The Need To Standardize Descriptive Cataloging*

The Need To Standardize Descriptive Cataloging*BY STANLEY D. TRUELSON, JR., Librarian

Yale Medical LibraryNew Haven, Coinnecticut

ABSTRACT

Because there are too many ways to describea book, its presence may not be discovered in abibliography or catalog. Standardized descriptivecataloging is needed to solve this problem and alsoto eliminate wasteful duplication of cataloging.The Anglo-Aniericant Cataloging Rules and theCOSATI Standard disagree on choice of mainentry, and the Library of Congress does not followthe AACR all of the time. But the essence ofstandardized cataloging is widespread availabilityand general acceptance of the data, regardless ofprinciples followed. Local adaptations in standardcataloging data are necesary, but those whichaffect all copies of a book, not just unique featuresof particular copies, must be made available foruse by all libraries by correction of the standardcataloging data. The national structure for com-municating standard cataloging data today ismainly printed tools, but tomorrow local libraryterminals on-line to a shared computer data bankmay provide the instantaneous access needed. Theproblem of getting the wider community of libraryusers to standardize their citation practices is moredifficult to solve, but hope for improvement liesin making access to standard data easier.

THE biggest problem of descriptive catalogingis that there are too many ways to describea book. Whether we want to know if a bookis in a particular collection or how it is relatedto other books in a collection, we often are un-able to make one another understand whichbook we are talking about, because we are un-able to agree on using the same way of describ-ing that book. This is more of a problem forpersons not trained as librarians than for thosewho are supposed to be experts on such matters,but librarians, too, can be stymied by the in-consistencies of descriptive cataloging.By "describing a book" and "descriptive

cataloging," I am referring primarily to descrip-* Presented as part of a general session on

"Current Cataloging Problems" at the Sixty-seventhAnnual Meeting of the Medical Library Associa-tion, Denver, Colorado, June 13, 1968.

tion of the contents rather than the physicalconstruction of a book and particularly to thechoice of main entry and the form of entryheadings.Some people say that careful descriptive

cataloging is unnecessary. We read about usingBooks in Print as a library's main catalog (1),and we hear proposals to copy without alter-ation the words on a title page as a sufficientguide to finding a book when one looks for itlater in a catalog. Such shortcuts undoubtedlywould work some of the time, but not enough ofthe time. Moreover, they would do little toshow the relationship of one title to anotherwhen one is looking, for example, for all vol-umes in a series or for several editions of a workwith varied titles. There is a need for descrip-tive cataloging, as opposed to mere copying.But it is a major problem that variations indescriptive cataloging sometimes confuse ratherthan clarify the bibliographic record.

In the January 1968 Library Trends, Tateand Wood point out that "the difference be-tween one library's cataloging practices andthose of another" is the result of local decisionsbuilt up over the years "by analogy with therules and examples found in the various editionsof the cataloging codes." They add that some-times the cataloging rules "are even tailoredto fit the cataloging practices of individual divi-sions and departments wthin the same library."This lack of compatibility between systems,they conclude, "is time-consuming and thereforecostly. It may even obstruct access to thedocument" (2).One can only agree. The great card catalogs

of our large libraries, proud products of decadesof devoted work, massively complex apparatusesas they are, and amazing tools for findingbibliographic needles in bibliothecal haystacks;all have their omissions and their inaccuracies.Who is to say that the special practices whichshaped one catalog were any more successful in

Bull. Med. Libr. Ass. 57(1) Jan. 1969 21

STANLEY D. TRUELSON, JR.

serving library users than those which shapedanother, when over the years dozens of mindsadded dozens of different interpretations of thesupposed special needs of one library's userscompared with another's? While striving for anelusive perfection in their catalogs, librarianshave caused problems beyond those they havesolved.The call for standardization is not new, but

it is being restated these days with renewedinsistence. The journal, Natutre, in an editorialof December 2, 1967, asserted that "the time ...has come to deprive librarians of the right todevelop independent cataloguing systems in themisguided notion that they thus serve the in-terests of their readers. No cataloguing system,"it continued, "can be ideal, but the advantagesof a common system override the virtues of anyone of them, especially now that computershave arrived in libraries" (3). And Verner Clappagreed when he wrote in the July 1967 LibraryTrends that "one lesson has been consistentlytaught by the experience of the last twocenturies, namely, that uniformity of practice-a common standard-is basic" (4).Hand in hand with the need to eliminate

confusion by standardizing the way in whichwe describe a book is the need to eliminateduplication of work by sharing the efforts ofOur catalogers. Cataloging talent and moneyare in short supply, and we have long beensquandering our limited resources by duplica-tive cataloging in libraries throughout our land.Librarians have a direct responsibility for seek-ing a way out of such wasteful overactivity.FortuLnately, the time is nearing when we cando better.

