaugurate the venture, made by Mr. Floyd E . Orton when he was director, was a bold one. Not all of the problems which have been en- countered were foreseen and it is a tribute to the ingenuity of Miss Herrick, who has been in charge of the project from the beginning, that they have been so successfully solved. The procedures necessary to create such a catalog have been, I feel sure, successfully estab- lished. The problem of the effective use of any catalog is one in which the college library's role as a teaching instrument must be continuously stressed.—Robert E. Moody, director, Boston University Libraries. time wish to f o l l o w Boston's experiment f o r its total collection, there may be institu- tions that would wish to consider a classi- fied catalog f o r a part of their collections. F o r these latter and f o r all interested in media of subject analysis this resume of the w o r k i n g program f o r a classified catalog is prepared. T h i s catalog w a s begun in September, 1 9 4 8 coincident with the start of a re- cataloging project designed to cover the central library and several of the other college collections. A t that time it w a s decided to use the L C classification system where total recataloging seemed needed. Certain units were in good condition and there w a s no justification f o r reclassifying those colleges at that time but procedures were established whereby they could be in- cluded in the union subject catalog. T h e s e w i l l be discussed later in this paper. Inasmuch as there is no relative index to the L C classification, it w a s necessary to establish one. T h e terms used in this index come from three sources. T h e first, in point of choice, is always a heading from the L C subject, heading list. T h e assign- ment of subdivisions is more flexible. In addition to using those provided in the official list, terms are selected from the schedule tables and f r o m their separate indexes. T h e references that correspond to see references in a dictionary catalog are chosen f r o m the Subject Heading list or from current usage. W i t h a classification as detailed as the L C it is possible to prepare a subject index 418 COLLEGE AND RESEARCH LIBRARIES that is more of a precision tool than is possible with the alphabetical subject cat- alog. G r a n t i n g that this precision of ap- proach is desired, the various aspects of a subject can be separately presented in such a w a y that automatically those which may have no immediate value in a particular problem are excluded from a search. F o r example, " N a t u r a l i s m in literature" is a phrase subject heading separated alphabet- ically from its parent heading, " N a t u r a l - ism." In a dictionary catalog it is used to cover both general works in P N 5 6 . R 3 and the same topic in separate national litera- tures. T h e classified catalog subject refer- ence can read: Naturalism B 828.2 Literature (use in) P N 5 6 . R 3 French PQ295.N2 Likewise a topic that appears in more than one field may be presented with its distinctions in one reference: Ions Chemistry QD547 Physics QC702-721 In a classified catalog the see reference is a direct reference to an entry such as Polar regions G 6 0 0 - 8 3 0 instead of P o l a r regions see Arctic regions; Seismography Q E 5 3 1 - 5 4 1 instead of Seismography see E a r t h - quakes, etc. If the see reference term has multiple possibilities of number references the following type of card is set u p : Spanish America F 1 4 0 1 - 1 4 1 8 For specific aspects of this subject consult the subject reference in this Catalog under Latin America. Criminology HV6001-9920 Treatises HV6025 For specific aspects of this subject consult the subject reference in this Catalog under Crime and criminals. It may be observed that in such cases the basic or more important number range is given, providing, as with a simple see ref- erence, a direct guide f o r the user. T h e subject index and its accompanying authority file are in card form. A t present the subject index is interfiled in the main catalog. T h i s serves to permit broader access to the subject headings than if they were consolidated in a separate file with a thousand or more in one drawer. Actually the first decision to use only a two-part catalog was made with the thought in mind that the change-over from a dictionary catalog should be as simple as possible. T o confront the students and faculty with a new classification, a new catalog in an un- familiar form, and three separate files of cards, would, it was believed, arouse such adverse reaction that no f a i r trial could be obtained f o r the classified catalog. I f , upon the completion of the reclassification the subject index is published by photostatic reproduction, as is planned, there can then be provided a portable index to the collec- tion. Although this index w i l l always be in process of growth, and it is expected that the first publication of it w i l l be fol- lowed by