special libraries 65 (10)

Upload: anonymous-z2qzethyho

Post on 03-Apr-2018

237 views

Category:

Documents


1 download

TRANSCRIPT

  • 7/29/2019 Special Libraries 65 (10)

    1/68

    D E C L M I ~ E K965, Val. 56. No . 10

    Cooperation in Library Processing,Computer Input Programs,Conimunications Networks, andVocabulary Control . . . "Auto"- andAutomatic Indexing Processes

  • 7/29/2019 Special Libraries 65 (10)

    2/68

    MINI-GRAPH. . . when you need catalog cards in a hurryDuplicate as many cards as you need in your o wn

    library. Ca rds ar e automatically fe d, coun ted an dprinted a t the rate of two per second.

    The Mini-Graph, specially designed for repro-ducing catalog cards, offers unusual convenience,flexibility and economy. Simple to operate-savestime and money!

    Write for full information. Mini-Graph Duplicatorwith a starter set of stencils and i n k . . . $275.00.

    G A Y L O R D B R OS ., I N C . L I B R A R Y S U P P L I E S S Y R A C U S E , N. Y . / S T O C K T O N , C A L I F .S'ECIAL LIB RAR IES is published by Special Libraries Association, month!y Se tember to April,bnnonthly May to Argust, at 73 Mam Street, Brattleboro, Vermont O S X n . Editorla1 0g c e s : 31 East 10thE+-a.=t Npw V n ~ h e w Vnrk 1MOX Second elass nostam at Brattleboro, Vermont.

    POSTMASTER. Send Form 3579 to Special Libraries Association,3 i East 10 St., Ne w York, B.Y. 10003

  • 7/29/2019 Special Libraries 65 (10)

    3/68

    he Faraday Press announces29 major Soviet Scientific Journalsnow available for the first timeAUTHORITATIVE COVER-TO-COVER ENGLISH TRANSLATION

    REGULAR YEARLY SUBSCRIPTION -BEG INNING WITH JAN. 1965 ISSU ESCybernetics

    KibernetrkaBmonth ly, $1 S/year

    Problems of Informat ion Transmiss ionProblemy Peredachr Informatsir

    Quarterly. $100/yea1Soviet Electrical Engineering

    ElektrotekhnjkaMonthly. $160/yeat

    MagnetohydrodynamicsMagnrtnaya Grdrodrnamrka

    Quarterly. $9O/yearApplied Solar Energy

    Gel~otekhnrkaBrmonthly $1 O/year

    Theoret ical and Experimental Chemist ryTeo!et,cheskaya I Eksperrmental naya Khrmrya

    B~monthly$120/yearPolymer Mechanics

    Mekhanika Po1,merovB~monthly$120/year

    Soviet Appl ied MechanicsPrrkladnaya Mekhanjk aMonthly $160/year

    Soviet Physics JournalIzvest~yaVUZ FrzikaBrmonthly. $125/year

    AstrophysicsAstrofrzika

    Quarterly. $90/yearJournal of Appl ied Mechanicsand Technical Phys ics

    Zhurnal Pr~kladnorMekhanrkr r Tekhnrcheskor Fr z ~ k ~B~month ly. 150/year

    Soviet Progress i n Chemistry UKSSRUkrarnskrr Khimicheskrr ZhurnalMonthly, $150/year

    Moscow U nivers i ty Chemist ry Bu l let inVestnlk Mos kovs kogo Unrverstteta KhrmryaBimonthty, $110.00/year

    Dif ferent ial Equat ionsDrfferentsraf'nye Uravnen~ya

    Monthly. $150/yearSoviet Radio Engineering

    lzvestrya VUZ RadiotekhntkaBimonthly. $1 5/year

    Soviet Engineering JournalInzhenernyr Zhurna l

    B~m onth ly, 150/yearSoviet Aeronautics

    Izvest,ya VUZ Avratsionnaya TekhnikaQuarterly. $125/yearJournal of Appl ied Spectroscopy

    Zhurnal Prrkladnor Spektroskopr~Monthly, $15 0/~ earMendeleev Chemist ry Journal

    Zhurnal Vses Khrm O b-va rm MendeleevaB~monthl y, 160/year

    Combustion, Explosion, and Shock WavesNauchno-Tekhnrcheskie Problemy Gorenrya I Vzryva

    Quarterly. $lOO/yearChemistry of Heterocycl ic Compounds

    Khrmrya Geterotsrklrchesk~k hSoedrnenrrB~rnonthly, l20/year

    Chemist ry of Natural CompoundsKhrmrya Pr rrodny kh Soedrnenrl

    Bimonthly $1 lO/yearAppl ied Biochemist ry and Microbiology

    Prikladnaya B~okhrmrya MrkrobrologryaB~monthl y. 120/year

    Soviet M ater ials Sc ienceFizrko-Khrm~cheskaya ekhanrka MaterralovB~rnonthly . 1 15/year

    Soviet Mechanics Bul let inlzve st~ ya kad Nauk SSSR Mekhanrka

    B~monthly. 16O/~earM O S C O Wniversity Physics Bul let in

    Vestnrk M osko vsk ogo Unrversrtera FrzrkaBrmonthly. $1 IOIyear

    Journal of Engineering Phys icsInzhenerno-Frzrcheskri ZhurnalMonthly. $150/year

    Soviet Radiophysicslzvestiya VU Z Radlohzrka

    B~monthly. l25/yearSoviet Genet ics

    GenetrkaMonthly. $150/year

    Please add $5.00 for subscriptionsoutside U.S. and Canada.Order your 1965 subscriptions to these essential Soviet iournals from:he F a r a d a y P r e s s , I n c .84 FifthAvenue. New York. N. Y. 10011

  • 7/29/2019 Special Libraries 65 (10)

    4/68

    BRITISHMEDICAL JOURNAL

    Read the best British journal for complete coverage of research inmedical science and trends in clinical medicine throughout the world.Pzlblished weekly. Annual Subscription $21.00.

    The British Medical Journal publishes also a number of specialist journals thefollowing of which are listed below for your particulsr attention.BRITISH HEART JOURNAL. The Official Sournal of the British Cardiac Society. Coversall aspects of cardiology, including anatomy, physiology, and pathology. One of the foremostcardiological journals in the English language. Annual Subscription (six issues a year), $14.00.GUT. The Journal of the British Society of Gastroenterology. Covers the whole field of gas-troenterology-alimentary, hepatic and pancreatic disease and papers may cover the medical.surgical, radiological or pathological aspects. Annual Subscription (six issues a year), $14.00.JOURNAL OF CLINICAL PATHOLOGY. The Journal of the Association of ClinicalPathologists. Publishes spec~al ly commissioned articles on each branch of pathology wlthprominence given to its clinical application. Technical methods; notes on new or improvedapparatus; book reviews. Annual Subscription (six issues a year), $14.00. THORAX. TheOfficial Journal of the Thoracic Society. Publishes original work on diseases of the chest rele-vant anatomical and physiological studies by authors in Great Britain and many other coun-tries. Annual Subscription (six issues a year), $14.00. JOURNAL OF MEDICAL GENET-ICS. A new journal publishing articles on clinical, statistical, pathological, cytological and

    bio-chemical aspects of medical genetics. Published Quarterly. Annual Subscription $10.00.BRITISH JOURNAL OF PHARMACOLOGY AND CHEMOTHERAPY. Reports oforiginal work in all branches of the subject, including experimental chemotherapy. Biochemi-cal and pathological studies. Annual Subscription ( 3 volumes, 9 issues, in 1966), $48.00.0 BRITISH JOURNAL OF VENEREAL DISEASES fulfils the need for a journal devotedentirely to the world-wide problem of the treatment and control of venereal diseases. Its sub-ject matter is contributed by authors throughout the world. Published Quart erly . AnnualS~~hsc l - ip r i~~z10.00.

    ABSTRACTS OF WORLD MEDICINE. This journal providescritical abstracts selected from 1,500 of the world's leading medicaljournals. In addition the 'ABSTRACTS' publishes a series of ReviewArticles. Published from time to time, each will deal with one of the'growing points' of medicine.Published monthly. Annual Subscription $1 8.00.These famous journals should be included in every medical library.

    ORDER YOUR SUBSCRIPTION N O WSubscription orders only:

    BRITISH MEDICAL JOURNAL80 Brighton Avenue,Boston, Mass. 02134

    All enquiries to The Publisher,BRITISH MEDICAL JOURNAL

    B.M.A. House, Tavistock Square,London, W.C.l, England.

  • 7/29/2019 Special Libraries 65 (10)

    5/68

    SpecialLibraries

    Centralization of Library ProcessingCompatibility of Comp uter Inp ut and Programs

    Sharing Communications NetworksSharing Vocabulary Con trol

    "Autoindexing" and Indexing byAutomatic ProcessesDo-It-Y ourself R etrieval System Design

    Comparative Costs of Document Indexingand Book CatalogingGov ernm ent and Libraries

    Ad-Hoc Forum for Informat ion CenterAdministratorsN L W 1 9 6 5: T h e 1 6 8 - H o u r D ay

    I. A. Warhei tHillis L. GriffinArthur E. JonesMarjorie R. HyslopPhyllis B. BaxendaleClaire K. SchultzL. H. LinderRobert J . HavlikJoh n SherrodMrs. Elizabeth W. Owens

    Special Libraries Associat ionSustaining Member 707

    In Memoriam: M ar jor ~e aker 719 Genevieve FordNom inat ing Commit tee Report 1965-1966 729

    FeaturesMessage fro m Lilliput 714 B. Little

    Is This a Problem ? 72 6 Louise StoopsCurren t Concentrates of the Library W or ld 730

    Have You H eard 732Off the Press 736

    Editor: M A R YL. ALLISONAssistant Edi tor : EDYTHEC. PORPA Specia l Librar ies Com mit tee :Chairman: H O W A R D . B E N T L C Y

    ELLISM O U N TMRS.ELIZABETH . USHE RPapers published in SPECIA L LI BR AR IE S express the views oj the aul l~o rs nd do not represent the opinionor the pol~cyof th e editorial staff or the Publisher. Manuscsrpts submitted for publication must be typed doublespace on only one side of paper and mailed t o the editor. 0 Rep rint s may be ordered immed iately before oraft er publication. 0 Subscriptions: U . S . $10; foreign, $11; single copies, $1.50. O b y Special Libraries Associ-ation 3965.INDEXED I N Buciness Periodicals Index , Public Affair s Info rmat ion Service, Library Lit erature, Manage ment Inde xand Library Science Abstracts.

