College and Research Libraries " RITV ARS BREGZIS The Ontario New Universities Library Project-an Automated . Bibliographic Data Control System ONULP was established in 1963 to develop and maintain fiv e 35,000- volume academic libraries, and automated bibliographic control was planned from the beginning. Complete bibliographic information is keyed into a master record, which is then manipulated, selected, and reformatted for printout as shelf-list cards, author-title book catalogs, and subject book catalogs. Other bibliographic services can now be conducted as "by-products," including cooperative, analytical, con- trol, and other activities. THE IMMEDIATE OBJECTIVE OF THE ONTARIO NEw UNIVERSITIES LIBRARY PROJECT IN THE LATE SUMMER OF 1963 the Uni- versity of Toronto library received from the government of the Province of On- tario a request to compile by 1967 five 35,000-volume basic college library col- lections along with the necessary cata- logs for five new Ontario universities and colleges, some of which had to enroll their first classes in September 1964. The task appeared to be a dual one. In the first place, it was necessary to select, ac- quire, and process approximately $35,- 000 worth of books monthly; for this pur- pose a special project group had to be es- tablished. Second, it required a system of catalog production that could accom- modate this acquisition rate as well as keep currently updated five sets of cata- logs in five different institutions. In October 1963 the Ontario New Universities Library Project ( ONULP) was established with its own budget which included $1,300,000 for the pur- M'f. Bregzis is Assistant Librarian in the University of Toronto library. chase of books over a period of three and one-half years. The administrative or- ganization of the project was set up in the form of the customary three basic operations: selection, acquisition, and cataloging. The selection of the materials was arranged to be done py the under- graduate book selection office of the University of Toronto library with added project staff. For the acquisition and processing of the selected materials two sections were established under the tech- nical services of the University of To- ronto library. Since the processing of the ONULP materials was to be done as a separate project not integrated with the acquisi- tion and cataloging of the materials of the University of Toronto library, an op- portunity presented itself to choose the most suitable method for compiling the required catalogs. The expert advice of the Institute of Computer Science of the University and the readiness of the IBM Toronto Data Centre to develop the machine system made it possible to proceed with work immediately. The immediate problem for the 1 495 496 I College & Research Libraries • November, 1965 ONULP participants was to have func- tional catalogs which could not only list the receipts from the project promptly and accurately but also integrate mate- rials acquired by the participants them- selves outside of the project. The ex- perience of the University of Toronto library with multiple catalog mainte- nance indicated that these requirements could not be met economically by cus- tomary card catalogs. Since to meet this situation the need for book form catalogs was indicated, it became apparent that electronic data processing methods could be used advantageously for their com- pilation as well as for the bibliographic control of the library collections of the five new institutions. THE LoNG TERM OBJECTIVES AND THE PROBLEM OF BIBLIOGRAPHIC CONTROL Bibliographic control of a library col- lection essentially is control of informa- tion macro-units and includes the whole range of organization of bibliographic data for various conceivable purposes. The bibliographic information custom- arily given on catalog cards constitutes only part of the information that a li- brary requires for its services. The real long-term need therefore is for the li- brary to develop a bibliographic control system which can supply not only the customary bibliographic tools, i.e., cat- alogs, but which can be used also to ob- tain quickly, reliably, and economically any type and selection of information pertaining to any part of the total col- lection of the library. Since electronic data processing methods permit such a total and flexible approach, and since such an approach appeared to be in the best long-term interests of the new in- stitutions, the immediate task of solving the catalog compilation problem de- veloped into a long-term objective of bibliographic control. Bibliographic control sets a number of requirements which exceed those of customary library catalogs. First, a greater number of aspects of the biblio- graphic unit, i.e., the book, is incor- porated in the master record, rather than a selection of such aspects. Second, all data have to be explicitly recorded and unmistakably identified rather than left to individual interpretation. Third, or- ganization of data has to be rigidly sys- tematic rather than aimed to serve im- mediate convenience. Moreover, the form of recording this information has to have flexibility and compatibility which can serve a variety of present and future uses, some of which will emerge as de- sirable only following further