Cataloguing for Special Libraries: Administrative Perspectives

Joni Lynn Cassidy

President, Cassidy Cataloguing Services, Inc.

Administrators and Catalogers: A Bid for Mutual Respect

To the bottom-line administrator, cataloguing may seem to be on the fringes of librarianship. To the informed librarian, cataloguing is seen to be at the heart of librarianship; the central core of everything librarians do.¹

Michael Gorman, co-author of “Anglo-American Cataloguing Rules, 2nd ed.”

Today's economy and corporate politics demand respect for the bottom line in all types of libraries and information centers. Anxieties grow. Catalogers worry about compromises to the quality of their work while administrators search for ways to contain costs. But, do these groups have mutually exclusive goals? Within reason, I think not. In all library environments, there are ways to reduce cataloguing costs without sacrificing the essential quality of the work (the essential quality of cataloguing being its ability to get the researcher to the appropriate library holdings).

When all the costs of record creation and maintenance are considered, the database is the most expensive part of any online library system. But, it can also be the most durable, outlasting many generations of hardware and software. The durability and life expectancy of machine-readable records is directly related to the extent to which they adhere to standards.

The standards that apply to bibliographic records are generally considered the domain of catalogers. Yet this knowledge is essential to the design and implementation of online systems, as well as to long-range decisions regarding the staffing and procedures of the technical services department. From that perspective, an understanding of bibliographic standards and their application is equally important for administrators. In a recent Library Journal article entitled “The corruption of cataloguing,” Michael Gorman states, Computerization makes quality records easier to use and makes new ways of using them possible; it does not eliminate the need for that quality.² Mutual respect for the standards, and the quality they ensure, could serve as a basis for the formation of common goals between these two groups of library professionals.

A Modest Proposal for Compromise

STEP I

Be willing to accept existing records. Local notes and accommodations can be easily added to standard MARC records. This process is obviously much faster than creating original records. Records can be accessed from many different sources. If you have a professional cataloger on staff who can customize, edit and upgrade records, your sources include the bibliographic utilities (such as OCLC, RLIN, CATSS), downloadable CD-ROM products (including Gaylord's SUPERCAT, CDMARC Bibliographic from LC, Bibliofile, OCLC's CAT CD), record vendors (such as Marcive, FastCat and RetroLink), and, most recently, book jobbers (such as Blackwell N.A.). Records from these sources may require cataloger intervention to bring them up to current bibliographic standards, as set forth above.

THE EDIT/UPGRADE FUNCTION WILL INCLUDE

1. Check the call number to make sure it is still valid and confirm that it places the title in an appropriate place in the library's collection; correct where necessary;
2. Check subject headings to see if more specific headings have been established since the original record was created;
3. Check the authority file to confirm the form of corporate headings; correct where necessary;
4. Confirm all series headings; correct where necessary;
5. Upgrade all punctuation and form to adhere to AACR2 Rev.;
6. Check the fixed fields for missing elements (such as date type, dates, type of record, etc,) that provide internal sorting information to the online system, Upgrading and customizing could also be performed by a library paraprofessional, with professional supervision and review.

STEP II

Customization of catalog records can include such steps as: (1) add specific holdings notes, (2) add location codes, (3) add in-house subject terms, (4) add contents notes, (5) add other local notes such as access information. The library must determine if the level of customization being performed is actually necessary for all items cataloged, and if the institution can afford such detail. An examination of technical services procedures and workflow would net a more cost-efficient department and outline the items which require this level of customization.

STEP III

If you don't have a professional cataloger on staff to customize/edit and upgrade records, consider using a cataloguing service such as Special Libraries Cataloguing, Cassidy Cataloguing Services, and OCLC's TechPro. Many regionally-based cataloguing services are also available, such as Lundahl Enterprises (Chicago); Jenkins Law Library (Philadelphia); SIMA (Washington, DC); Costabile Associates, specializing in government libraries (Washington, DC); and Library Specialists, Inc. (Atlanta). They are staffed by professional catalogers who are responsible for meeting current bibliographic standards and keeping up with the changes in the industry. They should be able to customize your records to your local specifications-If you go record shopping, remember that this is a product like any other. You will get what you pay for. A Hyundai is not the same as a Lexus, and neither are their price tags or repair records.