  • 7/29/2019 Special Libraries 65 (10)

    6/68

    Special Libraries AssociationPutting Knowledge to Work

    OFFICERS DIRECTORSPresident WILLIAMK. BEATTYALLEEN HOMPSON Northwestern Universitq MedicalGeneral Electric Company, San Jose, California School, Chicago, IllinoisPresident-Elect HELENEDECHIEFDR. F. E. MCKENNA Canadian National Radways,Air Reduction Company, Inc., Murray Hill, New Jersey Montreal, QuebecAdvisory Council Chairman PHOEBE . HAYES Secretary)HERBERT. WHITE Bibliographical Center for Re-NASA Facility, Documentation, Inc., College Park, Maryland search, Denver, ColoradoAdvisory Council Chairman-Elect RUTH NIELANDERMRS. HELENF. REDMAN Lambermens Mutual CasualtyLos Alamos Scientific Laboratory, Los Alamos, New Mexico Company, Chicago, IllinoisTreasurer GORDON . RANDALLJEANE. FLEGAL IBM Research Center, YmktoujnUnion Carbide Corp., New York, New York Heights, New YorRImmediate Past-President MRS.DOROTHY . SKAUWILLIAM . BUDINGTON Southern Regzonal Research Lab-The John Crerur Library, Chicago, Illinoi~ oratory, US. Department of Agri-culture, New Orleans, Louisiana

    EXECUTIVE DIRECTOR: BILL M. WOODSSpecial Libraries Association, 31 East 10 Street, New York, New York 10003MEMBERSHIP

    Dues: Sustaining-$100; Active-$20 (Paid For Life-$250) ; Associate4 2 0 Afilhe--$15 ; Stgdent-$2 ; Emeritur-$5; For qualifications,privileges and further information, write Special Libraries Association.

    - - -PUBLICATIONS

    *Business and Industrial Libraries in the. . . .nited States, 1820-1940, 1965 7.00Checklist for the organization, opera-tion and evaluation of a company li-. . . . . . . . . . . . . . . . . . . . .rary, 1960 2.00Contributions toward a special library. . . . . . . . . . .lossary, 2nd ed., 1950 1.25Correlation index document series & PB. . . . . . . . . . . . . . . . . . . .eports, 1953 10.00Creation & development of an insur-. . . . . . . .nce library, rev. ed., 1949 2.00Dictionary of report series codes, 1962 12.75Directory of business and financial serv-ices, 1963 . . . . . . . . . . . . . . . . . . . . . . . 6.50. . . .irectory of special libraries, 1953 5.00*Guide to metallurgical information(SLA Bibliography no. 3 ) , 2nd ed.,. . . . . . . . . . . . . . . . . . . . . . . . . . .965 7.00Guide to Russian reference and languageaids (SLA Bibliography no. 4), 1962 4.25Handbook of scientific and technicalawards in the United States and Can-. . . . . . . . . . . . .da, 1900-1952, 1956 3.00Literature of executive management. . .SLA Bibliography no. 5 ) , 1963 4.25*Latest publications

    Map collections in the U.S. and Can-. . . . . . . . . . . .da ; a directory, 1954 3.00National insurance organizations in the. . . .nited States and Canada, 1957 3.50Picture sources, 2nd ed., 1964 . . . . . . . . 6.75SLA directory of members, as of July. . . . . . . . . .5, 1964, 1964 .members 2.50nonmembers 10.00SLA directory of members, as of Octo-. . . . . . .er 15, 1962, 1962 .members 2.50nonmembers 6.00Source list of selected labor statistics,. . . . . . . . . . . . . . . . . . .ev. ed., 1953 2.00. . . .ources of commodity prices, 1960 5.00*Sources of insurance statistics, 1965 . . 8.25Special Libraries Association personnel. . . . . . . . . . . . . . .urvey 1959, 1960 1.00Special libraries: how to plan and equipthem (SLA Monograph no. 2) , 1963 5.55Subject headings for financial libraries,. . . . . . . . . . . . . . . . . . . . . . . . . . .954 5.00Subject headings in advertising, market-ing, and communications media, 1964 5.95*Translators and translations: servicesand sources in science and technology,. . . . . . . . . . . . . . . . . . . .nd ed., 1965 14.50U.S. sources of petroleum and natural. . . . . . . . . . . . . . .as statistics, 1961 6.00

    SCIENTIFIC MEETINGS-Subscription, $7.00; Foreign, $8.00SPECIAL LIBRARIES-Subscription, $10.00 ; Foreign, $1 1.00; Single copies, $1.50TECHNICAL BOOK REVIEW INDEX-Subscription, $10.00 ;Foreign, $11.00; Single copies, $1.50UNLISTED DRUGS-Subscription, $20; Single copies, $2.00SLA serves as the US . sales agent for selected Aslib publications

  • 7/29/2019 Special Libraries 65 (10)

    7/68

    The first-and only-index to insurance statistical materialSources of Insurance Statistics

    Elizabeth Ferguson, EditorAssisted by KatharineE. Cook and Mrs. Ruby C. Fangemann

    A detailed index to statistics published on a regular basis by commercial firms,insurance organizations, and some government agencies. Health, Life, andProperty-Liability Indexes have United States and Canadian sections and util-ize the terminology unique to their fields. Subject references give abbreviatedtitles of publications, data frequency, and period of time statistics cover. Mostdata refer to private insurance companies, but there is some coverage of vol-untary insuring organizations and the VA program. Annotated list of publica-tions and list of publishers' addresses.

    192 pages 1965 Cloth $8.25SPECIAL LIBRARIES ASSOCIATION

    3 1 East 10th Street, New York 10003SLA serves as the U .S . sales agent for selected Aslib publicatlons

    PAPERSONMICROFILM

    U p - t o - d a t e c a t a lo gl i s t i n g m a n y n e wt i t l e s a n d d a t e s o fnewspapers now pre-served on microfilmby Micro Photo.LIBRARIANS:Save space...consult thiscatalog to order backfileand current microfilm ofnewspapers in your area.Send for free copyM I C R O P H O TO

    D I V I S I O NBELL & HOWELLCOMPANY1700 SHAW AVENUECLEVELAND. OHIO 44112

  • 7/29/2019 Special Libraries 65 (10)

    8/68

    Xerox introducesPush-Button Copyingof Catalog Cards.Now you can reproduce catalog cardsin seconds with the Xerox 914.You can copy four different cards a t onceand make any number of copies vow want.

    Here's how you copycatalog cards in secondAll you need a re the originals and stand-ard card stock.

    1. Jus t load the card stock into the Xerox914 and push a button.

  • 7/29/2019 Special Libraries 65 (10)

    9/68

    When the due date rolls around, th e bor-rower and the book are instan tly identifiedfrom the information on the borrowing

    2. Place the catalog cards in this specialholder so the cards will be positioned prop-erly on the scanning glass.

    3. Dial the number of copies you want andpush another button.Now, just stand there. The Xerox 914does the rest.The 914 will even copy worn and frayedcards and make copies that are often bet-ter t han the originals.

    If you want to make a change on a cata.log card, you merely dab on Xerox Snopakecorrection fluid and type in any changes.The Xerox 914 never makes a mistake.Every copy is an exact copy. No moreproofreading.If thi s is the only job the Xerox 914 didfor you, you'd be 'way ahead. But there'smore. Much more.Eliminate typing overdue book notices.

    Many libraries that have a 914 copierar e converting to a new system that identi-fies borrowers of overdue books a s well asthe books themselves.To borrow a book, a reader fills out a bor-rowing card with his name, address anddetai ls of the book.

    card.Simply mail a xerocopy of the borrow-ing card to the borrower.Compare this method w ith the methodyou are now using. You may be surprisedby the amount of time and money you save.

    Cut down outside printing.Most libraries need a mountain of ma-terial copied in a year. Everything fromadministrative memos and telephone direc-

    tories to bulletins and acquisition lists.The Xerox 914 can handle these routinecopying jobs quickly and economically.How much will all this cost?

    You don't have to buy t he machine. A11you have to pay Xerox for ar e the numberof copies you make, based on a minimumnumber of copies per month. Plus a smallmonthly rental.What if you're opening a new branch?

    You'll need a complete set of new cards.Xerox will do th is job f or you in a fractionof the time an d cost i t usually takes.Fill out the coupon and we'll send you freeof cost our new 24-page booklet "XeroxSystems and Services for Librarians."

    Considering the amount of materialyour l ibrary has to have copied every day,it only makes good sense for you to get intouch with the people who make the mostadvanced convinp machine in th o ~ * r n r l +

    Send to : Xerox CorporationRochester, New York 14603.Name

    Address

    Stat e Zip Code

    XEROXX E R O X C O R P O R A T I O N , I 1 O C U E S T E R . N E W "OR6 14603 . BRANCH OFFICES

    IN P R I N C I P A L V. S. CITIES. C A N A D A : XEROX OF CANADA L lM l i ED , TORONTO.O V E I I S E A D : R I N X X E R D X LTO.,LONDON; FUJI-XEROX CO. . LTD. ,TOWO,

    8073 J O I N T L Y OWNED W lT H R A N I ORG IN IS I T ION , LTD.X E R O X A N D 91 4 A R E T R A O E M A I 1 K E O F XER OX C O R P O R A T I O N .

  • 7/29/2019 Special Libraries 65 (10)

    10/68

    It's here !A brand-newdirectory thatshows you over13,000 scientificand technicalperiodicalsfrom all overthe world!

    The first volume of Ulrich's INTERNATIONALPERIODICALS DIRECTORY, 1 1thedition,is now available for immediate delivery.I t gives you capsule descriptions of magazinesin every scientific, technical and medicalsubject. Included are periodicals from over100 different nations, as well as theU.S. and Canada.Unlike other guides, Ulrich's provides fu l ldetails on each magazine. . address. .editor. . . f i t carries advertising, book reviews,an index.. .where it's abstracted.. .PLUSlots of cross-references to facili tate your job.Here is an up-to-date revision of the directorylong considered the most authoritative toolin i ts field. It' s been compiled byEileen C. Graves, formerly of the PeriodicalsDivision, N.Y. Public Library, in consultationwith other professional librarians.Order your copy of Ulrich's Volume 1 oday.Once you see it , you'll agree i t wasworth waiting for.