CONFLICTING STANDARDS

One might suppose that the publication of anew cataloging code in 1967 would have solvedthe problem of standardizing, but not so. Thenew Anglo-American Cataloging Rules (5) area great improvement in cataloging theory andprinciple (6) over the red and green bookswhich had guided catalogers since 1949 (7, 8).Yet, while Seymour Lubetzky's catalytic criti-cism in 1953 of the earlier rules for entry,which he called "vague in design and weak instructure" (9), stimulated the many years' plan-ning which led to the new book in blue, thenew code does not correct all the logical defectshe described, but instead embodies compromises

in the face of operating realities; namely, cer-tain long-imbedded practices of large librarieswhich represent investments too great tochange.

Most importantly, the new rules did not re-sult in total agreement on principles. We nowhave, for example, both the North AmericanText and the British version of the Anglo-Am?lerican Cataloging Rules: equal, perhaps, butnevertheless separate. And, as if they were notenough, we have for that peculiar breed ofdocument known as the technical report, apeculiar set of special rules published in 1966,called the COSATI Standard (10).The COSATI Standard is particularly pecu-

liar in its choice of main entry. Before ex-amining this choice, let us consider the needfor a main entry, as did Lubetzky in his 1953critique (11). There are those, especially somecomputer advocates, who say that, since acatalog or a computer file may contain all use-ful entries for a work, it does not matter whichof them is the main entry. But it does. Multipleentries are not always possible: many findinglists, such as library order files, library unioncatalogs, or published bibliographies, enter awork only once. In order to search those filessuccessfully to see whether they include aparticular work, one must use the same entrythat appears or would be apt to appear in thefile. So long as choice of main entry differs,bibliographic access may be prevented.

North, in the October 1967 Special Libraries,says that "the Anglo-American Rules, and theCOSATI Standard are in remarkable har-mony...." Yet she reasserts the divisive as-sumption of the COSATI Standard that "oneof the principle distinctions of report literatureis the relative unimportance of the personalauthor" and that "the organization responsiblefor the report has been generally recognized asthe most useful author citation" (12). Warheitstated in 1952 the main reason for this view."Many continued research programs, as re-flected in a series of progress reports," he said,"do not always have the same personal authors;they do have the same corporate author" (13).This arguLment, however, overlooks the fact thatin the academic community research projectsoften move from institution to institution rightalong with the research personnel involved.As a resuLlt of the clash of the main entry

ruLles of the COSATI Standard with those of the

Buill. Mecl. Libr. Ass. 57(1) Jani. 196922

THE NEED TO STANDARDIZE DESCRIPTIVE CATALOGING

Anglo-American Cataloging Rules, a scientificpaper written by a medical school professor, forexample, is entered under personal author, asit should be, when published as a journal articleor a monograph, but when published in theformat of a technical report with essentiallythe same scientific content and author re-sponsibilities, it is entered under the name ofthe author's employing institution. Sometimesthe work is published as a journal article, andreprints of that article are cataloged and dis-tributed under the procedures for technical re-ports with the technical report accession numberstamped across the face of the reprint.

