Cataloging & Classification Quarterly

Volume 22, Number 2 1996

By Elizabeth N. Steinhagen, News Editor

It is the purpose of this column to collect and disseminate information on all aspects of cataloging and classification. I would like to include news concerning you, your research efforts, and your organization; in fact, I'd like to try and expand coverage to include information about cataloging activities all over the world. Thus, in the name of our many readers, I am asking you to send any pertinent materials, notes, minutes, and including the following:

Research and Opinion



Please send all correspondence to Elizabeth N. Steinhagen, General Library, The University of New Mexico, Albuquerque NM 87131-1466.

Cataloging News for this issue:


The following reports were contributed from the AMERICAN LIBRARY ASSOCIATION Annual meeting, Chicago, June 22-29, 1995.

A program titled "Crisis in Subject Cataloging and Retrieval" was sponsored by ALCTS Cataloging and Classification Section, Subject Analysis Committee and the RASD Management and Operation of Public Services Section, Catalog Use Committee, on June 25, 1995.

Arlene G. Taylor, (University of Pittsburgh), in her talk "Introduction to the Crisis," stated that there has been an erosion of confidence in subject cataloging, which is frequently thought not to be cost-effective. Signs of the crisis are 1) an administrative push to cut back or eliminate subject cataloging, 2) lack of sufficient education in the theory and practice of subject analysis, leading to a lack of understanding on the part of non-catalogers, 3) a widespread negative view of Library of Congress Subject Headings (LCSH), and 4) a view of classification as only a way of arranging items on a shelf, and therefore clearly dispensable in an age of online information. Reasons for the erosion of confidence are 1) the availability of keyword searching, which many people think is sufficient, 2) the difficulty of subject analysis in an expanding universe of knowledge--including the increasing variety of materials, and of different formats, not all of which are suitable for traditional subject analysis--increasing variation of word usage even in the same language, the appearance of new subjects requiring new terminology, and the use of multiple thesauri with little or no attempt to relate them to each other), and 3) the "since it can't be perfect" syndrome, i.e., since subject analysis is subjective anyway, so why bother?

Francis Miksa, (University of Texas at Austin), spoke about "Bibliographic Control Traditions and Subject Access in Library Catalogs". Suggesting that we need a broader perspective, partly historical, and a new approach and methodology, he discussed 1) bibliographic control as a general model and the various traditions of bibliographic control, and 2) the measure of a single bibliographic item, and how much information about it belongs in an entry in a bibliographic control system. Bibliographic control is any attempt to gain power over the information-bearing objects which comprise the bibliographic universe. The universe of knowledge is intangible and ordered, and resides in information-bearing objects, while the bibliographic universe is tangible--being made up of objects--but unordered; bibliographic control consists of identifying and ordering bibliographic objects so that they can be retrieved and used to help people reach the universe of knowledge. The types of bibliographic control that have arisen are--in chronological order--1) bibliography, 2) library cataloging, 3) indexing and abstracting, 4) documentation and information storage and retrieval, 5) archival enterprises, and 6) records management. The nature of a single bibliographic unit--that is, the basis of an entry in a bibliographic organization system--differs among these traditions of practice: in archives, it is a collection from a single source, in records management a group of records, and in library cataloging it was originally one book containing one work by one author.

The first breakdown of this ideal was the appearance of information-bearing objects containing more than one work, such as transactions of learned societies, periodicals, etc.; the solution to this breakdown was analytical cataloging, and the result was the rise of indexing and documentation. The second breakdown, originating in indexing and abstracting, was the discovery that subject access is not limited to a work as a single bibliographic item, and that it is not simply concerned with "aboutness". The response to the second breakdown was the fragmentation of the concept of the unity of a work into the concept of the work as a conglomeration of topics, forms, and genres. Therefore, library cataloging is two breakdowns behind, and still operating with a simplistic view of a document as a unit.

Miksa cited some examples of books that contain far more variety of material than a few subject headings would indicate, and pointed out that each field has its own genres that must be dealt with by going well beyond subject headings. He characterized library cataloging as a sense-making operation, pointed out that library cataloging is the leader in the use of controlled vocabulary, and suggested that it would be useful for people from the different bibliographic control traditions to communicate with each other.

Sheila Intner, (Simmons College), talked about "Subject Access Education: Oxymoron or Obligation?". She said there are two aspects of subject analysis: 1) determining subject content, and 2) expressing it in one of several ways. It is an essential part of the organization of knowledge, which library schools are obliged to teach. Users expect subject access, and if they rely entirely on descriptive cataloging for it, they will usually not find what they need, since title words are often not useful for subject access. One of the most important parts of subject analysis is determining the subject content of an item, and this can't be taught. We lack a coherent, unified theory of subject organization, and there are rules that limit subject access (for example, using subject headings only for topics that make up at least 20% of a book, and omitting contents notes; these are both based on a card environment).