technolog- ical development in information proc- essing. The master record-a bibliographic in- formation file in machine-readable form -therefore has to accommodate a num- ber of services, some of which would be developed at a later date. A printout record of the collection, or shelf list, is the most essential of the required con- trol records. Catalogs arranged by au- thor, title, or subject and possibly classi- fication are the most immediately re- quired output forms to be generated from the machine-readable master rec- ord. Other services include specialized lists of books and reading lists by sub- ject, language, etc.; periodic information to teaching staff regarding recent acqui- sitions in their respective fields of inter- est; records for automated circulation control; provision for tying in special subsystems, e.g., serials control; biblio- graphic data transmission to and from other institutions; financial and acquisi- tion records; up-to-date working tools, e.g., subject lists; and statistical data re- garding contents, patterns of use, and growth and size of the library collection. THE MASTER RECORD-A BIBLIOGRAPIDC INFORMATION FILE IN MACIDNE- READABLE FoRM The format of the master biblio- graphic information file is designed to facilitate the developing of the services enumerated above. In addition, since the ONULP bibliographic control system .r Ontario New Universities Library Project I 497 Cflt.L NuM f?£R 5 C uTTER Nas l~ c u~ss scfY'UMBti{ I 2 sll.. ~ 1111111 Ill FIG. 1-0NULP DATA SHEET- TITLE RECORD WORKSHEET ~ Sil.~ F ~ f'lc>. H Vm e. I~ tOfT/OW T ~ ()f" G t1 fVOLS. T H 1111 11111111111 Ill I I ORIGIN C! c 1.. $P£C. T I A )ItT£ ~SPR7 11. T N y y G * II I I I Ill I Ill It- ~~oc. ;; OltiTE • nr c c CAT date: RI!."'V: date: CLAS: REV: date: c ~ I~ o 5£.Q R :s ~ No. , S T Ia fio~ I Ill JiLl Ill lwl'" t .,... ~ ~~~~ ~ ~ '::::~ N 1~:;. ~ ~~ I~ ~ ~ ~;;: ~ .!:; ~ e: ~ ~ ~ ~ ~~~~ ~ ~~ .~ II ! J I I I I I I l I I I I I I !1 I I U l I I I I I 1 I I I 1 I l I r a 11 U l I I I T I I l I I 1 l I I I ~I l l I'Nj"lrl .. ~' ,. .. 0' -:-a: ..... ~~ ~:i. ~~ .~,~ ~~ ~~ ~: 0' (V(\1 """' "" ~jp'{ coN co?'l rPr ( ,.p~ ('oPI COt~r· Ceo PI roPI ( rY~ 01>'1 CoP~ Coh cor~ CoPt (rrv ( C' tJY' ,.JL ~~L. "'.jL N L N L NL .¥ L I K L ~~L ¥1~ N' l4 ~ ~ ~ .t L :. ~. ~ :~ N ~ ~ K -v. ~ 0.~ a. 8. 0 0 0 . g_ 0 . g. 0 0 o. ~ fl. ~ 0 0 o. g. 0. ~ 0 d . , 0 0. ~ "·: ~ . . C. r!. C . · cr I I I . I I I I I ..,101/1 -II ,.), lr. a~ N~ ~~ ilN 1\lFru ~ ... ~~ ~ ~ ~3: t.~ ~:: ~~ ·~ ~ .-. 0~ ~ ~~ ~~ 498 1 College & Research Libraries • November, 1965 serves also as a pilot system for the study of the University of Toronto central li- brary's automate4 bibliographic control requirements, extra features are incor- porated which will permit testing of the present forms and methods of biblio- graphic expression, notably the system of catalog entry and that of subject terminology. The master record is compiled by key- punching the bibliographic data into Hollerith cards and storing the informa- tion on magnetic tape. The data format of this master record, which is down- ward compatible with more general or less selective formats of properly identi- fied bibliographic data, provides for a maximum length of fifteen hundred char- acters for each bibliographic 'unit (title) which is subdivided in variable length fields for the various categories of biblio- graphic information, which in turn sub- divide into elements within a category.1 Each bibliographic unit, category of the unit, and element of the category is in- dividually addressable. This arrange- ment allows for the various bibliographic data components to be processed and ar- ranged in any desired combination. The Master Record consists of a title record tape file which contains the com- plete information pertaining to the in- dividual titles recorded, the name au- thority tape file, and the subject author- ity tape file. The authority files contain all names and subject terms that are necessary to cover the records for all titles as well as all references and history statements pertaining to the names in the name authority file and all references and scope notes pertaining to the subject terms in the subject authority file. 1 This maximum was established as a multiple of the sum of the lengths of bibliographic data cate- gories and elements as established in several surveys. The survey conducted by the University of Toronto library indicates somewhat higher character counts than the Columbia, Harvard, and Yale medical li- braries' survey. Cf. Peter M . Sprenkle and Frederick G. Kilgour, "A Quantitative Study of Characters on Biomedical Catalogue Cards-a Preliminary Investiga- tion ," American Documentation, XIV (July 1963 ), 202-206 . Fixed length fields are allotted for data cate- gories and elements that by their nature are not likely to v ary. The data in the master record are identified by a system of codes which serve as addresses of the data units, cate- gories, and elements, as well as provide the structural organization of these com- ponents. This structural code system per- mits addressing bibliographical data on any desired level of specificity. For the compilation of the master record a specially designed data sheet is used which reflects the structure of the data format of this record (Fig. 1). The following categories of data constitute the record for each title: 1. Call num her of the title. This cate- gory is further subdivided in the fol- lowing elements: a) main class. b ) subclass. c) class number. d) first Cutter number. e) second Cutter number. f) date. 2. Form of publication. 