There is no more logic to calling a universityor its subdivision the author of a faculty mem-ber's scientific or scholarly work when it hasreceived a federal grant to support his researchand writing than when it pays him from tuitionincome and gifts. And there is no more logicto creating an artificial corporate authorshipwhen the work is published in near-print, dis-tributed by the Clearinghouse for Federal Sci-entific and Technical Information, assigned AD,PB, or XYZ accession numbers, and indexedin the Government-wide Index to Federal Re-search and Development Reports and its com-panion abstracting services, than when it isdistributed in hard covers by a commercialpublisher and thus listed in the main part ofthe National Library of Medicine CurrentCatalog instead of in the new Technical ReportLiterature sections.Worse than the breakdown in logic is the

breakdown in producing standard catalogingdata. But the National Library of Medicineaccepted these illogical distinctions in practicewhen it announced that "papers which reporton research in progress, usually being performedunder contract with a U. S. Governmentagency," would be cataloged under rulesadapted from the COSATI Standard and listedseparately in the 1968 issues of the CurrentCatalog (14). As a result of NLM's applicationof this policy, medical school faculty papersare entered under the names of the schools,whereas, if the same papers were in journal orbook format, the schools would be identifiedonly in small print under the authors' namesor in a footnote and not at all in the catalog orindex entry. Even student theses for the M.S.degree are listed in NLM's Current Catalogwith the name of the university and its sub-

ordinate center as the main entry heading,merely because of the financing and format Ofthe published versions (15).

These examples, it is true, are exceptions tothe typical progress report, which is moreephemeral in content and more efficientlycataloged as an institutional report than as ascientific paper. But a way to distinguish thepreliminary laboratory or corporation researchreport from the true scientific paper or mono-graph of personal authorship, when either canbe clothed in the covers of a technical report,apparently has not been discovered. TheCOSATI Standard thus may have approachedcompatibility with the Anglo-American Cata-loging Rules, but it has not quite offered theultimate embrace.

VARYING PRACTICESSuch variations in principle are not the only

roadblocks to standardization. Once we finallyagree on policy, we still must face the differingdegrees to which it is accepted and followed.The country's leader in cataloging, the Libraryof Congress, follows the Anglo-American Cata-loging Rules only under very special circum-stances, an approach which it calls "superim-position" (16). "Superimposition" means thatonly new titles-and only some of them-arehandled under the new rules and that the in-consistencies and poor practices of the pastusually must remain untouched. The only realstandard possible in this melange of practiceand occasional principle is the standard that"what is shall be": the cataloging descriptionwhich has any possibility whatsoever of be-coming the established version is that whichissues from the Library of Congress on itsprinted cards, computer tapes, or publishedcatalog volumes. Perhaps, however, this situa-tion is not as bad as it might seem.

ESSENTIAL ELEMENTS

Principles or guidelines, in the end, do notguarantee consistent practice; they only encour-age it. Beyond cataloging principles stands theproduct that they lead to, the cataloging recorditself. One of the reasons that the experimentseveral years ago of cataloging in source, wherethe cataloging record was printed on the versoof a book's title page, offered so much promise,even beyond the easy availability to libraries ofcataloging copy, was that readers, booksellers,



and librarians alike could agree on the standarddescription of a particular book, because thereit was in black and white with even the blessingof the publisher himself. One may hypothesizethat the essential elements of standardizedcataloging are widespread availability and gen-eral acceptance of the data for a specific bookand that conformity to any particular catalogingcode is less important than these two elements.It may be annoying when, for example, oneyear's conference proceedings are entered undera main heading radically different from the pre-vious year's, but even this inconsistency wouldbe less serious if everyone agreed to use thesevariant headings for their respective volumes:at least we could make each other understandwhich book we were talking about, even if therelation of that book to other volumes in itsseries remained cloudy.

Shared cataloging thus contains already thepotential for becoming standardized cataloging.If access to the cataloging records of the Libraryof Congress and the National Union Catalogwere rapid and easy for all users of books,those records, in spite of their internal incon-sistencies, could become universally used cata-loging descriptions of the books they cover.They could, that is, provided that a mechanismis created for correcting errors.

LOCAL ADAPTATIONS

Someone may object that it is impossible forall libraries to accept a shared cataloging recordwithout all kinds of local adaptations to fit spe-cial circumstances. How do we handle the pleth-ora of local adaptations? Are they merely the re-sult of downright refusal to be satisfied with eventhe thought of using an established entry?We cannot sidestep the need for local adapta-

tions by pretending that there is no need. Whilethe needs are probably far fewer than librarypractice admits, there are instances in which alocal library must revise the standard catalogingrecord in order to give effective service. For ex-ample, in the National Library of MedicineCurrent Catalog (2: N-566, Jan.-Sept. 1967)there is an entry as follows:

Citizens Commission on Graduate MedicalEducation. The graduate education ofphysicians. [Chicago, 1966] xii, 114 p.... I. Title.