revisions, the printed copy w i l l still provide major access to the collection. I t is obvious also that the publication of this index might offer f o r the first time a relative index to the L C classification that could be used by other libraries. T o date there are approximately 1 8 , 0 0 0 cards in the two index files, and these cards vary from those with one direct number refer- ence or with a reference to a small range of numbers to entries that run to several cards and have multiple subdivision refer- ences. T h e longest reference now is f o r the United States and covers twenty-four cards and contains 1 7 2 subdivisions. These represent almost all the major subdivisions that w i l l be used, but the chronological sub- OCTOBER, 1953 419 divisions under " H i s t o r y " are excluded. T h e s e latter are put onto guide cards and the authority card to the U . S. entries re- cords these in red to distinguish them f r o m regularly represented headings or sub-head- ings. O n l y spot editing has been done on the subject index. I t is expected that some entries w i l l be reduced by w i t h d r a w i n g subdivisions or moving them onto guide cards. In general, under some topics that would usually be consulted only by advanced students, and where the possible number range is not in itself very large (such as Indo-Iranian philology), f e w e r subdivisions have been used than under topics of wider scope and interest. A t present if a subject reference card runs to a second or third card, it usually repre- sents an area that covers considerable space in the catalog. T h e r e f o r e , this is still a more rapid approach to the whole field than searching the many subdivisions in a dic- tionary catalog. T h e subject references are set up in straight alphabetical w o r d by w o r d arrange- ment to prevent any confusion in locating headings composed of more than one w o r d , or those with commas, dashes or preposi- tions. F o r example, a reference card reads as f o l l o w s : Boston University Card Chemistry Addresses, essays, l e c t u r e s , e t c . Q D 3 9 Analytic Q D 7 1 - 1 4 2 Apparatus Q D 5 3 - 5 4 Elements QD466- 467 Dictionary Catalog Form of Heading Chemistry Chemistry—address- es, essays, lectures. Chemistry—Analytic Chemical apparatus Chemical elements Although the last two entries above ap- pear on the chemistry reference card in a variant of the official f o r m they w i l l also appear in the index in the L C f o r m , being, in effect, see references in the University's index. T h e r e is one other f o r m of reference that has required a local decision and that is f o r a criticism of a w o r k that is not provided with a separate number in the L C classi- fication. A critical study of Jude the Ob- scure, f o r example, presents no problems f o r the classified catalog as T a b l e X I in the P N - P S schedules can be applied. A study of Keynes' " E c o n o m i c Consequences of the P e a c e " by Etienne M a n t o u x , however, re- quires a local decision as the L C number H C 5 7 cannot be used as a direct reference to this criticism as it also applies to the Reconstruction Period 1 9 1 9 / 3 9 as a whole. F o r this, the f o l l o w i n g f o r m reference is provided: Keynes, lohn Maynard, 1 8 8 3 — E C O N O M I C C O N S E Q U E N C E S O F T H E P E A C E F o r material on this subject consult the Classified Catalog H C 5 7 . F 4 6 1 Mantoux, Etienne. If a second criticism is received a second call number is added to the card. T h i s is one instance when the index may appear to be somewhat cumbersome to a user. T h e r e is a considerable amount of re- typing of reference cards as the reclassified materials join new books already in a subject area. T w e l v e hours of typing and two hours of revision are required each week to keep the additions to the index up to date. T h i s w o r k is done by a student typist and revised by a junior cataloger. T h e typing cost is from sixty-five to seventy-five cents per hour and the revision cost is f r o m one dollar to one dollar and twenty-five cents per hour. W h e n this experiment w a s started the administrator of the project stated that professional cataloging costs would be con- siderably higher than f o r a traditional re- 1 The Library uses Biscoe time numbers in place of Cutter Author numbers. 