Quality control issues and subject expertise will play important roles in cost determination and evaluation. But, he library is always in control of the final product by monitoring adherence to the standards and writing the specifications for local customization. If the library is not happy with the product, determine the source of the problem and revise the service contract accordingly.

STEP IV

Seek out and consider joining coordinated cataloguing projects in your region. As an example, the Cleveland Law Library joined a consortium of 16 area public libraries in which the Cleveland Public Library provides cataloguing for all members using OCLC. The librarian at the Cleveland Law Library commented that she sometimes had to correct records because the catalogers at the public library were not familiar with law materials. But this process took much less time than full cataloguing would have.

STEP V

There will always be a small percentage of new titles requiring original cataloguing, Try not to let these pile up. It's better to invest a few hours a month dealing with what catalogers call the “ugly titles” than to let them accumulate and form a backlog made up entirely of problems. If the library is acquiring some of the more routine records from outside sources, its professional catalogers will have the time to devote to original cataloguing.

STEP VI

Supplements, transfer binders and other forms of updates can be cataloged separately, but they don't have to be. Alternately, the cataloger can add a note to the record for the main volume(s) such as “Kept up-to-date by cumulative supplements” (pocket parts, revised volumes, etc.). In many cases, the actual item-level information will be recorded in the check-in file or circulation module of the online system. It may not need to be duplicated in the bibliographic record.

Anatomy of a MARC Record: What Everyone Needs to Know

Library automation is the common ground where administrators, reference professionals and technical services professionals meet.

When shopping for an online system, be sure to discuss the issue of MARC import/export with the vendor and make certain that you fully understand the sales representative's answer. Realize that it is often their job to mask the whole truth about their product, making it your job to probe for complete answers.

Originally designed to enable the Library of Congress to a print catalog cards more efficiently, the first generation of MARC was called MARC I. That was before formalization of format, but after pilot testing. The next generation, introduced in 1968, is called LCMARC, MARC II, or USMARC II. Just to make life interesting, there is also OCLC MARC and RLIN MARC.

The most important difference is in the record storage. USMARC storage capacity is divided into physical blocks of 2,048 characters, each block being packed with RC data. A new record is not required to start a new block, thus maximizing the amount of data that can fit on a tape. OCLC and RLIN write records to tape in an older format that requires each record begin at the beginning of a new block. Space between the end of the record and the beginning of the next block is not used. Other differences occur in the definition of a number of local RC fields.

Figure A is probably most familiar to librarians who've used OCLC or RLIN. It is the RC display, showing the MARC tags at the start of each new field. 245 is the title field; 260 the publisher, etc.; 600's are subject headings; 700's are added entries. Many librarians, even catalogers with years of experience, think the presence of those tags on the left constitute a MARC record. This is just the DISPLAY, designed to make it easy to read. Figure B is the USMARC II communications format. As you can see, those same tags have been stripped off and moved to the top to build an intricate directory of this record. The first 24 positions make up the “Leader.” It tells us the size of the record measured in the number of characters; whether it is a new record or one that has been changed; the type of record and the bibliographic level. The remainder of the leader positions are internal controllers, with some variation from OCLC to RLIN to US RC.

The rest of the directory is made up of 12-position segments, each representing a MARC field. Look at the segment that is circled at the end of the third line. The first 3 numbers are the MARC tag = 538. The next 4 numbers tell us how long the field is = 0314. The last 5 numbers tell us the starting position of this MARC tag in this record starts at position 00578.

It's the ability of a piece of software to create, or re-create, this directory that determines if it's truly a MARC product or not. Some software packages can import a MARC record and convert it to it's own proprietary format. But not all of them can export a MARC record, re-creating that leader and directory. If you ever plan to move your records into a different piece of software, you have to know if the product can EXPORT a RC record, not just import one. Be aware that even if you don't plan to move your records, it will probably come to pass sometime in the future.