Except for subject entries, no other added

entries are traced. While the main entry isproper, this cataloging record poses a prob-lem, because it is incomplete. The work isa report commissioned by the American Medi-cal Association and distributed by its Councilon Medical Education, and a reader may lookfor it under the name of either the AMA or itsCouncil. And like many reports of major in-terest (for example, the DeBakey Report and theReport of the Warren Commission), this one ispopularly known by the name of the chairman ofthe reporting body; in this case, as the Millis Re-port. Any library hoping to retrieve this docu-ment whenever needed must make at least twoadded entries not in the established catalogingrecord. My own library catalog now containsa record for this work much different fromthe one it, itself, was responsible for establishingin the National Union Catalog (Jan.-March1968, pt. 1, p. 64).Another situation in which the cataloging

record must be altered by the local library iswhen a data error, such as a misspelling, occurs.The usual percentage of human errors creepsthrough even on the Library of Congress'sMARC tapes. (The computer, it seems, is notmuch better than the typewriter in helping usto spell. While it is possible for a computer tocompare an entry with previously establishedentries and report similarities so that a humaneditor can examine them to determine whetheran error has been made, the computer cannotdo much if the file contains nothing similar tothe newly entered name.) When a local librarydiscovers an error of this kind, it is compelledto correct it.

These examples, along with the previousmention of variant headings for congress vol-umes, should be sufficient to demonstrate thatthe need for local adaptations in shared cata-loging is not merely an illusion of perfectionistmembers of the library profession; it is some-times a genuine need, and means must exist tomeet it.

Certain notes must be added to catalogingdata in local libraries to describe unique fea-tures of their copies, such as an inscription or adefect. These can be added to the establishedrecord as local option notes. But the examplesjust given are of basic elements of catalogingrecords which are inadequate for general use,as well as for the particular local library thatrevises them. What is required, therefore, is a

Bull. Med. Libr. Ass. 57(1) Jan. 196924


feedback method whereby essential correctionsto the standard record may be proposed by anyuser, reviewed by a coordinator of some kind,and acted upon when appropriate in order togive all users the benefit of the needed changes.In the national on-line computerized systemwhich we may have in the future, such updatingshould be fairly painless. But, for the present,it is very difficult to erase the entry in the pub-lished catalog or printed card and notify allwho used the earlier version of the changes.This is one reason why advances in computerstorage and distribution of cataloging data areimportant-indeed, essential-to the solutionof the major problem of descriptive cataloging,our present inability to obtain and use astandard cataloging record which is satisfac-tory. The fact that computers may soon tacklethis storage and distribution job, including up-dating where needed, is the bright ray of hopeahead in this field.

COMMUNICATION STRUCTUREThe structure for communicating standard

cataloging data in the United States today con-sists of published volumes of the NationalUnion Catalog; printed cards, proofsheets, pub-lished volumes, and magnetic tapes of the Li-bary of Congress; Library of Congress catalog-ing published in Publishers' Weekly andAmerican Book Publishing Record; and some-times alternate versions of the data in the Na-tional Library of Medicine Current Catalog andthe Cumulative Book Index, to name thesources which come foremost to mind. Recentarrangements by the Library of Congress toaccept the cataloging done for the nationalbibliographies of some other countries have in-creased the speed and comprehensiveness ofdata available from the present communicationstructure. But, the structure tommorrow (ormore likely, the day after tomorrow) may be ashared computer data bank fed by many li-braries, although primarily by the Library ofCongress, the National Library of Medicine,the National Agricultural Library, and largeuniversity and other research libraries, just asthe National Union Catalog is fed now. Fromthis data bank, libraries across the countrywould be able to obtain instantaneously theestablished cataloging record for any book bymeans of an on-line information terminal or,if no record were in the data bank, to enter their

own original cataloging for the benefit of allsucceeding inquirers of the data bank. The verysmall libraries, so long as the cost of informationterminals continued to be beyond their reach,should be able to arrange for the services of aterminal at another nearby library, perhaps bymailing batches of requests to be searched inthe computerized data bank.