3. Brief edition statement. 4. Edition code. 5. Number of volumes. 6. Size. This category consists of two elements: a) height of the book. b) thickness of the book. 7. Origin of the publication. This cate- gory is further subdivided in four elements: a) country of origin. b) city of origin. c) language of the publication. d) date of the publication. 8. Any special aspect other than those enumerated. 9. Paging and illustration statement. The two elements of this category can be separated if desired. 10. Principal author (main entry). 11. Conventional title. 12. Title. 13. Full length edition statement when different from 3, above. 14. Imprint. 15. Series note. 16. Other notes (except Contents). Ontario New Universities Library Project 1 499 17. Secondary bibliographic approaches (tracings). This category is subdi- vided in five elements as follows: a) subject term. This element is further subdivided in the fol- lowing forms: i) name as subject term. ii) topical subject term. iii) descriptor term. b ) added entry ( other than title and series). c) title added entry. d) alternate title added entry. e) series added entry. Up to ten approaches of each group can be assigned to one title record. 18. Contents note. This category can be used in a systematic organization mode indicating the numbers or vol- umes of the contents. If so identified, the individual parts of the contents function as analytical approaches to the larger work. 19. Copy and location. This category consists of two elements: a) copy number of the title. b) location of the copy. All data categories and elements, except for 1-9, are of variable length. The data format of the name authority file and the subject authority file is com- parably structured. During the required processing operations access to and utili- zation of the references, and history and scope notes recorded in these files, is performed automatically, based on a comparison between the names and sub- jects in the title record file and the au- thority files. The information recorded on the data sheet is key punched using one Hollerith card for each line of data on the data sheet. All data are entered in the master record in the form of character codes designed to provide output in upper and lower case characters modified by diacritical marks where required. For this purpose an 026 keypunch with modi- fied A-2 keyboard is used. The key- punched information is pedodically con- verted to magnetic tape on a 8 K IBM PN6080. K5 1958 card 1 King, William Francis Henry, 1843-1909, ed. Classical and foreign quotations; a poly- glot dictionary of historical and literary quotations, proverbs and popular sayings, compiled and edited, with translations and indexes. New York, F. Ungar [1958?] lxviii, 412 p . 0 SEE NEXT CARD PN6080. K5 1958 card 2 1. Quotations I. Title ERIN 1 0 00404 FIG. 2-SHELF LIST RECORD CARDS JX1730. A4 E15 Eayrs, James Georges, 1926- The art of the possible; government and foreign policy in Canada. [Toronto] University of Toronto Press [1961] viii, 232 p. Includes bibliography. 1. Canada-Foreign relations administration I. Title BROC ERIN GLPH SCAR TREN FIG. 3a-AuTHOR-TITLE CATALOG LISTING -MAIN ENTRY RECORD 1401 for processing on the IBM 7094 model II. The 7094 processing consists of extensive editing of the various codes, fields, and record lengths and of sorting the entire file in call number order. All required information in the desired out- put form is derived from this master record. THE CATALOG CoMPILATION SYSTEM At the present time the master record is used for the generati0n of shelflist cards and codex form catalogs of the project which were the immediately re- quired tools for the ONULP biblio- graphic control system. 1;'he automated catalog compilation 500 1 College & Research Libraries • November, 1965 JX1730. A4 E15 The art of the possible Eayrs, James Georges, 1926- The art of the possible 1961 viii, 232 p. BROC ERIN GLPH SCAR TREN. FIG . 3b-AuTHOR-TITLE CATALOG LISTING -TITLE ADDED ENTRY RECORD ML105. B16 1958 Slonirnsky, Nicolas, 1894- ed. Baker, Theodore, 1851-1934 Biographical dictionary of musicians. 5th ed. 1958 XV, 1855 p. BROC ERIN GLPH SCAR TREN FIG. 3c-AUTHOR-TITLE CATALOG LISTING -NAME ADDED ENTRY RECORD QL799 .. R78 Bibliotheca biotheoretica, v. 9 Russell, Edward Stuart, 1887- The diversity of animals 1962 viii, 151 p. BROC ERIN GLPH SCAR TREN FIG. 3d-AUTHOR-TITLE CATALOG LISTING -SERIES ADDED ENTRY RECORD Austria-History-1867-1918 Jaszi, Oszkar, 1875-1957 DB91. J3 1961 The dissolution of the Habsburg Mon- archy 1961 xiv, 482 p . BROC ERIN GLPH SCAR TREN DB85. M35 May, Arthur James, 1899- The Hapsburg Monarchy, 1867-1914 1960 X, 532 p. BROC ERIN GLPH SCAR TREN FIG. 4-SUBJECT CATALOG LISTING -SUBJECT ENTRY RECORDS system derives its effectiveness from the systematic structure of the data format of the master record, from which there- quired bibliographic information is se- lected and reformatted according to the formats of the shelflist cards (Fig. 2), the records in the author-title catalog (Fig. 3), and the subject catalog (Fig. 4). For the shelflist record "full catalog information" as defined by the ALA rules of entry and the Library of Congress rules for descriptive cataloging, is select- ed from the master record. Added to this information are also copy numbers and location designations. Similarly "full cat- alog information" is selected from the master record for the generation of the main entry record2 to be included in the author-title catalog. Information used for all secondary en- try records, however, is more selective, and certain categories of data are dis- played in abbreviated form. The second- ary entry records, both added entry records for the author-title catalog and subject entry records for the subject cat- alog contain only the call number, the secondary entry, the main entry, ab- breviated title, brief edition statement, date of publication, and paging, as well as location information. The abbrevia- tion of all secondary entry records is performed automatically by the com- puter. The only instances in this format- ting where manual guidance is exercised are the insertion of a special character on the data sheet indicating the end of the brief title and the separation of paging from the illustration statement. Since the master record tape file is in call number sequence, the shelHist rec- ords when produced are in proper filing sequence without any further rearrange- ment. Both the author-title and subject entry records, however, must be sorted in alphabetical sequence. 2 In this paper a distinction between the terms "entry" and "entry record" is made. "Entry" refers only to the name (personal or corporate), title, or subject term under which a work is entered. "Entry record" refers to the whole record of a work dis- played in a catalog; usually this term is qualified as main entry record, added entry record, or subject entry record. Ontario New Universities Library Project 1 501 The filing arrangement of entry rec- ords in the catalog is designed to be per- formed automatically by the programed system without any attempt to regulate it at the time of preparing the biblio- graphic information input. It is assumed that the arrangement of entry records in the catalog depends more on the relative characteristics between these records than on the specific characteristics of the individual record. Generally the filing order attempts to adhere to the customary sequencing of alphabetic information. The basic order of sequencing in this system is: blank, period, dash, comma, alphabetic char- acters A to .Z, and numeric characters 0 to 9. This sequencing system is applied to filing of bibliographic units according to a computer-constructed filing field which consists of a specified number of characters from the key categories of the bibliographic unit. The selection of cate- gories and number of characters in each category was based on sampling and analysis of appropriate sections of the University of Toronto library catalogs, and the selected lengths vary with the type of the entry record. In applying this filing formula, the sys- tem disregards nine qualifying terms at the end of the entry. Another modifica- tion of principal importance is the disre- garding of initial articles in all main and secondary title entries. A comprehensive list of articles in twenty-seven languages is used. Omission of an initial article found in this list is activated by the cor- responding language code. Modest as these modifications of pure character-by- character filing are, they are impressively costly in terms of programing and op- erational factors. The results of this filing arrangement are not expected to produce an arrange- ment identical to that specified by any of the customary filing rules. It is con- ceded that the customary semi-syste- matic filing arrangement cannot be justi- fied by the complexity of the required programs and operational costs that would be involved in its attainment. Moreover, it is not certain that the pub- lic would find the customary filing se- quences more convenient than consider- ably more standardized alphabetic ar- rangements. Further adjustments of the present filing formulas are expected to be necessary, and prolonged use of the ONULP catalogs should indicate the degree of validity of the basic sequenc- ing system of the bibliographic informa- tion in the catalogs. For the inclusion of references in the alphabetically-arranged title record tape file containing all main and secondary entry records, this file is matched with the name authority tape file. All refer- ences leading to names found in the title record file are set up in catalog for- TABLE OF FILING FIELDS Ontario New Universities Library Project MAxiMUM NUMBER OF CHARACTERS IN EACH FILING FmLD TITLE RECORD Title Title Field TOTAL Entry Added Subj. Author Main Date of Type Entry Entry Entry Entry Convent. Actual Public Field Field Field Field Field Title Title Field Author Main Entry 1 . . . . .... 39 . ... 7 10 3 60 Title Main Entry 1 • 0 • • . . . . • • 0 • 56 . ... 0 0 •• 3 60 Added Entry 1 39 10 • •• 0 7 3 60 ••• 0 .... Subject Entry . . . . • 0 •• 40 15 . ... 5 • • • 0 . . . . 