If the library purchases an online system that does not have RC export capabilities, make every effort to store a copy of all records used in the online system in USMARC II format for use with future generations of software. This is a service provided by many cataloguing services, record vendors and bibliographic utilities. Be sure to consider the maintenance of holdings, and other local information.

Retrospective Conversion

Librarians sometimes use the phrase “retrospective conversion,” or “recon,” to mean more than one operation. In its purest form, “retrospective conversion” means taking your paper shelflist and converting it to machine-readable format.

Depending on the state of the paper shelflist, the library may also want to use this opportunity to upgrade to AACR2 standards, or flip to LC classification, or update subject headings. In some cases, “retro” can mean going from no catalog to a new online model in one big step.

Going into a recon project, the first question to be answered is: Does the library want to convert the existing shelflist to MARC, or upgrade the records in one way or another in the process? Once that is decided, it's time to write the specifications for the project.

The next step is identifying companies that provide the services the library is shopping for. This will be like shopping for a car. Compare each company's service options, prices and staff qualifications. Be the enlightened shopper for your organization and realize going into the process that a $46 MARC record is not the same as an $8.00 MARC record. Find out what the differences are so that the library can make intelligent decisions about what is needed.

The library has control over the most important element of the recon project: the specifications for the selection of records. Consider the following when setting parameters:

1. Choice of call number
2. Choice of subject headings (LC, MESH, Sears, other)
3. Acceptance of pre-AACR2 records or not
4. Conversion of in-house notes from the paper shelflist or not
5. Other criteria, such as library of input

As with everything else in life, in recon, you get what you pay for. Know what the library wants BEFORE going shopping, or at least before making a purchase. That way, disappointment will be avoided later. If the library selects a small recon company, they may be able to come back to you in the middle of a project and say, “You didn't say that you wanted to keep library retention notes on your serials records, but we thought we should stop and ask you.” That gives the cataloger an opportunity to slap his/her forehead and say, “Oh, thank you for asking. Of course the library wants to transfer that information to my new records.”

But, be realistic. Don't expect big recon vendors to be able to read minds. If it's not in the specs, it probably won't be in the catalog. For the most part, companies working on a large scale usually have very few professional catalogers on their staff. The professional decision-making and supervision of the project must come from the library.

MARC Format Integration: Applications to Online Systems

The essence of format integration (FI) is that ALL FIELDS ARE VALID IN ALL FORMATS. For. example, fields such as 310 (frequency), 780 (previous title), and 785 (new title) can now be used in all formats, not just in serials. This enables the cataloger to describe the serial nature of all types of materials. Some online systems are written to import only certain fields from certain formats and you may need to talk to your vendor about upgrading for FI. Earlier phases of FI made certain fields obsolete. However, those fields are still present in records created prior to FI. Your import profiles will have to accommodate them for this purpose.

The major changes, as a result of Format Integration, phases I and II, are as follows:

1. Title added entries can be tagged in three different ways: 246, 730 and 740, This isn't “new,” but the way these fields are being used is “new.”
2. 246 @ Addition of subfield “i” preceding subfield “a” to display text. (EXAMPLE: 246 0 $i At head of title :$aState of New York). This option can be used if none of the terms generated by the second indicators apply.  Find out what your software will do with the subfield “i” information and how it will display. Find out if your software is programmed to generate the appropriate terms from the second indicator.
3. Some notes formerly tagged as 500 (general notes) in the books format are now coded with specific tags.
4. 505 @ Addition of subfield “t” for TITLE and subfield “r” for STATEMENT OF RESPONSIBILITY in the enhanced contents note, (EXAMPLE: 505 00 $t History of chocolate $r H. Hershey - $t Candymaking made easy / $r F. Nestle,) Can your software extract the subfield “t” information and index it with other titles? Can your software extract the subfield “r” information and index it with other authors?

The final stage of FI was the merger of the seven existing formats into one consolidated format. The fixed fields (008 include a character that the cataloger uses to designate the primary type of material.

The most important new MARC field is the 006. This will be used when an item crosses the old format boundaries; in other words, it could be described by more than one format.