CITING IN THE LITERATUREBut even if libraries succeed in standardiz-

ing, how can an entire nation ever agree inpractice on a particular cataloging entry for aparticular work, even if it should agree on therules for constructing entries? Not everyonecan be trained in the rules-certainly not thegeneral public of library users-and lists willcontinue to be made either in ignorance or indisregard of the rules, let alone with varyingresults from applying them.The sad complaint of a book reviewer in a

recent issue of Nature emphasizes this dilemma:"It is a pity," he writes,

that this good book is a "conference proceedings."As such, despite the useful and valuable informa-tion it contains, the book will be neglected. It willbe misplaced in the library, in that it will not beamong the current literature; it will be difficult torefer to and it will be tedious to write out thereferences to the articles which it contains.

This seems to be the fate of all conference pro-ceedings now, with the exception of those publishedin the normal literature such as Discussions of theFaraday Society. By being produced sometimesunder the name of an editor and sometimes not,they inevitably seem to be filed in the wrong places.Quoting articles from them in papers or in reviewsis always difficult: for anyone who does not agree,I suggest that they write out for themselves a shortform reference of the article contained in thisvolume by G. Porter on page 79, working fromthe full reference to the book shown above. Isuggest there will be as many versions as peopleattempting the problem and many of them willbe ambiguous and lengthy.

The reviewer suggests that a solution would befor publishers to make all their conference pro-ceedings numbered parts of series, which couldthen be cited by series title and volume, like aperiodical. "In this way," he states, "our li-brarians would both wish to obtain them andknow where to file them; they would be easy toreference" (17).

It is not likely, of course, that this would bea real solution, even if a consistent form of



series title were followed. One of the most am-biguous situations in descriptive cataloging iswhere there is a choice of either series titleentry on the one hand or one of the severalkinds of entries for a volume with separateauthors, editors, or title, on the other. It is notenough to cite a paper by Egger and Flynn asbeing in Progress in Brain Research 27: 165-182, 1967, when many librarians, contrary toour reviewer's hopes, may have the volume onorder with a single entry in their catalogs under"Adey, W. Ross, ed." with the specific title ofStructure and Function of the Limbic System,even though the order slip may state that thistitle is volume 27 of Progress in Brain Re-search. On the other hand, there are equaldifficulties when libraries cannot afford tocatalog series volumes as separate works (themonographic supplements to journals such asthe Scandinavian Acta, for example) and re-gretfully are compelled to trust that libraryusers will know the series title and volume andthus find the monograph that the card catalogfails to reveal under its author, editor, or title.The only apparent way to improve citation

practices is to publicize the use of standardcataloging records. The standard record coulddecide the question, for example, of whetherthe series volume should be cited as a separatetitle or as merely part of the series, and even fur-ther, could decide the exact entry heading. Withthe standard record on hand, we would all knowenough to cite volume 27 of Progress in BrainResearch, for example, not under editor Adey,but under "Symposium on the Structure andFunction of the Limbic System, Hakone, Japan,1965." We need to be able to point to a sacredtext, a bibliographic bible, as it were, in whichthe authoritative entry, the specific entry for thespecific work in question, is clearly, accurately,and fully established for all to use withoutquestion. This suggestion may have repellingovertones of authoritarianism, but it is intendedto help bring order out of chaos, perhaps an endthat justifies the means.How do we reach the community of citers-

authors who list books and papers as referencesin their own works? The chaotic world of free-for-all citation practices seems almost beyondreform. We must admit that it can never be ex-pected to reach the state of orderliness whichshould be possible in library catalogs. But inso-far as those catalogs have reasonably complete

cataloging data for each book, filed under mul-tiple entries, the brief or erratic citation in theliterature, so long as it is not grossly inaccurate,should permit a successful search of a library'sholdings' files. It is primarily in the use of single-entry lists where the so-called "dirty" citationcauses trouble, and it is because the choice ofmain entry is crucial to the use of a single-entryfile that the advantages of using standard entriesmust be publicized among authors and otherlibrary users. They will never all be persuadedor trained to discover and use standard entries,but it is very possible that many would changetheir habits as knowledge of the standard en-tries becomes easier to obtain. Some progresswould be better than none, and for this partialimprovement I think we should work hard:first, by intensifying our efforts to reach agree-ment among libraries on using an establisheddescription of a particular work, second, bydeveloping a computerized cataloging data bankand a rapid means of sharing it, and, third, bypublicizing and encouraging readers' use ofthe standard data.