    ULRICH'SInternationalPeriodicalsDirectoryVolume I ready nowVolume I1 (humanities, social sciences)ready 1966$15 net pp. each volume, clothbound

    . BOWKER C O M P A N YAvenue of the Americas, New York 10036

  • 7/29/2019 Special Libraries 65 (10)

    11/68

    These four scienti f ic an d engineering journals have been judged b y profes -sionals in th e instrumentat ion f ie ld to be the outstanding Soviet publ icat ionsin instrumentat ion.AUTOMATION AND REMOTE CONTROL 1Monthly; mathem atical ly or iented,emphasiz ing s tabi l i ty and opt imiza-t ion of automatic contro l systems.

    I INSTRUMENTS AND EXPERIMENTALTECHNIQUESI Bi -m on th l y ; devo ted p r im ar i l y tonuclear research and associatedinstrumentat ion.SUBSCRIPTIONS:

    OtherUS. Countr iesARC $ 6 0 .0 0 $ 8 0 .0 0IET 40.00 55.00MT 25 .00 35 .00

    I 4 5 .0 0 6 0 .0 0

    MEASUREMENT TECHNIQUESMonth ly ; covers measurement o fp h y s i c a l v a ri a b l e s p l u s t e s t a n dca l ib ra t ion o f measurement andrecording instruments.

    UIINDUSTRIAL LABORATORYM o n t h l y ; d ev o t ed t o m e t h o d s o fchemical analysis, phys ical investi-

    st.wrom 1957, except 19 60rom 1958 , excep t 1 96 1MT: f rom 195 8

    Complete volumes of back issues are avai lable at special pr ices. Pleasewrite for detai ls.Sam ple copies are available to l ibrarians w rit ing on organiza tional I

    INSTRUMENT SOCIETY OF AMERICA, Department R253 0 Wi ll iam P enn Place, Pi ttsburgh, Pennsylvania 15 21 9L

  • 7/29/2019 Special Libraries 65 (10)

    12/68

    SCIENTECHSCIENTIFIC &)) TECHNICALTRA VSLA TIOIVS

    QUALITY TRANSLATIONSALL MAJOR LANGUAGES

    BROCHURE ON REQUESTSCIENTECH P. 0 . BOX 18180 .CLEVELAND, OHIO 44118NEW FromWESTERN PERIODlCALS CO.

    The Ladenburg-Reiche FunctionTabulations created on the IBM 7094 at Rocketdyne Division of NorthAmerican AviationTable of Exponential FunctionsTabulations created on the IBM 7094 at Rocketdyne Division of NorthAmerican AviationEighth National Symposium of The Society of Aerospace Materials and

    Process Engineers on "Insulation-Materials And Processes For Aerospaceand Hydrospace Applications"Wave Mechanics Of A Free Particle by E. FisherContents include The Fundamental Constants, The Neutrino, The Elec-tron, The Proton, Unstable and Interacting ParticlesIEEE Mid-America Electronics ConferenceVolume 1-Electronic Systems Reliability, 1961Volume 2-Measurement and Instrumentation, 1963Volume 3-Measurement and Instrumentation, 1965Thermodynamic Properties of Individual SubstancesTables published by the Academy of Sciences of the USSR

    Exclusive Distributor

    WESTERN PERIODICALS CO.13000 Raymer St . , North Hollywood, California TRiangle 5-0555

  • 7/29/2019 Special Libraries 65 (10)

    13/68

    we'll duplicateall your catalog cards.. .. . . and return them to you within 72 hours. structions for your convenience . . . XeroxAny num ber of duplicate catalog cards, branch offices in principal U.S. and Canadiancopied photo-exact from your original. Same cities. Call the one neare st you-or writesize, durable, top-grade stock, precision- Xerox Corporation, Roch-punched an d trimmed, returned to you in the ester, New York 14603. I nsame order as the file you su bm it. . . Samples Canada, Xerox of Canadaand prices on request; special mailers and in- - Limited, Toronto. XEROX

  • 7/29/2019 Special Libraries 65 (10)

    14/68

    POLISHHISTORICALSOURCEMATERIALSMikolaj RejPOSTYLLA16th century evangelical serm onsbased on Erasmus' commentarieson the New Testament, contain-ing a rich, colorful picture of th e16th century Polish Church andreligiou s life. 79 woodcuts. E ditedby Institute of Literary Research,Polish Academy of Sciences. Pub-lished by Ossolineum, 1965. Intwo volumes, 34 x 23 cm. Vol. I,cri t ical notes, 321 pp. Vol. 11,photo offset repro duction of orig-inal text . 352 pp. $41.70Jan DlugoszANNALES SEU

    Latin chronicles of Polish historyto 1038, by the first Polish hist o-r i an a nd geographer . From amanuscript found in Paris. Liberprimus. Liber secundus. Preface;historical and geographical notes;Philological footnotes in Latin.Illustrated. Published by PolishScientific Pub lisher s, 1965. 25 x 20cm. 466 pp. Cloth, with jacket.

    $9.15Order from:ARS POLONAFOREIGN TRADE ENTERPRISE7 KrakowskiePrzedmiescie,WARSAW,PolandPostage will be added to price.

    Exp ert Service onMAGAZINE SUBSCRIPTIONS

    forSPECIAL LIBRARIES

    Faxon Librarians' GuideAva i l ab le on Request

    For the very best subscription service-ask abo ut our T il l Forbidden Auto-ma tic Renewal plan.

    F. W . FAXON GO., INC.515-525 Hyde Park Avenue Boston, Moss 02131

    Continuous Ser vke T o Librarzes Since 1886

    LOW COST PERIODICALA N D P A M P H L E T F I L E SA L L S I Z E S S A M E P R I C E

    $3.18 per doz. any assortment desired.12 doz. or more, $2 .8 8 per doz.FREE SAMPLE MAGAFILE sent uponrequest. You will receive it by returnmail along with handy size-chart andadditional details. No obligation or sales-man follow-up.

    THE M AGAFI LE CO.P. 0. BOX 3121 0 ST. LOUIS 30, MO.

  • 7/29/2019 Special Libraries 65 (10)

    15/68

    "THAT ROCHESTER MEETING O NENERGETICS.. .CAN WE GET COPIES

    OF THE PAPERS?"

    " I NEED A LIST OF ALL THE MEETINGS O NAEROSPACE THAT W ILL TAKE PLACE IN THE EAST

    NEXT FALL. CAN YOU GET IT FOR ME?"

    " I HEARD ABOUT A MEETING O N B IOLOGICALMEMBRANES THAT THEY'RE G O IN G TO HOLD THlS

    SPRING. WHERE CA N I GET MORE INFORMATION O N IT?'

    " I EXPECT TO HAVE A PAPER FOR THEACS RUBBER CHEMISTRY MEETING.

    WHAT'S THE DEADLINE FOR ABSTRACTS?"

    To find the answers to questions like these . . .and find them f as t. . . each for your copy of: . . .M S Technical Meetings Indexa quar t e r l y i ndex t o U.S. a n d C a n a d i a n h k e t i n g s i n e n g i n e e r i n g , t h e p h y s i c a lsc iences , chemis t ry , medic ine and the l i fe sc iences . Each i ssue coinple te ly. . . . . .ev i scd and cumula t ed . Indexed by: date of me eting keynrord locat ionof meet ing . . . ponsor ing organizat ion . . . eadl ine for abs t rac ts or paper s .11~12~1111~ t h ~ c r i p t i o f l :17 .50 U.S . t r m ! Crr~lrrdrr; 1 8 . 5 0 elsewhere.

    Send check or money order to: I MAIL THlS COUPON TODAY I I...................TM I S ( 13nclosed is $ for 1 ycar ~~lbscription.II l'leasc send sample copy an d morc information. ITe r j j dc nl Meeti ,rgr 1 S""".~.......................................................

    Itzformatio~z ervice 1 Org;cni/ation I................................................................22 Imperial Drive ( Addrcss

    I................................................................N e w Hartford, N ew York 13413 1 City .............................. Statc....................Zip............. II

  • 7/29/2019 Special Libraries 65 (10)

    16/68

    Science Citation Index1966coveragewill be

    1964coveragewent

    IS61coveragewas

    TO 1465P estimated)1115(as of September,more by year's end)

    S C I Journal Coverage is U P . . .U P . . .U P . . .UPIn 1961 we published the first Science Citation Index, indexing 613 keyjournals. The in itial list o f core journals selected emphasized multi-disciplinary ournals and basic engineering and science. . . physics, biology,chemistry and medicine. Ever since, we've been improving, upgrading,adding journals prac tically every day. In 1964 we added 87 journals and asource index with titles. In 1965 we added over 400 new journals. O n1966 our coverage will more than double our 1961 journal coverage,reflecting the suggestions and preferences expressed by our SCI audience. . . and we will continue to follow the same policy. In response to yourrequests, the 1966 SCI will include expanded coverage especially in theareas of oceanography, botany, agriculture, psychology, mathematics,radiation, metallurgy, nuclear science, paper research and informationscience. 13 ere is a partial list of information science ournals to be coveredin 1966:

    American Docum entation, Aslib Proceedings, Biblio theca Phonetica, Bu llet in of the Med icalLibrary Association, Communications of the ACM, Computer Journal. IB M Journal of Researchand Development, IEEE Transactions on Inform ation Theory, lnform ation an d Control, Infor.rnation Storage and R etrieval, Journal of the Association for Computing Machinery, Journal ofChemical Documentation, Journal of D ocumentation, Journal of Verbal Learning and VerbalBehavior, Kybernetik, Library Resources and Technical Services, Mechanical Translation.Methods of Info rmation i n Medicine, N achrich ten fi ir Dokumentation, Operational ResearchQuarterly, Operations Research. Phonetica, and Revue Intern ationale de l a Documentation.Every journal is indexed comprehensively. All items (except advertisements)in each journal are processed-all articles, editorials, letters, book reviews,corrections, meetings, etc.INSTITUTE FOR SCIENTIFIC INFORMATION 325Chestnut Street Philadelphia pa 19106 U S A

  • 7/29/2019 Special Libraries 65 (10)

    17/68

    NEW Simplif ied Indexing!saves up to 90% of

    Professional Searching Time!