60 Note that these fields are the maximum allotted. An actually shorter statement in any of these fields will transfer the balance to the following field, so that a total of sixty characters can be selected from the spec- ified eight filing subfields in order to determine the filing arrangement. 5021 College & Research Libraries • November, 1965 Alphabetical Sort! Alphabetical Sort II Subj. Authority \J Frc. 5-PROCESSING CHART-TITLE RECORD AND SuBJECT RECORD mat and added to the title record file (Fig. 5). Re-sorting of this file is neces- sary in order to place all references in their proper alphabetical positions. Subject entry record processing is per- formed in the same manner as processing of the author-title entry records. In ad- dition an authority list of legitimate sub- ject terms with references to these terms is constructed for the convenience of the processing staff. The last stage of the Cobol program for the 7094 processing cycle consists of setting up the output format according to the output type: card format for the shelflist records, and three-column page Ontario New Universities Library Project 1 503 for the author-title catalog and the sub- ject catalog (Fig. 6). This is done in the form of a 1401 tape output listing. This listing is performed on a 1403 model II printer equipped and programed to han- dle upper and lower case characters and diacritical marks. Each line containing diacritical marks must be printed twice using the space suppress feature of the printer. The printed catalog page containing 110 lines of 132 characters each ( 105 net lines of 126 net characters) and measur- ing 14 X 18% inches is too large for con- venient use. Photographic reduction to 60 per cent of the original format gives good readability and convenient size ( 9 X 12 inches). This format accommodates bibliographic information covering thir- teen titles on one page of the author- title catalog and seventeen titles per page in the subject catalog. This includes all required secondary approaches (sec- l•· Title: The ondary entries, references, and scope notes). An unlimited number of copies of the catalog can be obtained by offset printing from the photographically re- duced plate. Issue of the book form catalogs fol- lows a cumulative pattern. Monthly is- sues containing information added to the master record during the current month are prepared for each of the first two months of each quarter of the year. Quarterly cumulations are produced for the third and ninth month of the year, semi-annual cumulation for the first half of the year, and a total cumulation of the entire information store at the end of the year. This pattern represents the most acceptable ratio between the print- ing cost (i.e. , cost of reprinting the in- formation in successive cumulations) and the convenience of finding any given item by searching in the minimum num- ber of catalog issues (Fig. 7). FIG. 6-CATALOG: AUTHOR-TITLE Allemand, Maurice, 1872 - e B5lzac, Honore de, 1799 - 185 CAR TREN Le lys dans Ia vallee 1 9 63 441 P• QA471. S46 BROC ERIN GLPH SCAR etry 1963 PQ2167. P5 1964 BROC Allemand, Maurice, 1872- ed. Balzac, Honore de, 1799-1850 CAR TREN La peau de chagrin 1964 xxxvl, ~61 P• SCAR TREN PQ2369. A14 1962 z v. SCAR TREN Allen, Glover Morrill Bats. New York, ( 1962) x, 368 p. lllus. Bibliography: p. 1. Bats BRDC ERIN Allen, Glover Morrill Birds and their York, Dov e r Pub I ica x, J3B p. iII us. Inclu d es bibllo g 1. Birds I. Tit 504 I College & Research Libraries • November, 1965 ADDITIONAL BmuocRAPmc SERVICES The master file of bibliographic data is capable of providing other much needed services to the library besides the pro- duction of catalogs. There are four groups of such functions which can be based on or derived from the ONULP master record of biliographic data. First, the master record is designed to support a number of service functions and records. The individual institutions can automatically produce book cards and other circulation records for all their materials which are recorded in the mas- ter file, since the combination of individ- ually identified classification number, form of the publication, copy number, volume number, and location designa- tion can produce unique identification for each physical unit in the library's col- lection. 7 6 .-o i 5 0 u ~ ~ .8 U'l ~ ~ 4 1-< ........ 0 ~ E z ~ 3 00 ce t > -< 2 1 I I 0 No t cumulated Listing of library materials can be produced according to any characteristic or combination of characteristics, such as subject of any desired degree of specific- ity as provided by the Library of Con- gress classification system, the form of publication, the city or country of origin or language of the publication, the year, decade, or century of publication, or the location of the work in the library sys- tem. The systematically recorded classifica- tion numbers are planned to be used for current awareness services of biblio- graphic information to the teaching staff Third-yearl y-curr ent Q Quart erl y-c urre nt \8 Third-yea rl y- Bi-monthl y & sem i- to da te annuall y-cmren_t 0 Qu art erl y- to el a te Quar terl y & semi -· annuall y-c urrent ' "- Bi-monthl y-to dat e 0 ~ Monthly-to elate 0-- 10 20 30 40 50 60 70 80 90 100 T otal Numh er of C umul ati on Units to be Printed Durin g th e Year Frc. 7 -CUMULATION EFFECT AND PRINTING CosT > Ontario New Universities Library Project 1 505 SUBJECT, OA AUTHOR AEF"EAENCE "'U CLASS CA~~UT;E~:;M;E:S 1:1 EOITION 1§1 ~· I SIZE-r ORIGIN f !