Moving Forward: On Firmer Ground

The Guidelines for Technical Services in Private Law Libraries³, approved by the American Association of Law Libraries Private Law Libraries Special Interest Section in 1994, could be adapted to provide a foundation for the establishment of common goals between library administrators and technical services professionals working in all varieties of special libraries.

Guidelines relating to cataloguing are as follows:

1. The cataloguing function in small, medium and large libraries should be performed or managed by an on-staff degreed librarian or through a qualified cataloguing service.
2. The Library of Congress classification scheme and subject headings should be used with or without modification in small, medium and large libraries…
3. Small, medium and large libraries should have access to a bibliographic utility.
4. The current national catalog code should be followed in the provision of descriptive cataloguing in small, medium and large libraries. At present, this is AACR2R, which may be supplemented by use of the Library of Congress Rule Interpretations. In an online environment, it is strongly suggested that software which is compatible with MARC bibliographic records and which follows the MARC standards that are detailed in the USMARC Format for Bibliographic Data should be chosen. Any MARC standards followed by the bibliographic utility in use, such as OCLC, RLIN, WLN, etc., should also be followed.
5. Authority control should be provided within the local catalog of small, medium and large libraries. (This is implied by adherence to AACR2R. At a minimum, this means that the form of each heading will be uniform so that all titles properly collocate. At most, it could mean that a structure of cross references and authority notes are provided in the catalog for the user. Authority work is essential to a catalog, especially as it grows. Split files work against the user's location of all pertinent titles).
6. All libraries should maintain some kind of procedures/decisions file to ensure continuity of choice in cataloguing. Use of national catalog documentation such as AACR2R, Library of Congress Rule Interpretations, USMARC Format, Library of Congress Subject Headings, LC Subject Cataloguing I Manual, Library of Congress Classification… may substitute for a local cataloguing manual, if national standards are followed. Any local variations on such national standards should be recorded in the local manual.

A Personal Perspective on the Guidelines

To put the above into perspective, let me explain briefly what I do for a living. I founded, and now co-manage, a company that provides cataloguing to 68 libraries, most of them law collections. For many of these libraries, I manage their cataloguing and processing in lieu of an in-house technical services librarian. It is my responsibility to advise the head librarians on technical services issues, such as automation, and help them budget and plan for the future. In a very real sense, my staff and I function as part-time members of each library's staff.

I have been part of every group of technical services librarians who have worked on the development of the STANDARDS FOR LAW FIRM LIBRARY TECHNICAL SERVICES, including the first informal after-dinner discussion, which basically amounted to…“So, what do we think about this idea of standards?”

I wanted the STANDARDS to simply reflect the message “Do it right, or we put you in jail for malpractice.” While there were others who shared this sentiment, intellectually, we knew that any call for standard operating procedures would have to be supported by statistics documenting a movement towards adoption of such standards in libraries such as our own.

It was also possible to compile and issue STANDARDS, and justify their implementation as a matter of professional ethics. In 1992, Special Libraries Association published Malpractice issues in librarianship: an SLA information kit. It included the following statements: “The situation of the in-house librarian whose sole patron is the organization who employs the librarian is different from that of [public] reference librarians whose patrons are not their employers. The in-house librarian has an employment contract with the information user. A failure to perform satisfactorily may amount to a breach of the employee's duty of care inherent in the employment relationship. If so, technically under common law, the employer has a common law cause of action against its employee for any damages suffered due to its employee's negligence. […] There is a misconception that the non-licensing of our profession somehow protects us from liability, i.e., that a lack of formalized standards means that we cannot be held accountable for not meeting them. […] The information profession has not articulated standards by which to judge […] performance. We must define acceptable practice for information providers, not only in terms of ethical-standards, but also by establishing procedural guidelines to ensure the quality of our services.”

The American Library Association Code of Professional Ethics tells librarians that they “must provide the highest level of service.” Following ethical conduct, the next component of a professional's duty is the requirement to exercise reasonable care, skill and diligence in the rendering of professional services. Problems related to the lack of reasonable skill are more difficult to define than ethical problems. But, what the standard courts use to judge whether or not a professional breached a duty is what a member of the profession in good standing would have done under the same circumstances.