SUMMARYThe problem, then, is our failure to agree on

a standard way of describing a book so thateveryone can refer to it without confusion andlibraries can save cataloging time and money.The technological help of computers is grad-ually making possible a system for rapid sharingof new and old cataloging data among manylibraries throughout the country. Libraries un-able to maintain their own computers may beable to afford terminals linking with computersand cataloging data banks elsewhere. To dis-tribute standard cataloging data beyond librar-ies to the wider community of library users isa more difficult matter and one less likely ofadequate solution, but the prospects of at leastvisible improvement are good once librariansget their own houses in order.

REFERENCES1. GORE, DANIEL. A shortcut to book catalogs?

Libr. J. 93: 1110-1113, March 15, 1968.2. TATE, FRED A., AND WOOD, JAMES L. Libraries

and abstracting and indexing services-astudy in interdependency. Libr. Trends 16:360-361, Jan. 1968.

3. Where should books be kept? Nature 216: 881,Dec. 2, 1967.

4. CLAPP, VERNER W. Retrospect and prospect.Libr. Trends 16: 173-174, July 1967.

Bull. Med. Libr. Ass. 57(1) Jan. 196926


5. AMERICAN LIBRARY ASSOCIATION, et al. Anglo-American Cataloging Rules. North AmericanText. Chicago, American Library Associa-tion, 1967. xxi, 400 p.

6. FIELD, F. BERNICE. The new catalog code: Thegeneral principles and the major changes.Libr. Res. Tech. Serv. 10: 421-436, Fall1966.

7. AMERICAN LIBRARY ASSOCIATION. DIVISION OFCATALOGING AND CLASSIFICATION. A.L.A.Cataloging Rules for Author and TitleEntries. Edited by Clara Beetle. 2d ed.Chicago, American Library Association,1949. xxi, 265 p.

8. LIBRARY OF CONGRESS. DESCRIPTIVE CATALOG-ING DIVISION. Rules for Descriptive Catalog-ing in the Library of Congress. (Adoptedby the American Library Association.)Washington, 1949. vi, 141 p.

9. LUBETZKY, SEYMOUR. Cataloging Rules andPrinciples; a Critique of the A.L.A. Rulesfor Entry and a Proposed Design for TheirRevision. Washington, Processing Depart-ment, Library of Congress, 1953. ix, 65 p.(p. 61.)

10. U. S. FEDERAL COUNCIL FOR SCIENCE ANDTECHNOLOGY. COMMITTEE ON SCIENTIFIC ANDTECHNICAL INFORMATION. Standard for De-scriptive Cataloging of Government Scien-tific and Technical Reports. Revision No. 1.Washington, Clearinghouse for Federal Sci-entific and Technical Information, Oct. 1966.iii, 50 p. (AD 641 092, PB 173 314.)

11. LUBETZKY. Op. cit., p. 57-60.12. NORTH, JEANNE B. A look at the new COSATI

Standard. Spec. Libr. 58: 582-584, Oct.1967.

13. WARHEIT, I. A. Bibliographic identification andorganization. Amer. Docum. 3: 105-110,Apr. 1952.

14. Current Catalog to broaden coverage. Nat.Libr. Med. News 22: 3, Dec. 1967.

15. Lehigh Univ., etc. Cited in: Curr. Cat. 3: 49,Feb. 26, 1968.

16. Libr. Congr. Inform. Bull. 26: 183-185, March9, 1967. (Amplified in Libr. Congr. Cat.Serv. Bull. 79-81, 1967.)

17. BORRELL, PETER. [Review of] Reactivity ofthe Photoexcited Organic Molecule.... 1967.Nature 216: 900, Dec. 2, 1967.


The Need To Standardize Descriptive Cataloging*

Documents