420 • COLLEGE AND RESEARCH LIBRARIES cataloging program. In general this has held true. M a k i n g the necessary relative index with its attendant difficulties of rec- onciliation and delimitation of class numbers with subject terminology requires more time than preparing an alphabetical subject au- thority file from L C headings. T h e assign- ment of class numbers to represent each added entry takes, of course, more time and discriminating judgment than checking a printed heading. In those cases where L C cards are not available the time of subject cataloging is doubled by the necessity of establishing the subject headings and then assigning the class numbers f o r each. Con- siderable experience with the L C system is required to assure accurate interpretation of it. Although the neophyte in library school is told that a subject heading and a classi- fication number are not necessarily synony- mous it is true that the latitude of variation is narrower in a classified catalog. T h e r e - fore, each classification number used must be related to some subject heading or to some aspect of that heading and that indica- tion must appear on the subject reference card. Fortunately, from the viewpoint of progress, the major part of the collection has been covered by L C cards, although about 1 5 0 titles per month require local classification. Both new and reclassified books that are covered by L C card entries have classification numbers assigned by the catalogers without close checking (but with certain omissions, such as the use of form numbers in the H and J classes). T h e catalogers also classify new books without the L C entries and they are responsible f o r the accuracy of all descriptive catalog in- formation.2 T h e cards are then sent to receive classified catalog entries and refer- ences. In rare instances it is necessary to 2 A junior cataloger without Library School training does all descriptive cataloging for material not covered by L C entries. It may be subject to revision by the senior cataloger, if necessary. recall the book to determine the aspect of a subject it represents. T h i s occurs usually only in assigning secondary subject entries. Generally the added subject entries can be established f r o m the information at hand. If a text is classified as a treatise on chemis- try and has a second subject entry f o r tech- nical chemistry it is reasonable to conclude it can have an added classification number f o r a treatise on technical chemistry, or if Missons New Voyage to Italy is c l a s s e d in Italian travel of the 1 7 t h and 18th century and the added entry is f o r Europe—Descrip- tion and travel, it seems safe to assume that D 9 1 7 ( 1 7 t h century travel f o r E u r o p e ) is correct inasmuch as the author died early in the 1 7 0 0 ' s and the first edition was pub- lished in 1 6 9 5 . 3 A l l cards come to the classifier with a w o r k slip. Except f o r reclassified material this is one of a seven-part multiple form pre- pared by the O r d e r department. T h i s slip is then annotated to give the typist the in- formation she needs f o r added entries. T h e s e are indicated on the lower l e f t side of the w o r k slip. A l l added subject entries appear as "parenthesis numbers" above the original call number when typed. T h e y are traced on the shelf-list card (all tracings are on the shelf-list) in the lower l e f t cor- ner as ( J K 4 2 1 ) , ( H D 9 9 9 . 0 8 ) , etc. T h e lower right portion of the work slip is re- served f o r indicating if any new subject reference cards are needed and a notation such as " ( 1 ) " means " r e f e r to subject heading number one on the printed card and, using it, prepare a subject reference to the number indicated." A s the recataloging has advanced, very often it is not a ques- tion of making a new heading but of adding to a heading already made. T h e classifier, using a separate 3 x 5 card, indicates this addition in a pencilled note. F o r example, the first book on Radio broadcasting re- 3 Information on Library of Congress card. OCTOBER, 1953 40 7 quires an addition to the entry under Radio, so the pencilled card reads: Radio Broadcasting P N 1 9 9 1 - 1 9 9 1 . 9 W h e n the box of cards that has come to the classifier is completed the additions destined to go on classified catalog refer- ence cards are placed in a group at the end of the box. W h e n a card typist completes a box she not only w i l l have typed each set of cards but she w i l l have made all the new subject references required. A student typist then receives the group of additions and proceeds to w o r k on them, retyping principal entries when she cannot place the additions in proper alphabetical position. F r o m September, 1 9 4 8 , to J u l y , 1 9 5 2 , 44,676 titles4 were processed, including both new and reclassified materials. T h i s in- cludes 5002 titles f o r volumes outside the central unit. T h e s e latter receive f e w e r sub- ject entries (usually only one) than do titles in the main collection, which means that any statistics on the proportion of classified catalog entries to the titles in the total col- lection are not wholly representative. T h a t is, about 1 . 