There is very limited time for teaching people anything in library school; ideally, concepts should be taught first, but there isn't time, so tools are taught first, then rules, because these things are practical and teachable. Things that cause problems for students trying to learn subject analysis are: 1) the idea, surprisingly prevalent, that there is one right number and one right descriptor for a given item, 2) lack of subject expertise, 3) lack of coherence in the tools, and 4) the pretense that LC knows best, now being vigorously upheld by the utilities.

A beginning cataloging course is really a catalog appreciation course; it should aim to persuade students to take advanced cataloging, and those who don't should emerge from the beginning course with an understanding of the reasons for and value of cataloging.

Responses to the crisis include: 1) use of technology which makes it possible for users to get information directly, without the mediation of a librarian, thus subject analysis should be based on the needs of users since there are far more users than librarians; users don't know LCSH, and neither do reference librarians, 2) users and search strategies vary; LCSH minimizes flexibility, 3) subject analysts behave as if they had only two choices: LC or nothing, but they should be doing anything they think will help users (e.g., contents notes, Dewey numbers), and 4) one should not expect information to fit outmoded simple preconceptions.

Final considerations: 1) subject analysis exhibits some unreasonable dichotomies: some librarians say subject headings are useless, while others clamor for subject headings for fiction, form and genre terms, etc., 2) just because some administrators say we don't need subject headings doesn't mean it's true, and 3) it is unreasonable to insist on using only one kind of subject analysis, considering the variations in materials, users, etc. We should consider using different thesauri for different materials, sometimes more than one at a time. Some people need deep indexing and multi-level classification, while others need only keyword searching.

Thomas Mann, (Library of Congress), spoke about "Cataloging and Reference Work". His first topic was the continuing need for subject classification of books (i.e., for subject arrangement of books on shelves). He gave two examples of information that could be found only by taking books in a particular subject area off the shelves and looking through each one for the relevant information. The information exists in these books at the page and paragraph level, and this kind of searching could not be done if the books were not organized on the shelves by subject. Scholars, students, and journalists use this type of search quite often, but librarians generally ignore it or say that it is unimportant (partly because it can't be computerized, and some librarians think anything that can't be computerized is unimportant). The quality and level of research that can be done in libraries would be greatly diminished if this kind of searching became impossible.

Mann's second topic was the importance of specific entry in a controlled vocabulary. Use of the most specific entry is being abandoned because of the increased use of copy cataloging; general headings are being accepted in place of specific ones, and this leads to disaster. The items are effectively lost, because one never knows where to stop with general headings (since all general headings are potentially applicable), whereas with a specific heading, one stops when one finds the heading that fits most closely with the subject one wants> If works dealing with this subject all had the specific heading, one could then be sure that one had found all the works in the library on this subject.

The third topic was that the crisis is mainly due to reference and bibliographic instruction librarians, who are not telling users how to use the retrieval systems created by catalogers. They should tell users about the red books, about the importance of Narrower Terms (NT, including those that are alphabetically adjacent to Broader Terms (BT) as these cannot be found in screen displays), about the usefulness of subject headings from records for relevant items located by author, title, or keyword for finding similar items. (Of course, this will not work if the headings are at the wrong level of specificity!); and about the subdivisions of subject headings. Some bibliographic instruction librarians are telling users not to use LCSH, so the users are missing many--sometimes most--of the relevant items. If the retrieval system is going to work, reference and bibliographic instruction librarians have to explain how subject headings work, rather than concealing or even disparaging them. Michael Gorman, (California State University--Fresno), talked about "The Cost and Value of Organized Subject Access," saying that systematic subject access is the key to effective use of libraries, and it is therefore both cost-effective and cost- beneficial, even though many administrators don't think so. But there are problems, both inherently and in application. Good subject access maximizes both recall and relevance. Specificity is extremely important; it best meets the needs of most users, because the cataloger has already differentiated the items. It is also extremely important that a verbal subject system have a syndetic structure, so that the user can explore broader, narrower, and related subjects. The time spent by the cataloger in creating subject headings should be inversely proportional to the time spent by the user on retrieval; the canon of service of our profession demands adding that value at the front end instead of shifting the burden to (infinite numbers of) users. Direct and indirect benefits to the user increase with the amount of time spent on subject headings; if we believe that the whole purpose of a library is to make its collection accessible, we can't afford not to provide detailed access to collections. Effective retrieval is impossible without authority control (which however is free, since it is just cataloging done right).

Gorman contrasted the "howling desert" of the Internet with the well-ordered world of libraries, comparing the Internet to a used bookstore in which the bindings, indexes, and front matter have been removed from all the books and they are arranged in no order. The user searches for clumps of related material, but has no idea of its source. It may seem ordinary to go into the largest library and be able to find a specific item, secure in provenance and immediately usable, but this is beyond the wildest dreams of Net-surfers. We need fast and efficient access to recorded knowledge and information, because we have lives to live and can't spend time surfing; subject access is an essential part of this, and is vital for future seekers of truth.