    M A P S . C H A R T S - GR A PH S * D R A WIN G S A B S T R A C T S D O C U M E N T SPhotographically onto fully machine processable tab cardswith XEROGRAKARD*

    TAB CARD SHEETS- he Xerograkard system employs only con-Save the time which vent ional EAM and of f ice copy ing equipment wi thout modi f icat ion. It f i l l sthe gap between edge-notched cards and cost ly, sophis t icated e lec t ron icreally counts . . . or opt ical systems. Conversion f rom manual or convent ional EAM referencethe researcher's! systems is s imple and fas t . The image produced on th e card is as per -

    man ent as the pr in t in g on any tab card. Xerograkard cards comp lete ly e l im-inate the fa lse-sor t problem wi th super imposed coded sys tems.

    =Tr,hm=rk LET XEROGRAKARD HELP YOU FIND IT FAST!

  • 7/29/2019 Special Libraries 65 (10)

    18/68

    IMPORTANT NEW PHYSICS SERIESFROM CONSULTANTS BUREAU

    Place your standing order tod ay for books in these series. It w il l ensure the delive ry of newvolumes immediately upon publication; you wi ll be billed la ter. This arrangemen t is solelyfor your convenience and may be cancelled by you at any time.

    REVIEWS OFPLASMA PHYSICSAcad. M. A . Leontovich, Series edit01A systematic, multi-volume review of the pres-ent status of plasma theory, serving both as anintroduction for students and for researchersenterin g the field , and as a convenient, authori-tative, up-to-date presentation of current knowl-edge fo r wo rk e rs i n p l as ma phy s i c s . Th i scontinuing series, translated from Russian, isprepared by internationally known Soviet ex-perts. Each volume contains a number of in-tegrated tutorial reviews, covering in depth andi n breadth specific aspects of theory. In manycases, new material is presented.Volume 1A comprehensive introduction to "classical"plasma physics, containing authoritative paperson: Mo t~ on f Charged Particles in Electromag-netic Fields in the D rif t Approximation, by D. V.Sivukhin; Particle Inter actions i n a Fully IonizedPlasma, by B. A. Trubnikov; Transport Processesi n a Plasma, by S. I. Braginskii; and Thermo-dynamics of a Plasma, by A. A. Vedenov. Muchof the material in the f irst two papers is pre-sented here for the f irst t ime. Although thetheoretical analyses are quite advanced, theexperimental aspects of the subject are keptfirmly in view throughout.33 6 pages 1965 $12.50Volume 4Contains three papers: Hydrodynamic Descrip-tion of a Collisionless Plasma, by T. F. Volkov;Cooperativ e Phenomena and Shock Waves inCollisionless Plasmas, by R. Z. Sagdeev; andCoulomb Collisions in a Fully Ionized Plasma.by D. V. Sivukhin. The last paper contains newmaterial on the relevance of this topic to mirrormachines, provided by the author for the Eng-lish edition.24 1 pages January 196 6 $12.50Further volumes i n thi s series will be publishedduring 1966.

    LEBEDEV PHYSICS SERIESAcad. D. V. Skobel'tsyn, Series editorComplete English translations of the Proceed-ings ("Trudy") of the famed Lebedev PhysicsInstitute of the USSR Academy of Sciences pub-lished as Special Research Reports translatedfrom Russian.OPTICAL METHODS OFINVESTIGATING SOLID BODIES."Trudy" Volume 25Includes a major paper by N. D. Zhevandrov onpolarized luminescence of crystals. The secondpaper, by the late V. P. Cheremisinov, reports astudy of the vibrational spectra and structureof oxides; a final paper by L. A. Vainshtein con-cerns the calculation of cross-sections for ex-citation of atoms and ions by electron impact.19 4 pages 1965 $22.50COSMIC RAYS"Trudy" Volume 26Contains an account of the experimental in -vestigations into nuclear an d, electromagneticinteractions at high and ultra- h~gh nergles car-ried out i n the last few years in the laboratoriesand research centers of the Lebedev PhyslcsInsti tute.254 pages 196 5 $27.50RESEARCH INMOLECULAR SPECTROSCOPY"Trudy" Volume 27Devoted to spectroscopic investigations intomatter in various states of aggregation by themethods of Raman scattering and infrared ab-sorption. A special section IS devoted to themethodological problem of correct mg measuredquantities for instrumental errors.206 pages 196 5 $22.50Further volumes i n this series will, be publishedapproximately 6 months aft er t h e ~ r ppearance~n the original Russian.

    Q ONSULTANTS BUREAU227 We n t 17th Street . New York. N e w York 10011

  • 7/29/2019 Special Libraries 65 (10)

    19/68

    As in the past, the librarian is able to process only a very small propor-tion of his library inputs and has to rely on data centers and other index-ing-abstracting services to process the bulk of his materials; thus it be-comes increasingly imp ortant tha t their outp uts be comp atible. As a result,government agencies and technical societies are undertaking programsleading to the standardization of bibliographic materials.

    Centralization of Library ProcessingI. A. WARHEIT

    IBRARIANS HAVE been interested in cen-L tralized processing fo r a very long time.Some of the professional literature I ex-amined in preparation for this paper wasover 20 years old and some was publishedin 1965. Actually the question of centralizedbibliographic services has been with us forwell over a century; for effective utilizationof centralized indexing goes back to the firsttime a librarian used an early issue of C h e m -isches Z entralb ldt . Essentially, the need anddesire for effective centralized bibliographicprocessing has always been present. N o o neis really opposed to it, if it is done right. Al-though the reasons for support are rather ob-vious, it is well to repeat them here. Also, oneshould be aware of some new developments,which make centralization even more desira-ble an d in fact im perative. T h e specifics willbe developed in the following papers.The essential tasks of a librarian are tobuild and organize his collection and set upthe catalog, indexes, and other finding toolsso he can efficiently fill specific requestsfrom his clientele. This necessity to organizeeverything to suit parochial, specific needsgreatly influences the librarian to do special-ized processing. But this ideal of individu-alized specialized processing is no longerfeasible. In fact, in modern times, except foriery specialized libraries, specialized process-ng has never been really carried out.The librarian has always been hospitableo cooperation and coordination and the

    establishment of standards. Even the mostindividualistic librarian has appreciated cer-tain standards such as the 3 x 5 catalog card,descriptive cataloging form ats, and the stand-ard filing rules. It is really only in subjectanalysis of material that real differences ofopinion make themselves felt. For manyyears there was much interest in special clas-sification schemes for special collections.Many librarians developed their ow n schemes,there was a flourishing literature on the sub-ject, and usually every library school fea-tured it in its curriculum. Today, specializedclassification is pretty much a dead issue.Partially this is because there is less relianceon classification for finding materials. Thelarge, general classifications are more de-tailed, and a librarian simply does not havethe necessary time t o devote himself to con-structing and maintaining specialized classschedules.W h y S t a nd a rd s Are Need ed

    Today, the librarian is not only acceptinggeneral classification schemes but he is alsobeginning to demand standardized subjectanalysis. Th e reasons for this are many.1 . ARE AOF RESPONSIBILITY:oday the li-brarian is responsible for furnishing infor-mation that goes far beyond his immediatecollection. H e has to supp ly materials fromthe totality of available informa tion. A clientno longer says, "Do you have this in the

    rh is is the key note speech presented at a panel entitled "Cooperatio?l'hro ugh Autom ation: T h e Future in th e Nuclear Science, Engineeritzg, andWetals/M aterials Fields," wh ich wa s sponsored by th e Nzlclear Science2nd Engineering Sections of th e Science-T echnolog y Divisi on and th eWeta l~ /Mater ia l sDiv is ion at th e 56th Special Librar ies A~ soc ini i on C o n -!)enlion in Philadelphia, ] a m 8 , 1963. Th ree o f the papers giz,en at thebanel follow. Dr. W ar he it is Sepzior Systems Analyst , I B N Sys tems De-le lopment D iv is io~z , an Jose , Cal i fornia .

  • 7/29/2019 Special Libraries 65 (10)

    20/68

    collection?" but rather simply asks, "Willyou get this fo r m e?" H e doesn't really careif the item is owned by the library or not.With current finding tools, communication^networks, central data centers, and availablereproduction methods, a requester expects tobe furnished all the items he wants andneeds.The librarian thus must be able to get at aliterature he does not control and is, there-fore, dependent on others for the biblio-graphic control of this material. Again thisis nothing new. Only today it is a very muchlarger problem than it used to be, and it is amuch larger problem fo r the special librarianthan for the public or even the academic li-brarian.2 . Q U A N TI TY :he sheer physical growth ofliterature has made specialized, individualprocessing too expensive. More is being pub-lished both formally and informally. In fact,it is the latter, represented by reports, memo-randa, and other unique separates, that iscausing the greatest increase in the volumeof material the librarian must handle.3. S U BJEC T O V ER A G E:ot only has therebeen an increase in the volume of literature,but users' needs have been extending intomore subject areas. Modern science and tech-nology are much more interdisciplinary thanthey used to be. Furthermore, with thegreater use of the individual report andjournal article, the information being han-dled is much more specific. Greater detailedsubject knowledge is required to process thismaterial. T h e documentalist and subject spe-cialist have played an increasing role in thelibrary. A much higher quality of subjectanalysis is needed for effective library service.The general librarian finds that he cannotafford to process the enormous mass of in-formation to which his clientele must haveaccess. Nor is he adequately equipped toprocess the great variety of very technicaland detailed subject matter he must supply.The librarian may have excellent training inone or two specialities, but it is too much toexpect him to be an expert in everythinghe is called upon to handle.4. S P EED F SERVICE: aterial is being de-manded on an ever decreasing time scale. In-formation is consumed faster. In a sense ithas a shorter half life. People seem to learnabout documents before they are even pub-

    lished. AEC and NASA, for example, try tosend their microfiches to their member li-braries before Nuclear Science Abstracts andStar reach their readers. In other words,there is practically no time available foreven the most rapid bibliographic processing.5. SPANOF RESPONSIBILITY:n the past, alibrary user came to a central facility-thelibrary. Today a librarian is often called upo nto serve widely scattered personnel in a largeand complex organization. Library users areoften remote and must be served indirectlyby mail, phone, and messenger and throughbranch libraries, part-time clerks, and secre-taries. It is, therefore, difficult and often im-possible t o provide a user the necessary bib-liographic tools and aids a librarian mightbe able to fashion for a central collection.The isolated requester must depend on pub-lished indexes and abstract bulletins-ontools produced by centralized services.For all these reasons, the librarian mustprovide more and faster and better serviceto more people. He must have access tolarge masses of very technical literature,which is very widely scattered and which hecannot control directly. The professional li-brarian is being spread very thin, and he isbeing forced to look for better methods.Compat ib i li ty Be tween New Techniquesan d Services

