Throughout the history of the world, there have continually been attempts made to record culture by people who saw the importance of preserving the knowledge of their own civilization. The degree and nature of this activity varied as did the people, and have ranged from simple cave paintings to more sophisticated systems of classification seen in ancient institutions. Cultural relevancies and temporal developments have always been deciding factors in determining the kinds of subjects compiled and the fashion they were catalogued. One thing that has remained constant is the fundamental fact that the universe of knowledge is forever expanding and must be preserved. Currently, examples are the growing fields of Genetics and Computer Technology.
Library catalogues have a very extensive history and can be traced back to the libraries of Antiquity. In the 7th century B.C., important libraries in Mesopotamia had author and title catalogues that were posted on walls for user convenience. In the 3rd century B.C., Callimachus, Greek scholar and chief librarian of the Alexandrian Library, compiled a huge catalogue for the library called the Pinakes. Catalogues have changed dramatically over the centuries, having appeared in many formats, including clay tablets, papyrus scrolls, printed books and cards, microform, and the online versions that are prevalent today.
Computers have changed library environments and revolutionized information retrieval by allowing additional points of entry and quick changes in the system. However, they are quite expensive and with the economic realities affecting libraries today, keeping up with ever-changing technologies will be difficult. Coupled with the explosion in the publishing world, there is more and more out there to catalogue. From the mid-1800s to the present, preoccupation with the cost of cataloguing remains an underlying concern. The price of keeping track of and organizing the world of knowledge prompted Charles Jewett, Librarian and Assistant Secretary at the Smithsonian, to propose the idea of stereotyped plates, which were a series of preserved, mass-produced separate titles composed in adherence to a set of very strict rules (Jewett 1853,61). Although Charles Cutter feared for the loss of the art of cataloguing, he could see the economic advantages for libraries if they used the Library of Congress printed cards (Cutter 1904,65). Original cataloguing is time-consuming and very expensive.
The last 150 years have seen changes in the ideological purpose and organization of library catalogues. They have gone from lists of books containing limited amounts of information, to globally-interconnected indexes with vast bibliographic data. The passing of time has witnessed changes in cataloguing practices and technological evolutions, however, the proliferation of published knowledge and the cost of cataloguing are still problems libraries deal with. One solution has been to use copy cataloguing. This essay will present some of the problems associated with using cataloguing copy, and the kinds of policies and procedures that cataloguing departments must have in place when evaluating and accepting records from a shared database.
Charles Jewett strongly believed that a national library was needed, which would oversee a collective catalogue for libraries across the United States. His dream was that it would become an international organization built on sharing and cooperation (Jewett 1853,52). Such a library would permit scholars access to important books and manuscripts; show where intellectual disparities existed; and act as a facilitator for the progression of knowledge (Jewett 1853,54). In Catalog it Once for All: A History of Cooperative Cataloguing in the United States Prior to 1967 (Before MARC), Barbara B. Tillett writes that:
One should not lose sight of a larger goal to provide our users with access to the bibliographic universe while inventorying the holdings of our libraries. Cooperatively creating catalogs was seen as a way to reach these goals (Tillett 1993,4).
In 1876, many cooperative cataloguing enterprises began to appear, one of which was The Library Bureau, and it began to sell printed catalogue cards (Tillett 1993,5). The earliest issues of the Library Journal contained American Library Association (ALA) discussions on the subject (Tillett 1993,6). Melvil Dewey proposed that the ALA lay down some guidelines for cooperative cataloguing, which all libraries could follow, possibly filtered through the Library of Congress (Tillett 1993,6). Catalogue rules devised by Jewett and Charles Cutter were fundamental in successfully bringing the dream of a collective catalogue into fruition.
Over the next twenty-five years many ALA member libraries entered programs that shared cataloguing information. During the Depression and throughout the war years, materials were in short supply and very costly. Libraries not only shared bibliographic information, but also library materials (Tillett 1993,8). Jewett's vision of an interlibrary loan system took hold, allowing users to locate books in the country and giving them the ability to borrow them. In the 1890s, the Library of Congress took over the role of copyright depository. By this time, the Library of Congress printed catalogue cards were on the market and changed the direction of cataloguing in a very meaningful way.