The program on "Rare Book Cataloging for the Generalist Cataloger" was sponsored by ALCTS Cataloging and Classification Section; ACRL Rare Books and Manuscripts Section on June 26, 1995.

Eric Holzenberg, (Grolier Club of New York), discussed the definition of a rare book, which can vary from one collection to another, and the hybrid nature of such a book as both text and artifact. The part of rare book cataloging dealing with the text is often not very different from ordinary cataloging, but more notes and added entries may be needed than AACR2 would prescribe. A much more detailed description of the physical aspects of the book is usually needed to distinguish different editions, issues, and states; this is especially important in a shared cataloging environment, to help other catalogers to identify the item they are trying to catalog. The objective is a description of the ideal copy, and a careful description will help other catalogers to realize that their copies may have some parts missing. Notes on the binding and marks of provenance may also be important.

He then discussed an example of a title page of an early book, and how it would be transcribed using, respectively, AACR2 and DCRB. This was extremely useful in showing the differences between the two sets of rules and pointing out information that would be lost by using AACR2, which could be extremely important to users of the records for rare books. He emphasized the importance of always trying to find a published description of a book, since this can save the cataloger's time and also provide valuable information about parts of a book that may be missing from the copy in hand.

Elizabeth Johnson, (Indiana University), described the ideal rare book cataloger as someone who 1) is interested in cataloging rare books, 2) knows languages or is willing to learn, 3) has subject expertise, mainly in order to be familiar with the appropriate reference sources, and 4) has a high tolerance for detail even for a cataloger. She added that any investment in training and education in rare book cataloging will be rewarded, for both the institution and the individual. She mentioned several institutions and organizations that offer workshops and courses.

She then discussed other technical services issues as they relate to rare books: security--generally more important for rare books than for ordinary items--and preservation, including handling, since it is important to handle rare books carefully and to teach other people to do the same, climate control, disaster preparedness, pest control and dust abatement, light (many rare book collections are kept dark when no one is in the stacks), processing and shelf preparation (emphasizing the desirability of non-intrusive procedures), repairs (apart from some simple paper repairs that anyone can learn to do, it is best to leave torn pages, Scotch tape, etc., alone unless one has access to a professional conservator), and what to do with objects inserted into books (letters, clippings, pressed flowers, etc.; she recommended discarding the flowers unless one is working for the Hunt Botanical Library, and putting paper items into a separate enclosure and describing them in a local note in the bibliographic record).

Johnson concluded by saying that a rare book cataloger needs to develop a sense of when to stop (i.e., of when the record contains enough information and detail), and by pointing out that rare book catalogers should be included in discussions of new library automation systems or of changes in existing ones, because they are often the only people aware of how information in rare book records displays in the OPAC. The library has probably spent more on acquiring and cataloging rare books than on ordinary books, so it is worth providing adequate access to their bibliographic records.

Linda C. Ewbank
University of Pittsburgh


The ALCTS Serials Section Committee to Study Serials Cataloging met June 25, and 26, 1995. In the following, a summary of the meetings minutes; with complete minutes available from ALCTS.
On June 25, the Committee approved revisions to Guidelines for ALA/ALCTS/SS Committee to Study Serials Cataloging Minutes of Meetings. These revisions corrected minor errors and specified a more timely electronic distribution. The Committee also considered possible programs to sponsor at the 1996 LAMA/LITA conference. Next, Committee and audience members discussed the role and functions of the Committee, identified as the following: to serve as a channel to move proposals into the queue for possible rule changes; to respond to questions; and to identify broad topics for general discussion, with the prospect of identifying proposals for changes in rules and practice.
On June 26, the Committee heard reports from liaisons from LC/NSDP and CC:DA, and a report concerning Serials Cataloging Institutes. Then Crystal Graham read a paper, "What's Wrong with AACR2: A Serials Perspective," previously presented at the AACR2000 PreConference. Graham identified structural, philosophical, and practical difficulties in the use of AACR2 for cataloging serial publications. She explained that serials records serve as bridges to related information, a function grown more essential with the appearance of automated indexing and abstracting tools which require links to library holdings. What we need, she said, is a code that focuses on relationships and distinguishing features of serials "families." Committee and audience members discussed Graham's paper and proposed topics for future meetings. These included serials cataloging for computer files and other materials, conference proceedings, mega-serials, management of serials cataloging, second phase of format integration, use of serials cataloging information by local systems, and the fundamental question of whether we need bibliographic records if patrons require only article-level information.

Carolynne Myall.
Eastern Washington University

Publication announcements

The three documents listed below, issued by the Network Development and MARC Standards Office of The Library of Congress, are now available from Library of Congress:
Format Integration and its Effect on the USMARC Bibliographic Format (1995)
Update No. 1, March 1995, Appendix F: Format Changes
USMARC Specifications for Record Structure, Character Sets, and Exchange Media (1994)

Back to Vol. 22 Nr. 2 | CCQ Homepage | Tables of Contents | Informaworld |