~~6; 11:1 PROC. a§ S!~~~;~E § ~~~~ NUM~ER l 1 I 2 l DATE _[~~ ~~~ VDLS.IHG!l THJcm ~CITY I LANG l DATE l ~ lgl DATE OOGOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO 1234567B9nnuuMftftoaaw~un~~unn~•n»»~~nnu~w~~~«~«~uu~~~~~~~~~~~~~~"~"vuurnnnn~~nnnn~ 11111111111111111111111111111111111111111111111111111111111111111111111111111111 PA GES OR VOLUM ES I I ILLUS STATEMEN~T - . 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 ~JjjJ_ 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 4 4 4 4 4 4 4 4. 4 .u.i.l.5 5 5 5 5 5 5 5 5 5 s 5 5 5 5 s 5 s 5 s n L 5 5 5 5 5 5 5.s 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 .5 5 5 5 5 5 5 5 5 5 5 5 5 s -=-:r AUTHOR TITLE I I CONTENTS 6 6 6 6 6 6 6 6 6 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 77 7 77 7 7 7 7 7 77 7 77 77 77 77 7 7 77 77 77 77 7 77 77 7 77 77 7 77 77 7 7 7 7 7 77 7 77 CO PY I COPY I COPY I COPY I COPY I -~,;;. I COPY I COPY I COPY I COPY I COPY I COPY I COPY I COPY I COPY I COPY I NO I Loci IIO.,Loc l NO.ILocj NO JLocl NO.jcocl NO. ,LOcl NO. ,Loc l NO.,LOcl NO. ,LOcl NQ I Loci NQ I LOCI NQILocl NO.,LOc.j NO.,Locl NO.,Locl NQ I LOCI NQ I LOCI 8 8 8 8 8 8 8 8 8 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 ,9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 1 2 J 4 s 6 1 a 9 10 n 12 1314 15 16 o 111120 21 22232425 26 27 2129 3D 31 3233 34 35» 37 »3940 4142 43« 454647 «4950 51 5253 54 555157 ~sua" 62 ~M&511&71!69 ron 12 n ~ ~n n 1111 a FIG. 8-TITLE RECORD CARD of the colleges. The areas of interest (profiles) of the teaching staff would be expressed in several relatively general classification numbers. The system would compare these numbers with the specific classification numbers in the file of the recently-acquired materials selecting those items which have classification numbers equal to or more specific than those of the interest profile. The result- ing lists would then be mailed period- ically to the persons concerned. The compatible data format facilitates important cooperative services. The sys- tematic structure of bibliographic data in the master record permits flexible stor- age of the entire master file in a large electronic memory providing random ac- cess to this information for purposes of checking new acquisition requests against the master record, for cataloging in- formation contained in the master record, or for receiving and transmitting biblio- graphic data from and to remote loca- tions. Bibliographic reference informa- tion can be obtained from the common pool of the project. Also, information can be exchanged with other libraries using a similar compatible data format. It is ex- pected that regional union catalog ser- vices can be based on the centralized store of bibliographic data. The master file can be compiled by centralized or cooperative efforts. The format of sys- tematically identified categories and ele- ments of bibliographic data permit the sharing or centralizing of bibliographic processing. This cooperative effort can be further assisted by direct communica- tion of the machine-readable biblio- graphic data from library to library. All these cooperative functions of automated bibliographic information control prom- ise unlimited possibilities of application. The third group of automated func- tions which are facilitated by the sys- tematized data format of the master record is a number of analytical ap- proaches to the. bibliographical informa- tion. The detailed control of classification number values affords systematic subject approach which can respond to the en- tire range of specificity inherent in the Library of Congress classification system. Classification numbers are recorded in the master file oriented to a fixed loca- tion. The input form of this feature is indicated on the flollerith card by an arrow (Fig. 8). Similarly, a system of form codes identifies every bibliographic unit (title) as monograph, serial, journal, microtext, and so on. Combined with other parameters of the data format, such as language or place of publication, this aspect lends itself to a variety of applica- tions of analytical bibliography. The fourth group of automated func- tions pertains to the control of the library 506 I College & Research Libraries • November, 1965 collection. The independently function- ing authority files of the master record permit complete flexibility for accom- modating alternative forms of the key bibliographic elements in the master record. All variant forms are recorded in these files under the accepted stan- dard form which alone is directly linked with the affected bibliographic data on the title record. This structure not only provides the system with an automated reference network, but it also permits compilation of various working tools in any desired format for use of the process- ing staff, such as lists of subject terms, or name authority lists. For the wide range of statistical and inventory control purposes, records of required scope and format can be pro- duced as necessary. Administrative data which derive from the format of the mas- ter record include both the information about the physical size of the library's collection and the qualitative factors connected with the growth of the collec- tion. An important factor in administrative