    To meet these increasing demands, the li-brarian is not only working harder but ismaking use of new tools and enlisting theservices of other organizations. Technologyis providing him many new tools for proc-essing, communicating, and disseminatinginformation. These range anywhere from anoffice copying machine to a computer. As istrue with so many of these complex andexpensive tools, he has to share them withothers and, in many instances, is dependenton special technical personnel, whom hedoesn't control, to perform the necessaryservices.But more important than these mechanicaland electronic devices, the librarian tries tcdo a minimum of processing and have somcdata center d o it for him. T h e concept of thtcentralized info rm ation service has a fascina.tion for the non-librarian, for the persorwho is new to the technical informatior

  • 7/29/2019 Special Libraries 65 (10)

    21/68

    business. The many Congressional hearingsand reports by various Presidential commit-tees attest to this. These centers seem like asimple and easy solution to those who havenever been exposed to the grubby detailsof try ing to provide library service.However, there is a great deal of merit,if not in the grandiose, all-embracing cen-tralized service on the Russian model, atleast in the more traditional discipline orapplication-oriented documentation center.These information or data centers identifyand often even supply documents. Theymake the announcement bulletins, indexes,abstracts, and other searching tools an indi-vidual requester can use. The problem is toproduce products m ost suitable for the litera-ture searcher. And an increasingly importantrequirement is that the products from thesevarious sources be compatible-not only thatthey be compatible but that they be user-oriented and that they be understood by theuser-librarian.The sad history of some of the majorinformation services is that they have beendesigned without the ultimate user in mind.Weird and contrived file systems, special

    secret codes and so-called "machine lan-guages," esoteric compaction techniques thatcan be used only by special, unique com-puters, tricky language analysis routines,which seem to be better adapted for demon-strating the cleverness of the machine thanhelping the searcher, the crazy coding andavoidance of plain language to save tape andmake the program run faster, and, what isworst of all, the great neglect of the centralproblem, namely indexing, have all greatlydelayed and hindered the adoption of thenewer tools and methods that can be bene-ficial to the librarian and the library user.There is a great need for intelligibilitybetween the various services a librarian mustuse. Interchangeab ility of inputs is becomingserious. In the past when the "processor"was a human being, the trained librarian hadlittle difficulty in searching simultaneouslyGmel in and Chem ical Abstracts. Today theinflexible machine simply will not be ableto read NASA tapes, Index Medicus tapes,and ASM tapes in their varying formats, bitcodes, and file organizatio ns.The central processing agencies are be-ginning to realize that serious problems of

    compatibility must be overcome before theirproducts can be effectively used. T h e Clea r-inghouse has been working on standards fordescriptive cataloging of technical reports.There have been studies to reconcile the-sauri and authority lists used by the AECand the Defense Documentation Center. TheAmerican Society of Metals and the Engi-neers Joint Council are working together todevelop common standards for the Reviewof Metals Literature and Engineering Index.There is similar widespread activity in themedical-biological field. There is a strugglegoing on to develop a standard system forindexing chemical structures. The politicalscience people are now starting to search forsome standard system.In the field of reproduction, work is alsogoing forwa rd. M il-D specifications for a per-ture card reproduction of engineering draw-ings are now an industry-wide standard. Ofmore interest to librarians are the developingNASA-AEC standards for microfiche.Each aspect of all these cooperative effortscould and has consumed a whole conference.The important thing is that the librarian,who in essence is the real user of all thesecentralized services, be an active participantin developing these systems. If he isn't, then,as has so often happened in th e past, he willbe handed in ferior products that are not onlyuneconomical to use but ar e often useless.Centralized processing is here to stay. W ecan't do witho ut it. Let's g et it done right.

    ReferencesD A T A T RO LORPORATION.xperimental Study ofConvertibility between Large Technical IndexingVocabularies . . . by Wil l iam Hammond andStaffan Rosenborg. Silver Spring, Md.: August1962.MORSCH,ucile M. Cooperation and Central iza-t ion. Libvary Trends, vol. 11, no. 2, October 1953,p. 342-55.PAINTER,nn F. Convertibility Potential AmongGovernment Information Agency Indexing Systems.Library resource^- and Technical Services, vol. 7,no. 3, Summer 1963, p. 274-81.PIERSON,obert M. Centralized Cataloging: ItsImplications to Personnel. Library Journal, vol .90 , no. 4, February 15 , 1965 , p. 826- 8.POPECKI, oseph T. Bibliographic Information Ex-change. Library Journal, vol. 90 , no . 4, February15, 1965, p. 823-6.SHAW, Ralph. W ha t is Central in Center Services.Proceedings of th e Ame rican Docum entation In sti-tute, vol. I , 1964, p . 109 (abst ract ) .

  • 7/29/2019 Special Libraries 65 (10)

    22/68

    Features of computer input compatibility are considered using library cata-log records as an example. A common input can be utilized to meet dif-ferent local needs if a computer can identify the input elements, i.e., eachunit of catalog card data, and be programmed to manipulate the elementsto produce output in a different format. Achieving program compatibilityamong different types and models of computers is more difficult and costlyat the present stage of hardware development.

    Compatibility of ComputerInput and Programs

    HILLIS L. GRIFFIN

    I AM SURE librarians feel properly at homewith the topic of common input sincethey have used it for so many years. I refer,of course, to the Library of Congress printedcards, which have saved so much catalogingtime for so many libraries through the years.The cards serve as a common cataloging in-put, generated by one library to a standardspecification and used by other libraries,which may modify the information to suittheir unique needs and requirements.

    With more libraries placing their opera-tions on a computer, librarians are again con-sidering the problem of a common input inrelation to computer systems. They are look-ing for an L.C. card that can be read by acomputer to save the time required to re-keythis information for entry into a machinesystem. If all libraries used the L.C. inputwithout alteration, there would be no prob-lem, and the task would be relatively easy.But most libraries change the L.C. input insome way, major or minor, and it is at thispoint that the need arises for a common in-put that can serve as the means for achievingan uncommon output, which can be editedand adapted to the needs of an individuallibrary.

    This type of input has two functions: I )it serves as a guide to cataloging an item

    and 2 ) after some major or minor surgeryat the hands of the cataloger, the card itselfmay be used as the catalog document for anitem. If the surgery is major, the entire cardmay have to be retyped; if minor, offendinginformation can be crossed out or some newinformation added.

    Certain repetitive steps are performedwith the card in any event. The typist mustlocate the call number and perhaps type itin the upper left hand corner. The title mustbe typed across the top of the title cards,and subject headings must be placed in asimilar position on subject entry cards. Be-cause we know where pertinent informationis located on the unit card, we can bring itout and relocate certain items of informa-tion on certain types of cards that serve well-defined purposes in the catalog.

    The concept of identifying the variousitems of information within an entry is animportant one. W e don't have to write"author" before the author's name becausewe (and our library users) are intelligenthuman beings. Wouldn't it be nice if acomputer were bright enough to look at acard and cull these little tidbits of informa-tion out of i t? Th e problems of convertinga card catalog to punched cards or punchedpaper tape would be solved. W e would

    [Mr. GrijjGz is t he Infol-m atio~ z ystem s Librarian, Library Services Depart-me nt , Argonlze N at ioaal Ldboratory, Argo nne, I l l ino is . H e presetz ted th ispaper at a combin ed session of th e Science-Tech?zology Division's NuclearSc iem e and Engineering Sections a~ zdMeta ls /Materia ls Divis ion a t the56 th Specia l L ibraries Associa tion Convent ion i n Phi ladelphia , Jua e 8,1 9 6 s .

  • 7/29/2019 Special Libraries 65 (10)

    23/68

    simply be able to feed the card catalog into acomputer; it , in turn, would put out thisinformation in some machine-readable formupon w hich it could operate and go on toprint the catalog.Present computers, unfortunately, requireinput data in a form they can read. The li-brarian, then, must initially oversee thepreparation of the input in the form ofpunched cards or punched paper tape be-fore it can be accepted by a computer. W hi lethis may be no more trouble than typing asingle catalog card, it still consumes time.My feeling is that it takes less time to pre-pare a unit entry to a computer than it doesto prepare a full set of cards for a catalog,even if they are LC printed cards requiringonly the typing of call numbers and tracings.The reason is that a computer, given a unitentry, can be programmed to generate addi-tional records automatically for subject, title,and ad ded entries.Ident ify ing the Ind iv idua l Elements ofI n f o r m a t ion