Thus, our goals became crystal-clear: (1) Survey the private law library community to see what “members in good standing” were doing in their technical services departments, and (2) use that information to establish a set of “Procedural guidelines to ensure the quality of our services.”

The first technical services questionnaire was written by, and distributed to, members of the New York Technical Services Roundtable. The main purpose was to analyze the responses and refine the questions before distributing the questionnaire on a national basis.

After more re-writes, this time by AALL's Private Law Libraries Technical Services Division, the national survey was distributed through all AALL chapter newsletters and PLL Perspectives, the official newsletter of the Private Law Libraries Special Interest Section.

National survey results were tallied and results published in PLL Perspectives. With these results in mind, an Advisory Board of academic, private, county and independent law librarians, led by chairperson Suki Scott, was created to develop a list of standards to propose to the PLL Board. Each guideline reflects national bibliographic standards and actual practice in a majority of the libraries surveyed.

The “Standards” were re-named “Guidelines” to avoid the impression of strong-arm tactics. The now-named Guidelines for technical services in private law libraries were presented, discussed and amended at a technical services town meeting held at the AALL annual meeting in Boston in summer 1993.

By winter 1994, finalized Guidelines were submitted and approved by the PLL Special Interest Section Executive Board. They took three years to complete.

Summary

A good classified catalog or subject index is practically indispensable, particularly as a reference guide to legal treatises, monographs and governmental publications. […] Minus a good catalog a library loses much of its efficiency, for much valued material remains unknown to library users.

Lawrence H. Schmehl, law librarian for the New York County Lawyers Association, 1959

There is a definite need for cataloguing standards when you've reached the retro/automation stage. The most concise argument for the adoption of standard cataloguing practices was written by Automation Consultant Richard W. Boss in the third edition of his book, The Library manager's guide to automation:

The database, made up of cataloguing records, is the most important part of an automated library system. While hardware may be replaced from time to time (the norm is every 5 to 7 years) and while software is periodically rewritten to transfer it to other hardware, improve performance, or enhance it functionally, the bibliographic files, if well designed, will outlast several generations of hardware and software.

The machine-readable bibliographic file is analogous to the 100% rag cards that libraries use in their card catalogs. Those cards are intended to last indefinitely and to be transferred to new card catalog cabinets when the old ones are replaced. Similarly, the bibliographic file represents a considerable investment and should be of high quality and transferable.

The most cost-effective way to approach a retro these days is to access existing standard records and edit them to fit your needs. This can be done in a variety of ways, other than through OCLC, which is really a very expensive choice measured against the alternatives. A selection of sources (CONTACTS) for standard catalog records that can be imported into your online catalog software and edited for local use follows this article. Any cataloger will tell you this is a much faster process than original cataloguing; and, that using any kind of in-house system (classification, subjects, or description) means essentially that every title must be originally cataloged.

The cost of original cataloguing has been estimated at $50 per title by Harry Martin of Harvard University, and $75 per title by the Library of Congress. Purchasing existing records should cost between $6 and $25 per record, by comparison.

Another consideration from the administrator's point of view of is that any in-house system will prevent you from being able to use outside labor on a temporary basis if you lose your cataloger due to illness or job mobility. My company had three calls at the beginning of the summer from libraries that, for one reason or another, needed temporary cataloguing help for a short time. We were able to service two of them, but the third didn't use LC classification or subjects and we were not able to help them out because all our billing rates assume that work done for any library can be added to our master database for possible use in the future. Anything non-standard can't contribute to the master database and thus creates prohibitive costs for the client.

In conclusion, an administrator's awareness of the expanded cataloguing options available to an organization's library through the use of the above suggested standards and varied records sources, should allow for an atmosphere of high-quality and cost-effective information access decisions.

References

1. Gorman, Michael, The Corruption of Cataloging, Library Journal, Sept, 15, 1995, Pg. 330
2. Gorman, Michael, The Corruption of Cataloging, Library Journal, Sept. 15, 1995, Pg. 340
3. Managing the Private Law Library 1995: Affecting the bottom line. New York: Practicing Law Institute, 1995. Pgs. 193-210