decisions of every library director is the actual physical size of the collection and its integral parts. Recording of thickness of every item added to the collection is intended to facilitate calculations for such purposes. Moreover, the physical size of the collection combined with clas- sification data can produce vital informa- tion about the distribution of shelving arrangement and capacity at any given time. Planning of library buildings and the arrangement of materials in the li- brary can thus be based on calculations of empirical factors rather than estimates and i~tuition. Planned regulation of the growth of the library collection accord- ing to specific requirements can be facil- itated by combining the various data elements identified in the master record. ;Enlightened building of research library collections which still is the weakest aspect of library administration can be made possible by periodic analysis of the collection in terms of all bibliographically qualitative aspects, such as subject mat- ter, form of publication, periods, etc. Similarly, analyses of the use of the col- lection can easily be obtained and cor- related with other characteristics of the collection. ' CONCLUSIONS AND PROBLEMS To date the master record has been tested only for two purposes: for the production of shelflist records in card form and for the compilation of catalogs in book form. No complete evaluation of the system therefore is possible at this time. Certain general observations, how- ever, can be made. Throughout the design stage of the ONULP bibliographic control system it appeared that during its operation the project would not require handling of either excessively complex or large quantities of bibliographic information. This allowed compromises to be ac- cepted that kept the design of the sys- tem within limits of economic feasibility and immediate usefulness. No compro- mises were made, however, of a prin- cipal nature affecting the structure of the system. It was assumed that one of the functions of the project was to test the feasibility of such a system for potential- ly larger and more complex requirements which would necessitate more detailed identification of a larger number of bib- liographic elements. One of the observations which in- dicates a potential for important further development is the facility of the sys- tematic structure of the bibliographic data format which is employed in the compilation of the master record to ac- commodate communication of biblio- graphic information. Such communica- tion in digital data form may be carried out through the intermediary of punched cards, magnetic tape, teletype, or tel- ephone network for the purposes of dis- tribution of catalog information and for reference access to centralized stores of bibliographic information. J Ontario New Universities Library Project 1 507 This potential brings out the impor- tance and need for a generally accept- able compatible data format for biblio- graphic information so that not only a generally agreed minimum of appro- priately identified bibliographic data would be recorded and thus could be re- trieved by any participant of a coordi- nated system, but also so that standard- ized encoding structure could make these data readily acceptable for a vari- ety of uses by the recipient. The data format developed for the ONULP master record has purposely maintained a maximum possible similar- ity in form with customary forms of bibliographic data in order to accom- modate as much as possible formerly existing work patterns. The operation of the input system for the past year bears out this expectation. Only a ininimum of adjustment in customary work proce- dures was required; the change from the customary card layout to the ONULP data sheet (Fig. 1) did not require more than a few days. The same applies to the format of the printed catalog in which the close correspondence between the customary forms and those adopted for the ONULP catalog format greatly facil- itates the transition from a manual to an automated mode of operation. The catalog compilation system in- cludes a number of experimental fea- tures which eventually should help to evolve better methods for several dif- ficult aspects of bibliographic data proc- essing. The programed editing of biblio- graphic data which abbreviates all sec- ondary entry records in the catalog ap- pears to indicate the feasibility of auto- mated editing processes as a means of adjusting bibliographic data received from an extraneous source to fit a local data file which has certain characteristics peculiar to the individual library. The completely automated formation of all entry records for the catalogs has demonstrated the value of the unit con- cept in bibliographic data re-formatting. Formats of various required bibliograph- ic data have been obtained from one bas- ic input format, the str~cture of which is oriented towards flexibility rather than towards preselected bibliographic re- quirements. The display of the various entry records appears to indicate that in book form catalogs the customary distinction between "main" and "secondary" entries becomes less important. The structure of the master record and the ONULP cat- alog compilation system have taken over the control of relationships between