    The basic problem of preparing univer-sally usable computer records is that ofidentifying each element of the input rec-ord. An author must be so identified to acomputer, which should know whether heis a corporate or personal author throughsome coding scheme. The title must beidentified, a s must the classification numb er.Cutter number, imprint, pagination, head-ings, and so forth. Given such clues, how-ever, the computer can identify and manipu-late each of these items of information (orfields) to give the desired end product.Given an in put in which the sequence of in-formation is 1 ) author, 2) title, .?)pub-lisher, 4) imprint , 5 ) series note, if present,6 ) classification number, and 7) Cutter num-ber, the computer could file this informationin any desired sequence within a given entryand could place the entries themselves in anysequence. Note that only the informationthat appears for each item need be identified.If there is no series note (field 5 ) , the com-puter can be programmed to sense this factwhen it reads a field 6 entry immediatelyafte r a field 4 entry.A list could be produced in call number(field 7 & 8) order by author (field 1)

    with other fields in normal order, omittingfield 4 (imprint) and field 5 (series note)and with the classification number and Cut-ter number appearing both at the head ofeach entry (at the filing point) and withinthe entry itself. This is possible because eachelement of information in the entry has beenidentified to the computer, and this identifi-cation makes it possible for the computerto manipulate the data to human specifica-tions.Since it is possible to produce machineinput in which each element of informationcan be identified by a computer it shouldfollow, then, that if it can be identified byone computer, it can also be identified byanother computer, and that if the original in-put conventions are known, it is possible totak e th is i n ~ u t nd translate it into anotherformat. If one wants to accept the originalinformation as it stands, simply translatingit into one's own format, a computer can dothis extremely rapidly. If one wishes tomake changes or edit the input, then a com-puter can reformat the input, print what ithas done in the desired format, and producea file that may be used to enter changes.One would also have the ability to changeonly tho se par ts of the data-the classifica-tion number and Cutter number, for exam-ple-that needed to be chan ged to make theentry acceptable to one's requirements.The key to the whole problem of com-patibility is identification of each logicalelement in the entry, plus a knowledge ofthe conventions that established the infor-mation within each element. Given this in-formation some one could, for example,take my magnetic tape file of bibliographictechnical report entries and make a file thatcould be used in his library for all reportswith the same report numbers. He couldsimply keypunch a card showing the reportnumber of each report and use this numberto select from my file the bibilographic datafor each matching report number.One example of a general machine-read-able input product has been generated as apart of an experimental program within theAEC Division of Technical Information.Citations and descriptors for each item ap-pearing in Nztclear Science Abstracts havebeen ~ u n c h e d in machine-readable form.Nearly every piece of information appearing

  • 7/29/2019 Special Libraries 65 (10)

    24/68

    in the bibliographic part of the NSA cita-tion has been included. This product is thenused by several different groups, each re-quiring different portions of the informa-tion. One group may require corporateauthors, and another may not. One may re-quire that author's names appear first-name-first, although they are carried last-name-firston the card.How do we go about satisfying theserequirements from a single basic input,making this input serve the needs of manydifferent users, each with different require-ments? The problem is merely one of identi-fying each discrete element of informationand tagging it with an identifying codewhenever it appears. The information mustalways appear in the same way, e.g., au-thor's last name first whenever it appears.Each entry must have a master entry identifi-cation number, type-codes for each differenttype of information, and sequence numbers,which tie together the punched cards withineach type of information. A type 2 card,for example, is always a title, and type 2cards with sequence numbers 90-99 are acomment. They may be included as a part ofthe title (because they are a type 2 element)but may be identified as a comment or de-leted entirely because the computer can tell,from the unique sequence number, what theyare. If you don't want the comment, havemachine instructions in your computer pro-gram that will tell the machine to bypass atype 2 card with any sequence number 90-99.If you want to identify it somehow as a com-ment, have machine instructions in the con-version program that will tell the computerto put the word "comment" in front of thisinformation when i t is reproduced.The basic entry has a fixed limitation onlength, but it actually need be only as longas is required for the actual information it-self. W e have allowed, for example, for atitle 6,300 characters long. In actual prac-tice, however, we punch only as many cardsas are required to show the title information.Editing to Satisfy Local Requirements

    If the input format is well known and ifthe discrete items of input information arewell identified, it is no problem to edit theoriginal input to meet local requirements.If authors are entered uniformly last name

    first, then it is no problem to invert themto a first-name-first format for output pur-poses. It is also quite easy to reduce a longtitle to a pre-determined maximum lengththat may be a part of jorir record format. Itis quite reasonable to combine rigidly fixedfields fo r certain information (pagination,date, etc.) while retaining the advantages offixed-variable length fields for other infor-mation that is more variable in length (suchas author, title, etc.). This gives extremeflexibility in the input product and makesthe editing process considerably easier.Certain logical steps may also be taken asthe result of the presence or absence of cer-tain information. If there is no date. forexample, the field reserved for the date inthe input record will be blank. The computercan be programmed to print NO DATE orDATE UNKNOWN (or some similar in-formation) on the output record if this isdesired. It can also substitute an alphabeticmonth for a numeric month, e.g., JANU-ARY for Ol/XX/66.This illustrates that the presence or ab-sence of certain information-can be a usefulindication to the computer program, becausethe computer has the ability to test for thepresence or absence of any character or char-acters and to take certain actions as a resuitof this test. The preceding example showshow the computer can be programmed totake an alternative course of action in theabsence of certain data as well as in its pres-ence. Certain constant information can besupplied by the program as the result oflogical testing, such as the prefix "19" tothe date. Spacing of the paper may occuron the printer, the computer may punch acard or write a tape record, or some otheraction may be taken dependent upon logicaltesting.The user output record format need not bethe same as the format of the input records,whether it be printed output or machineablerecords. Suppose that the input is furnishedin strings 70 characters long, but that youwish to use this information to write catalogcards at a maximum line length of 47 char-acters per line. The problem is to break upthe long 70-character input string to 47-character output lines without chopping offwords in the middle and to indent the firstline of the paragraph five spaces. To do this

  • 7/29/2019 Special Libraries 65 (10)

    25/68

    one would read in the first string of 70characters and find the position of character43 (which is the end of the first line withthe indentation). If character 43 is a blank,then characters 1 -4 3 may form the first line,and the string can be moved to the outputarea. If character 43 is not a blank. thenthe program should test character 42 for ablank, and if this fails, test character 41,and so on back. When a blank is found, thestring from character 1 to character ( 43 -n )is moved to the output area, and the addressof character (43-11) + l is stored. This is theaddress of the first character of the nextline. Testing for the end of this line, whichis 47 characters in leng th, begins at character(43-11) + 1 + 47, and the blank testingroutine is repeated again until an inter-wordblank space is located. When less than 47characters remain in the work area. the nextrecord of 70 characters may be read in andthe process continued to the end of the in-formation for that entry type. In this waythe input data is unpacked to the outputformat. A similar procedure may be usedfor repacking short strings to longer strings.Programm ing Prob lems in Convers ionof I n p u t

    The machine problems encountered insuch conversion schemes are not difficult oncharacter-oriented computers, such as theIBM 140 1 or Honeyw ell 200, but may provemo re difficult-although certainly no t im -possible-on wo rd-or iente d comp uters, suchas the GE 225, IBM 7090, or CDC 16OA.The whole procedure is entirely dependentupon proper identification of each segmentof the input and upon the fact that this in-put must be in a certain form. Rigid limita-tions need not be imposed upon the lengthof input data fields, although certain fields,e.g., date, lend themselves very readily tofixed fields. It is important to capture andidentify all data in th e original inp ut if thereis even the remotest possibility that it willbe of use. The expense of doing this at ini-tial input time is relatively low, but it can bequite costly to have to go back and retrievethis information later.It is important that full information begiven without abbreviation of titles or au-thors. Standard abbreviations may be usedif they are used consistently and are properly

    punched each time they are used. They maybe expanded at some later time if necessary,using a dictionary of abbreviations vs. fullspelling in the computer. Actually, the timeit takes to compute title length and reduceoversize titles to a maximum number ofcharacters may be more expensive than sim-ply punching them as they stand. The com-puter may edit them, if necessary, but theycan always be retrieved in full if later ex-perience indicates that they should be usedin full. Artificial limitations need not bemade, given the capabilities of present andprojected computer systems.Compat ib i li ty be tween Com puters an dCo mp u te r P ro g rams

    Given proper identification of the inputelements, it is not difficult to use a commoninput and tailor it to local requirements.The only real requirement is that the inputformat be known to the user and that allconventions be well defined. W it h this infor -mation, it is possible to write computer pro-grams to accomplish the conversion veryeffectively on almost any com puter. W he nwe talk about compatibility between com-puter programs, however, the waters growsomewhat more murky. Now we are talkingabout hardware. Computers come in differ-ent sizes and configurations, and the dif-ferences between computers, even of thesame series, may be considerable. In a way,the situation is not unlike automobiles, someof which have their engine in the rear, andsome in the front. Some have power steer-ing, and some don't. They all provide trans-portation, but the parts are not interchange-able.And so it is with com puters and with com-puter programs. If I write a computer pro-gram for my computer and I utilize the spe-cial features that are available on it, theprogram will not operate on your computer,even if it is the same general model as mine,if your computer doesn't have the specialfeatures required for operation of the pro-gram. If my program requires 16,0 00 posi-tions of core storage, and you have availableonly 4,000 positions of core storage, theprogram will be useless to you. The situa-tion is somewhat analogous to trying totransport in a single trip a n en tire basketballteam to a game in a small foreign sports

  • 7/29/2019 Special Libraries 65 (10)

    26/68

    car. A bus would d o the job very well but aspo rts car-never !If your computer is the same as mine andhas all the features required by my program,you can obtain the same results I d o by usingmy program as there is some degree of com-patibility between com puter programs.Even the so-called common programminglanguages such as FORTRAN and COBOLhave d ifferent restrictions fo r different com-puters to take advantage of the strong pointsof each machine. FORTRAN, especially, hasbecome strongly oriented toward user hard-ware with the introduction of many dialectsof the language. COBOL may appear at-tractive when writing for word-oriented ma-chines, such as the IBM 7090 and CDC3600, but it compiles a rather inefficient pro-gram for a machine such as the IBM 1401.Program optimization is especially importantin handling information applications, sincethere is generally a large volume of inputand output. This contrasts with most scien-tific applications, which have relatively littleinput or output but a great deal of internalactivity within the computer in arriving atthe answer.

    It is important to realize that all com-puters don't speak the same language in-ternally. They can all be programmed to dothe same job, but the instructions may bedifferent and the approach may be different,depending upon the computer and the in-structions available for use on it.Probably the best answer to program com-patibility will continue to be good programdocumentation, explaining through the useof flow charts and coding sheet commentswhat is being done at each step in the pro-gram. This is especially valuable whenchanges must be made. Proposed new com-puter languages and the proposed compati-bility of these languages from machine tomachine may answer the problem in part,but if your system has six tape drives and mysystem has two tape drives and eight diskfiles, it would require not only reprogram-ming but even, perhaps, a different approachto optimize the system to the hard wa re avail-able. In short, there is no ready answer tothe problem of program compatibility, andprobably the best that can be hoped for isthat you will be fortunate in finding a goodprogram that will work with your hardware704

    and will, perhaps with some modification,d o the job you want to do. Simulators, emu-lators, and translators only attempt to foolone computer into think ing that i t is another,generally at some penalty in operating time.Again, the application determines whetherthe cost is reasonable or whether reprogram-ming should be considered.