5 cards per title are prepared f o r the classified catalog, but this figure would be slightly higher if a separate count were kept for the Chenery collection. One per- son working approximately half time has been able to carry the w o r k . T h e main- tenance of a distinct division between pro- fessional and clerical activities in all the work of the department is, of course, to be expected. A s in most libraries there have been periods where this distinction has been virtually nonexistent. W i t h rare excep- tion, however, the classifier has been able to devote the equivalent of half-time w o r k to the classified catalog. T h i s responsibility in- cludes the assigning of the added subject 4 Excluding pamphlets that are selectively cataloged with no subject entries and are primarily for reserve room use. entries with their proper references and the checking of tracings f o r all other added entries. Decision on series added entries w i l l be transferred to a union serials cata- loger if this position is established. A t pres- ent new and reclassified serials and series are distributed among the catalogers in accordance with their separate areas of re- sponsibility. I t is not possible to isolate all the costs of establishing and maintaining the classified catalog as it has always been a part of the total project of reclassification. It is, how- ever, possible to separate and estimate cer- tain parts of its cost. T h e number of cata- logers and clerical staff assigned to the recataloging program has varied, but the work on the classified catalog has been main- tained by one classifier quite consistently on a half-time basis f o r the forty-six months from September, 1 9 4 8 , to J u l y , 1 9 5 2 . A n average of 9 7 1 . 2 titles per month have re- ceived classified cataloging on the half-time basis. T h i s would, it is believed, compare favorably with the w o r k of a subject cata- loger, but it would not equal in quantity the cataloging possible f o r a dictionary sub- ject catalog under similar conditions. If the fifty-six hours of clerical w o r k required to carry the additions to the index are esti- mated at the maximum w a g e levels in effect at present, the monthly costs are $ 4 i . 6 o . 5 If 9 7 1 . 2 titles are prepared per month, the unit cost is then $ 0 , 0 4 2 . T h i s does not in- clude the cost of new references that are made by the catalog card typists. T h i s operation is not separated from their w o r k but a figure of .05 per card has been es- tablished as typing costs f o r regular catalog cards. N e w reference cards require less typing time, therefore it seems possible that the actual cost of a new reference is the same as that of preparing an addition. 5 That is, 75^ per hour for typing and $ 1 . 2 0 for revision. 422 • COLLEGE AND RESEARCH LIBRARIES T h e professional labor costs per title could be established f o r a similar situation by taking the prevailing salary scale f o r ex- perienced personnel and balancing it against estimated output. It should be expected that the average number of titles classified in the first f e w months w i l l , of necessity, be lower than the average over a long period of time when experience in the methods begins to take effect. D u r i n g the forty-six months period the whole project has had the equivalent of one cataloger f u l l time for 1 6 5 months includ- ing all vacation allowances. T h e total proj- ect has i n c l u d e d : — ( 1 ) Reclassification and recataloging of the Chenery Library collection (not yet completed) (2) Cataloging and classification (with no responsibility for card typing or prepa- ration of materials in the branch unit) for the two colleges and one bureau collection.6 ( 3 ) New library collection established for a small college with all technical services completed in the central unit. F o r the purpose of estimating the total expenditures the full-time service of the classifier is included above as the other half of that position has included administrative responsibility for the reorganization pro- gram as well as certain responsibility f o r personnel and other activities f o r the central library. A total of 1 1 4 , 7 6 0 volumes have been processed in this period, averaging 695.5 volumes per person per month. One of the expected economies of this type of catalog is the rapidity with which subject headings can be changed when a new term supersedes an older form. Less anticipated is the very real economy in num- ber of cards needed. T h e dictionary subject catalog's use of overlapping terms results in multiple entries that can be avoided in a 6 Three collections now having centralized processing joined the program after the two on which the figures in this article are based. classified catalog. F o r example, the ma- terial classified in P N 4 0 0 1 - P N 4 3 2 1 very often has both Oratory and Elocution as subject headings on L C cards. In another field, Still's Underground Railroad has three subject headings assigned by the L C . 