As budgets got tighter in the early 1970s, and space for library storage became limited, there was an escalation in shared purchases and interlibrary lending (Tillett 1993,8). This brought the problems still hampering library sharing to the forefront:
Interlibrary loans and collection development require bibliographic control for the identification and location of materials in the bibliographic universe. Bibliographic control in the form of traditional cataloging is expensive, so how can libraries provide bibliographic control at a reduced cost? (Tillet 1993,8).
The same issues plaguing libraries since the 1800s are continuing to concern library administrators. It may be just a pipe dream to think users will have the bibliographic universe at their fingertips, but individual libraries still need to deal with the hundreds or thousands of books and other materials that come into their own institutions every year. Original cataloguing for all of these items is an extravagance most libraries can no longer afford, as they use their shrinking budgets in more creative ways. Copy cataloguing for most items is the answer to reducing the amount of cataloguing that must be done. In Developing New Roles for Paraprofessionals, Mary M. Rider states that:
Librarians and administrators have begun to emphasize the increasing need for higher levels of cataloging productivity to offset shrinking budgets and reduced staff at many institutions. Efforts are also being made to eliminate or reduce large backlogs of uncatalogued materials that have been accumulating in many larger libraries (Rider 1996,27).
Automation of library cataloguing departments became the way libraries could begin sharing catalogue copy and rid themselves of backlogs (Smith 1994,1). Emerging from this transition was a splintering of the library community as one side advocated using copy cataloguing to provide much quicker access for users, while the other side insisted that striving for perfect records was more important (Smith 1994,1). Tillett discusses Andrew Osborn's theory that much of the problem came from the critical nature of cataloguers and how they viewed their colleagues' work habits (Tillett 1993,29). Osborn pointed out that cataloguers were unable to accept the near-perfect records of the Library of Congress. They felt compelled to modify them to the point that original cataloguing was the outcome (Tillett 1993,29). Backlogs developed because librarians did not readily accept records without extensively checking them first (Rider 1996,27).
Original cataloguing is seen as the ultimate in the library world. Many librarians believe the records are practically free of mistakes since librarians with MLS degrees do the lion's share of the work. How realistic is this theory when the majority of library school graduates have only a course or two in cataloguing and must undergo a rigorous period of training when they begin their careers? Paraprofessionals are given the same training (Rider 1996,29). The real differences occur in the types of training given and how diligent, dependable, and mindful a cataloguer is. A degree cannot make an individual careful, nor can it make a cataloguer commit to making user access their overall objective. Human inequalities and errors are the cause of discrepancies seen in shared databases, whose only measure is the quality of the inputted data. Regardless, there has been a definitive move away from the idea of a perfect database, on to an emphasis of meeting the needs of users. To do this, libraries have employed the help of outside vendors (Rider 1996,27).
Cataloguing costs libraries a great deal of money. In order for copy cataloguing to become an efficient and effective choice, close adherence to the rules laid out in the Anglo-American Cataloguing Rules 2 (AACR2) must be followed, and agreement on which elements should receive the most concentration from cataloguers. A crucial element is the identification of which AACR2 rules cause the most trouble because of ambiguity and overelaboration (Inter 1989,23). Since so much copy cataloguing is done by paraprofessionals, clarification is extremely important. They should not have to wade through some of the jargon found in AACR2; jargon that reminds some of us of legal briefs. Traditionally, copy cataloguers have adapted the catalogue records that member libraries have contributed to utilities such as RLIN and OCLC (Rider 1996,26). Most cataloguing units maintain a list of libraries they prefer to take copy from; such things as the incorporation of local practices, the numbers of errors, similar cataloguing practices, and how closely the copy comes to LC copy standards are some of the considerations (McCue 1991,66). LC copy is considered preferable for most libraries, and is more often selected when there is a choice (Taylor 1988,296).