    T h e Dividing LineThere has been much evidence of, andmany words have been written about,the so-called info~ wza tion explosiotz ,particularly during the past fifteenyears. There is no doubt that we areconfronted with a serious situation,not only in physically coping with thevolume of paper, but also with the in-tellectual di sc ip lin e~ involved in thegeneration, preservation, and utiliza-tion of the information itself. Th e firstawareness of the problem, perhapsnaturally, was felt by librarians anddocumentalists. Only recently has thisimpact been of growing concern to thescience-research peop le themselves.An analysis of this situation, a sum-mary of which appears in TLD 91-65-071, has been made by a staff memberof the Battelle Institute. Several sig-nificant conclusions are drawn : (1)that scientists and engineers prefer todelegate to others the task of findinginformation, while they continue moreprofessionally satisfying work ; (2)that professional abstractors and in-dexers do a more objective job thanauthors do in identifying informationfor subsequent retrieval; and ( 3 ) thata combination of fantastically hugecentralized document depots and a net-work of specialized information analy-sis centers, supported by the govern-ment but privately operated, willincrease in number and importance.Hence, the dividing line begins to takeform between the laboratory researchscientist and the engineer, on the onehand, and the technical literature spe-cialist, on the other. Each will have hiscontribution to make toward the futureexpansion of knowledge.Reprinted from Trrhnlcul Library Bulletin,no. 8 4 , June 1, 1965, of th e AC ElectronicsDivis ion, Genera l Motors , Mi lwaukee .

  • 7/29/2019 Special Libraries 65 (10)

    27/68

    This article, premised on the rapid handling and manipulation of infor-mation by computers, discusses the problem of moving informationquickly from its source to the user, especially when long distances areinvolved. It offers wire communication as a solution and describes variousterminal equipment that may be used. Sharing a communication networkis presented as a way to reduce the cost to individual users.

    Sharing Communications NetworksARTHUR E. JONES

    C MPUTERS ARE COMING into their ownin their ability to handle large volumesof data. According to the prediction ofDavid Sarnoff of RCA, we will have in thefuture, "Computers capable of storing allof the information presently contained inall the world's libraries, responding to com-mands from human voices in different lan-guages, and automatically translating thespeech of one country into the spoken wordsof another."W e read and hear such statements every-day, but the big problem faced by those whoare located some distance from a computeris how to get at this store of information.The answer lies in the field of communica-tions. It is fairly simple to communicate witha computer in an on-premise situation, butwhen the distance is extended a few milesor many thousands of miles, there is theproblem of expense. One way to cut downon the expense is to share it with someone.Possible Communications Networks

    A group of libraries or even several groupscan be interconnected by the simple deviceof using the common carrier systems pro-vided by the telephone companies and West-ern Union in the United States, and RCA,ITT, and others overseas. Each library, usingTWX or telex, can dial any other libraryusing the same system and transmit a re-quest or a reply. Each of these systems works

    much like the telephone exchange, exceptthat the communication is written ratherthan spoken.Another way to interconnect is througha leased line, which can have any one ofseveral brands of communicating machinesattached to the terminal ends. In sharing acomputer, each library could be connected in-dependently with a computer location by aleased communication circuit.For libraries wishing to exchange infor-mation among themselves as well, a deviceis needed to tie all the independent circuitstogether and to switch the circuits or themessages so that libraries can talk to eachother. There are a number of such switchingdevices on the market. Without going intodetail, this equipment comes in a rathersimple form for switching a small numberof stations on a single circuit and becomesmore complex as it is necessary to switch alarger and more complex network of sta-tions. There is a growing trend to use com-puters to switch messages or circuits. Whynot use the computer that is going to beshared? This is possible, although it mightbe more desirable to use a smaller buffercomputer as the switching device to allowthe larger computer to operate more effi-ciently.

    Depending on the location of the librariesto be interrelated, it could be very expensiveto have each one independently connected to

    A condensation of a p q e r ptzseuted to t he Science-Ter-huology Division'sNuclear Science and Engineering Sections and th e MetalslMaterials D i-vision at the 56 th Special Libraries Association Conv ention in PbiIadeIphia,Jane 8 , 1965 , by the Supervisor of Communica tion S e ~ ~ i c e s .andia Cor-porat ion, A lbu qu erq ~e ,N e w ;Mexico .DECEMBER965

  • 7/29/2019 Special Libraries 65 (10)

    28/68

    a comp uter location. O ne way to minim izethis expense is by having several stations onthe same circuit. This setup works well be-Cause each station on the line usually needsthe line only for short messages scatteredthroughout the day. By taking turns, allmessages can be sent over the same line, andthe cost is shared among all users. If trafficincreases and another line must be added, itis still an adv antage to share.Still another example of sharing is a net-work using computer switching and trunk-ing. The trunking concept places several cir-cuits between the two points having thegreatest traffic. Then each station is fed intothe closest terminal point of the trunk, as theseveral primary circuits are called. Switchingequipment at these terminals searches foran open line and connects a calling stationthrough to the station at the other end.I n p u t a n d O u t p u t E q u i pm e n t

    Whatever communication system is used,it should be kept in mind that input and out-put devices come in many specialized forms.Librarians should not hesitate to considerthese special features. The display equipmentfor reference now in use and still on thedrawing board promise an almost unlimitedability to retrieve information. Input unitsare equally diversified.A typewriter of some kind is usually usedin conjunction with a computer. In the pastthe typewriter has been only an administra-tive device for control; input and outputwere handled by off-system preparations orprocessing of punched cards, paper tape, ormagnetic tape. Today a typewriter can beused as direct input/output. With one avail-able unit, a librarian can type an abstract ora catalog card on an typewriter, and t he copywill enter directly into a computer. In theprocess, the copy can be corrected if neces-sary. After the information is stored, it canbe retrieved on this same typewriter; if de-sired, the information can be revised oradded to an d then stored again.With the use of a cathode ray tube and aTV-like screen, graphic information can bedisplayed for reference and copied if neces-sary. While information is displayed on theface of the screen, it can be changed eitherwith a device called a light pen or with a

    typewriter. A requester in a distant referenceroom can be provided with a TV creen andan automatic page or card turner. When '1call is placed at the main reference desk, theproper reference book or tray of cards willbe placed in position in front of the camera,and the requester can proceed on his own.Micro film, in th e form of 35mm reels oraperture cards, can be examined on a T Vscreen; if the user wishes to retain the in-formation, it can be stored digitally by acomputer and retrieved at will. Informationfrom other sources, e.g., a computer or atypewriter, displayed on the system can beconverted into completely developed micro-film. Apparently the present equipment wasdesigned with engineering drawings inmind; however, libraries have been usingmicrofilm extensively in the last few yearsand could utilize this kind of equipment toretrieve and work with information storedon their microfilm reels. Several of the fac-simile manufacturers have features by whichmicrofilm can be scanned and transmitted toa distant poin t. It is received as a half-size tofull-size document.Facsimile offers an excellent, though pres-ently expensive, way to transmit pages ofinformation between libraries. Use of fac-simile eliminates what might be called"down time" when the reference documentis traveling by mail to and from a borrower.With Fax, a giveaway copy can be transmit-ted, and the reference document is immedi-ately available for th e next bo rrower.I have focused attention up to this pointon written or pictured information, but it isquite possible that libraries can use a morevolatile type of information for certain ref-erence problems. If so, they can telephone acomputer, ask a question by signaling withthe touch tone dial, and receive a spokenanswer. Answers might be limited to suchthings as informing the caller that a certainperiodical has or has not been received, butit's quite possible that there are other goodapplications.The more exotic developments in com-munication for the home may well provideeventual spin-off benefits to libraries. Thetelephone companies are working with theidea of reading gas and light meters over thetelephone line in the still of the night whilepeople are sleeping. Work is going on that

  • 7/29/2019 Special Libraries 65 (10)

    29/68

    will allow the housewife to pay her bills Using the minimum mileage in each exam-by phone. She would call the computer at ple given for TWX, the cost of a dedicatedthe bank and , by usin g a touch tone dial, di- line using 100 words-per-minute-equipmentrect that X dollars be deducted from her ac- would be approximately as follows:count and credited to the account of thedress shop o r gas company. 1 11Miles-$1 34 per mo nth40 1 Miles-$393 per mo nth2,001 ~ile s-$1 ,120 per monthCom municat ion CostsCost of communication circuits remainsa major factor when considering transmittingdata and other information over any distance.The cost situation has been changing overthe last few years, and it is reasonable to ex-pect a continuing reduction.Where short distances are involved, thereis no great saving advantage to be gained by

    sharing, but as the miles stretch out, theadvantages of sharing increase. Each groupplanning to share facilities should look atthis possibility. Several libraries planning toshare a computer at a given location mayfind themselves grouped around this loca-tion. At first, it might appear that eachshould simply connect directly with the com-puter location. Upon examination, it willusually be fo und that sha ring a single circuitwould be more economical. For example,consider 11 major cities in the area borderedby New York in the north, Columbia, SouthCarolina, in the south, Cincinnati in thewest, and Norfolk in the east. If a library ineach of these cities were connected to theOak Ridge Computer Center, it would re-quire approximately 4,500 miles of circuits,using the air line miles from each point.~ h e i e ircuits w ou ld cost a p p r o x i ~ a t e l y$4,450 per month. O n the other hand, ifall these stations shared the same circuit, theair line mileage would be approximately1,50 0 miles at a cost of approxim ately $900per month.Commercial TWX cost is based on 11bands of distance from the calling location,and the rate is slightly different for eachband. The following three samples will givesome idea. A call to a place 111 miles to185 miles away would cost 30 cents perminute in one minute increments. From 401miles to 550 miles would cost 45 cents perminute, and 2,001 miles and over wouldcost 70 cents per minute. If traffic betweenthe various points is heavy enough, it willbe worthwhile to use a dedicated circuit.