1 . Underground railroad; 2. Slavery in the U . S . — F u g i t i v e slaves and 3. Slavery in the U . S . — A n t i - s l a v e r y movements. T h e book is classed in E 4 5 0 which applies directly to subject headings number one and two. T h e third subject heading is covered adequately by the classified catalog refer- ence to E 4 4 1 - 4 5 3 . T h e r e f o r e f o r this volume one card only is needed in the classified catalog to represent three ap- proaches. T h e s e are examples of the numerous headings that so overlap that the dictionary catalog user distinguishes very little value in their separation and that the classified catalog user quite happily finds in one place. T h e precision of the L C classification schedules allows the reverse of this situation to appear less frequently than would be the case with a broader classifica- tion system. T h a t is, rarely is one L C number used for two subjects requiring two subject references that may result in con- fusing a reader coming from one of the references and finding what may seem to him irrelevant material. W i t h a consideration of the cost of a classified catalog should go some statement of its value to the user in comparison with a traditional subject catalog. N o quality statistics are available f o r either type at present. Some definite observations may be made, however. Boston University's cata- log has been in public use since its begin- ning. F o r a major part of the four year period students and faculty have consulted it without assistance. 1 9 5 1 was the first year a freshman orientation program w a s offered and the same year a serious attempt was made to provide a readers' service desk OCTOBER, 1953 40 7 in the catalog area during peak periods of use. Encouraging results have, of course, been observed and both of these activities are now permanent parts of the library service program. B u t , it has been proved that undergraduates, graduates and faculty, with no library staff help are able to use the relative index and go f r o m it to the classified catalog and locate material. T h e Reference department is most en- thusiastic in the results of its use of the catalog. Staff members have commented upon the extra help they receive f r o m the subject references. T h e subdivisions alpha- betically placed on the cards often serve to narrow their search to a specific number or may draw to their attention the very aspect that they need. T h e consistent use of the subdivision general works under every major subject area provides a sure guide to broad treatment suitable f o r the under- graduate w r i t i n g a paper or f o r the spe- cialist f r o m another field w h o is intent on getting some basic material in a subject re- mote f r o m his own activities. Several minor experiments in arrange- ment of the catalog have been tried out. A t first it w a s proposed to have selective subject representation in the M a i n catalog. T h e difficulty of arriving at an acceptable choice of this kind prevented development of the plan. Inverse chronological filing was used until the spring of 1 9 5 1 . I t was abandoned as it seemed to cause confusion and served no readily apparent need.7 T h e filing is now alphabetical by main entry un- der each classification number. G u i d e cards are used extensively and contain both the class number and the subject term. Mention was made earlier in this paper to 7 The shelf list, because of the use of a time number as a book number, provides a chronological survey of the major holdings in any field. the method of treatment f o r representation in the classified catalog of material in col- lections of other colleges that may, f o r ex- ample, use the Decimal classification system. T w o types of entry are possible. F o r a collection which in no probability w i l l ever be reclassified, the main entry is in the form of a parenthesis L C number above the loca- tion symbol and D e w e y call number. If f o r some reason it seems advisable to delay decision on possible reclassification (as, f o r example, the School of N u r s i n g L i b r a r y may wish to consider the new R schedule f o r adoption), the parenthesis number is again used above the location symbol but no D e w e y call number is given. W h e n cards f o r either type of material are sent to the classifier, if they are printed cards the suggested L C number is placed in proper position on the w o r k slip. T h e classifier then verifies and checks it for needed refer- ences and decides if any added subject en- tries are to be used. In general it seems f a i r to say that as the total project of reorganization has pro- gressed it has been possible to set up this type of subject control without unduly higher costs in professional and clerical labor. H o w much f u r t h e r along the recat- aloging project would be had a dictionary subject catalog been set up cannot be esti- mated. T h e only comparison possible is the output per month of this project balanced against a similar recataloging program in a library with experienced personnel and ap- proximately the same percentage of titles having L C copy available. T h e really important question—Does this method of subject control offer sufficient advantage to the users to w a r r a n t its use? — c a n only be answered subjectively at pres- ent. W e believe it does. 424 • COLLEGE AND RESEARCH LIBRARIES