the various approaches to the bibliographic unit (title). The catalog itself is freed of this control task. The reader consult- ing the catalog is primarly concerned with finding the names related to a bib- liographic item, and he is usually little interested in the relative specific weight of these names connected with the item . It seems possible that the book form catalog may bring relief to the problem of "choice of entry" in particular for publications in which several persons ot corporate entities, according to the pres- ent entry rules, aspire to the status of "main" entry resulting in a highly per- plexing confusion. The ONULP bibliographic control sys- tem is also designed to produce author analytical entry records for the items listed in the contents of any bibliograph- ic unit whenever the requirement for analytical entry records is specified. For this application the facility of data proc- essing methods is expected to provide a much needed service in an area of work which is very cumbersome and time con- suming when performed by manual methods. The printed output of the catalog compilation system is expressed in the form of one type font which contains upper case and lower case alphabetical characters, numerical characters, and the following special characters required for adequate display of bibliographic in- formation: period, comma, colo~, semi- colon, question mark, apostrophe, round brackets, square brackets, diagonal slash, 508 1 College & Research Libraries • November, 1965 underscore, plus sign, hyphen, dollar sign, percentage, asterisk, number sign, equal sign, musical flat sign, calligraphic letter "I", quotes, ampersand, and all dia- critical marks required for the Roman alphabets of all major languages. The printing of ONULP catalogs is done on an IBM 1403 model II printer for which a special print chain with the enumer- ated characters was developed jointly with the Florida Atlantic University li- brary and the Yale University medical school library. The resulting graphical display of bibliographic information leaves many things to be desired. The major non-Ro- man alphabets, several type fonts and sizes, and a sharper letter outline are re- quirements which cannot be readily met by the impact printing technique. Elec- tromagnetically-optical methods appear to be holding promise for a widened scope and enhanced appearance and clarity of the automatically produced catalog. Some of our existing bibliographic methods and philosophies restrict the ONULP master record and the catalog compilation system to a limited degree of specificity and exactitude of the bib- liographic data which can be usefully processed and communicated. This lim- itation is not insignificant for approach through name entry; it is, however, se- verely restrictive for systematic subject approach. The design and the initial operation of the ONULP system have shown that subject headings by virtue of their structure do not lend themselves to systematic organization of concepts and their terms which are used to describe the contents of bibliographic units. Pres- ently the only practically available meth- od for securing a somewhat systematic subject approach to large library collec- tions of general subject scope is through classification. This method, however, does not permit more systematic subject matter organization than our classifica- tion schemes. It is therefore clearly indicated that the most pressing long range require- ment for adequate control of biblio- graphic information is research into sys- tems of entry or an equivalent to it, and in particular into systems of organization of concepts and their terminology. Other restrictions of the 0 NULP sys- tem are purposely imposed and can be eliminated if desired. The present sys- tem is programed to handle only in- formation about book materials; minor adjustments can provide for other forms, such as micro-materials or audio mate- rials. More considerable expansion of the system is required in order to adapt it for handling large and specialized research collections: as indicated earlier, the identification of a number of categories and elements of bibliographic informa- tion has to be carried further than pro- vided in the present data format of the ONULP system. The ONULP bibliographic data con- trol system has demonstrated the fea- sibility of the application of electronic data processing techniques to biblio- graphic control. Even if information re- trieval, meaning mechanized access to factual information, is not yet feasible, mechanized access to macro-units of in- formation-bibliographic control-is well within the range of immediate practical application. The ONULP system has in- dicated that many benefits can be derived from such an application, but it has also pointed out that serious and concentrated effort has to be directed toward research into the philosophical and theoretical bases of bibliography even more than technological aspects and techniques. Before further transition from the print- ed page technology to magnetically- stored information technology expressed in detailed micro-units of information can take place, the basic structure and behavior of information generation and transformation have to be established for the application to macro-units of in- formation. ••