    Perhaps the greatest advantage accruingfrom sharing is the ability to reach all sta-tions at one time with a single transmissionof a message. This saves time lost in retrans-mission, and in some instances it saves pre-paring the message over again for transmit-ting t o each location if they are not connectedby a network. Information is captured in ma-chine language and can be reused or sharedwithout additional labor. Once a network isjustified and in use, there is a tendency tocommunicate more freely, and this leads toeven greater cooperation.Cooperation implies the sharing of re-sponsibility as well as the sharing of bene-fits, and there are some responsibilities. Forexample, networks are normally establishedbecause there is a common goal to bereached. Networks, like libraries, are run bya set of rules, and a responsibility exists toabide by these rules. Those contemplatingparticipation in a network must keep logsand trace messages.Conclusions

    To avoid being immobilized by the sheervolume of library acquisitions, libraries needto move quickly to the establishment of fastcommunication channels so that specializa-tion in one library can be shared economi-cally with others. Special equipment forcapturing data for storage, for retrieval, fordisplay, or for transmission is fast becominga reality. Now is the time to start planningfor the use of shared facilities to enjoy thegreatest possible interchange of information.

    SLA Sustaining Mem berThis is an addition to the Sustaining Mem-bers for 1965 listed in November SpecialLibrnries :

  • 7/29/2019 Special Libraries 65 (10)

    30/68

    Three general types of controlled vocabularies-classification schemes,subject heading authority lists, and thesauri-are compared for the gen-eration of printed indexes and production of stored indexes for internalinformation retrieval. Problems in sharing vocabularies concern vocabularystructure, subject content, and methods of manipulation or processing. Thethesaurus appears to offer the greatest flexibility for various types of in-dexing and greatest adaptability to varying degrees of automation. Tworecent developments in the area of vocabulary sharing are described-theCOSATI Subject Category Lzst and the joint effort of the American So-ciety for Metals, Engineers Joint Council, and Engineering Index to de-velop parallel thesauri and mutual computer processing systems.

    Sharing Vocabulary ControlMARJORIE R. HYSLOP

    Two ASPECTS of information storage andretrieval that are prominent in the mindsof documentalists and librarians today are:I ) automation and 2) sharing, which carriesthe connotations of compatibility, convert-ibility, interchangeability, and standardiza-tion. There are many facets to these two as-pects, but the one of concern here is the roleof vocabularies in systems that can be auto-mated and that can be shared. Mechanizedindexing, storage, and retrieval systems existthat do not rely upon controlled vocabularies,and among these are systems that can be au-tomated and shared ; permuted title (KWIC)indexes and selective dissemination of infor-mation (SDI) systems are two examples.However, the controlled vocabulary is animportant tool for increasing consistency ofindexing and precision of searching.A controlled vocabulary in the informa-tion retrieval context is an indexing vocabu-lary-an authority and guide to preferredterminology and a device for showing asso-ciations between words.There are three general types of controlledindexing vocabularies:

    1. Classification Schemes2 . Subject Heading Authority Lists3 . Thesauri

    There are also innumerable specializedvocabularies or hybrids of these three typesthat defy any attempt to force them into ageneral category. For purposes of simplifica-tion, problems in sharing will be consideredfor these three general types only.There are many points at which vocabu-laries can be shared and many ways of doingso. Ideally, complete sharing would involvethe utilization of the same vocabulary, in-cluding identical terms and vocabulary struc-ture, by two separate organizations for in-dexing separate bodies of documents. Suchcomplete sharing would presume that bothorganizations serve closely similar purposesand clientele, and this situation is not oftenencountered in practice. Therefore, sharingis much more likely to be on a partial basis.The problems involved in sharing fall inthree categories, which are discussed in thefollowing three sections.Vocabulary StructureIt is not the purpose of this paper toanalyze the structure of vocabulary buildingin detail, as this has already been done mosteffectively ( 1 ) .* However, basic differences* See numbered references a t end of paper

    -Mrs . Hys lop i~ A ~ ~ o c i a t eirec tor of th e A S M Documentat ion Ser t&-e ,American Society for Metals , Metals Park, Ohio. This is a condensedversion of th e paper she presented to th e Nuc lear Science and En gineeringSections of the Science-Technology Divisi on dnd th e Metals/M aterialsDivision at the 5 6t h Special Libraries Association C onventio n i n Phila-delp hia, Jutze 8, 196J,bat the references hnl e b een ndded.

  • 7/29/2019 Special Libraries 65 (10)

    31/68

    in structure of classification schemes, subjectheadings, an d thesauri should be reviewed.The structure of a classification schemeshows word associations by means of hierar-chies or family trees leading from genericterms to mo re specific terms at various levels.It is a rigid structure, which imposes thewell-known problem of terms that may fallin more than one family tree. Thus, the clas-sification scheme offers little latitude for ex-pressing different po ints of view-a deter-rent to sharing.The structure of a subject heading author-ity list, on the other hand, is loose, flexible,and simple. Its principal eleme nt of structureis the alphabetical array that provides wordassociations by bringing together all terms ofsame or similar root spelling. W or d associa-tions are also provided by the traditional"see" and "see also" references. Still anotherstructural refinement is the provision of vari-ous "levels" of indexing-m ain entry termplus sub-term, plus sub-sub-term, and so on.Sharing the alphabetical structure of a sub-ject h ead ing list is widely practised.The structure of the thesaurus ( 2 ) com-bines features of both the subject headinglist and the classification scheme. Like thesubject heading list, its terms are alphabet-ically arrayed. Like the classification scheme,it exhibits hierarchy by means of "broader-narrower" term designations, synonymousterm, and related term designations. Al-though the hierarchies are not so discreetlydisplayed , they go beyond the con fines of th etraditional classification array by permittingany term to appear in as many hierarchies asmay be appropriate. It is, thus, the most ver-satile of the three types of vocabularies inshowing wo rd associations.SubjectContent

    While traditional and special library clas-sification schemes have been w idely shared asintellectual subject guides for manual sys-tems, the shareability of their precise termi-nology in an automated system is severelylimited by their rigid structure. The same istrue of subject heading lists, but for the op-posite reason. Their loose alphabetical struc-ture so readily permits insertion and modi-fication of vocabulary terms tha t, even thou ghthe structure may be shared, the tendency is

    to adopt completely different terminologyor languages for each automated system.Sharing of terms in thesauri has been muchmore widely practiced, as will be noted inthe specific exam ples.N o matter what the vocabulary, two gen -eral problems are involved in sharing termi-nology-1) the concepts peculiar to a pa r-ticular discipline and 2) agreement on theterms themselves.The first problem can be illustrated by thefields of metallurgy, chemistry, and nuclearscience. In metallurgy, an intricate system ofword association is required to represent al-loys. For precise indexing and retrieval, alloyhierarchies can be based on composition(aluminum-copper alloys, aluminum-copper-magnesium alloys, etc.), innate characteristics(austenitic steel, dispersion hardening alloys,maraging steels), on properties (heat re-sistant alloys, magnetic materials), or onapplications (bearing metals, brazing alloys).Almost all these hierarchies overlap, and theproblem of controlling them in a vocabularythat provides sauce for the gander as well asthe goose is not a simple one.In chemistry a parallel problem is build-ing suitable terminological associations forchemical compounds, particularly organic,and in nuclear science difficulties are con-cerned with relationships between elementaryparticles, nuclear structures, isotopes, andsymbolic representations peculiar to physics.In dealing with this problem, not onlymust all of the word or concept associationsrequired within the detailed terminology it-self be provided, but also if vocabularies areto be shared, this detailed terminology mustbe made to fit within the framework of thebroad terminology cutting across narrowerfields of interest.Agreement on selection of terms is bothan interdisciplinary and a cross-disciplinaryproblem. T he best way to achieve such agree-ment seems to be by enlisting the assistanceand guidance of g roups of exp erts represent-ing a wide range of subject matter, regardlessof whether they know much about vocabu-lary building, indexing, or classification.Compromises and arbitrary decisions are un-avoidable, but once an optimum vocabularystructure has been designed and the opinionsof such experts are channeled in to this struc-ture, there is hope for developing a usable

  • 7/29/2019 Special Libraries 65 (10)

    32/68

    and acceptable indexing vocabulary withmaximum retrieval potential.Manipulation to Produce Indexes

    The bibliography of documentation andinformation retrieval over the past ten yearsis more liberally sprinkled with descriptionsof automated information processing meth-ods, techniques, and hardware than with anyother variety of published papers. Startingwith punched cards, both edge-notched andmachine manipulated, the documentalist's in-terest has ine&bly focused on the computeras the ultimate tool for information retrievaland, therefore, for adapting vocabularies toautomated indexing.Problems in mmipulation of the threetypes of vocabularies vary depending uponwhether the purpose is to produce a printedindex by automatic methods or one storedinternally in a mechanized IR system.Use of classification schemes for produc-ing printed indexes is rare, although not un-known. Some recent work along this line,which utilizes the Universal Decimal Classi-fication, is most interesting. Originally de-veloped for indexing Meteosological alzdGroustropbysical Tit les ( 3 , 4 ) , it has sincebeen extended to Geosc ieme Abstrac ts (5 ) .A good possibility should exist for adaptingthe computer programs used to produce theindex to other disciplines, which could bebased on UDC.Use of classification schemes for internalinformation retrieval indexes manipulated bycomputer has met with little success, al-though, according to Robert R. Freeman (6)there is. "a need for a fresh look at the useof data processing equipment in conjunctionwith classification systems in the light of ad-vancing technology. Earlier conclusions, gen-eralized from experience with punched-cardequipment, were too pessimistic. Computerscan render significant aid to humans in-~ o l v e dn improving and using existing clas-hification systems."The best collection of classification vo-cabularies (whether automated or not ) isundoubtedly that contained in the SLA Spe-cial Classifications Center at Western ReserveUniversity, Cleveland (7).Subject heading authority lists are thetime-honored \-ocabularies for generation of

    printed indexes, primarily because the alpha-betical arrangement is inherent. Examples ofcomputer generated indexes based on subjectheading lists are numerous, and new ones arecontinually popping up. The