There are currently many ways to get copies of cataloguing records to use in a library database. Two, which are not recommended, are to download them off the Internet or from a local catalogue. There are also CD-ROM products available such as the Library of Congress's Bibliofile, WLN's Lasercat, and the National Library of Canada's Amicus. In addition, vendors or book jobbers like Yankee Book Peddler (now called YBP Library Services) and Blackwell North America(1) sell books with the MARC records included. This is called 'ready cataloguing.' There are also bibliographic utilities that specialize in database products, including OCLC; SLIMS for small special libraries; Western Library Network; and ISM (formally UTLAS). Both good and bad records can be found in all of their databases. Yet, a comparative study done by McCue, Weiss, and Wilson found that LC copy was "relatively high in quality regardless of origin ... the study showed that there was no significant difference between LC and the best of the RLIN member libraries" (McCue 1991,66,74).
It's not as simple as saying that your library will buy cataloguing. There are many things to take into consideration, including finding a system that will "provide the best response rate for your library, and be at a price that can be afforded" (Taylor 1988,286). The other consideration is whether your library will give records back to the system for others to use. Some require libraries to reciprocate, which means extra time and extra work for staff. Since a library is only a storage facility for materials without a catalogue, libraries need to give special attention to what goes into their databases. The quality of a catalogue is how both users and other libraries judge the calibre of a library's excellence. If a library presents an inferior, mediocre, and non-integrated catalogue, the staff who work there are judged accordingly. Consequently, library administrators must make themselves aware of the pros and cons of each system.
Several problems can occur when using cataloguing copy. They include errors in MARC coding; varying forms of entry; problems with punctuation; lack of local practices; typographical errors; insufficient call numbers; discrepancies that cause serious problems (i.e. authority control issues); the separation/integration of series or conferences (Taylor 1988,328); and if cataloguing in Canada, copy could contain American Subject Headings rather than Canadian (i.e. Eskimo instead of Inuit). Overall, authority control problems are an ongoing concern. Additionally, there are many records in databases that have seen a variety of cataloguing practices and could be filled with rules and interpretations of rules that are no longer acceptable (Smith 1994,8). It's important for libraries to issue policy statements that require cataloguers to use the newest copy available, especially if it is LC, to take advantage of its updating system (Taylor 1988,332).
What cataloguers are looking for is the most appropriate record, using an exhaustive search process, while working independently. The goal is to recognize an appropriate record when they see it; this is what takes experience. What every library wants are perfect records, but they often settle for a compromise of sorts. Copy cataloguing is here to stay and libraries must have policies in place for the evaluation and acceptance of records.
Libraries need to do a lot of research before deciding on which cataloguing network to use. Asking other local libraries is a good place to start. A periodic review of the service should be required, with changes made if necessary. When libraries join a cataloguing network, they should have a set of guidelines already developed to save time and money (Taylor 1988,334). Will the copy cataloguing be done with CD-ROM products that are very cost effective, or will the most current records be made available online? A library will also decide whether or not to use unedited copy. If so, from which source or sources will the copy be accepted from? Will the library have a blanket policy of using LC copy without making any changes? It will also be necessary to develop procedures for handling copy cataloguing that requires editing to integrate it into the library's online catalogue. Since there is no point in using copied records if everything in them is going to be changed, it is more cost effective to determine which fields will be concentrated on and edited, and which can be eliminated altogether.
Identifying the fields in records that are commonly problematic usually involves changing access points and correcting errors (Taylor 1988,330). For example, depending on who has done the cataloguing, there will be more or less particular attention paid to certain fields. Since LC cataloguers have been trained to concentrate on the physical description and access points, and less on spacing, punctuation, and subject headings, their copy and corrections to other copy will reflect these tendencies (McCue 1991,71-75). In doing an analysis of catalogue copy in OCLC and RLIN databases, Sheila S. Inter found that:
Twenty-five AACR2 rule errors occurred in descriptive headings in OCLC's records; 29 occurred in RLIN's. These errors related mainly to the choice of main entry versus added entries. This relatively large number of errors relating to headings exhibits a startling lack of understanding of AACR2 Chapters 21-25 among cataloguing experts (Inter 1989,12).
Training is an important element in the equation, and time should be spent on figuring out the most important points to cover. One solution offered by Stephen S. Smith is that training should incorporate a mentorship program, like the one used at the cataloguing department of the University of Illinois at Urbana-Champaign Library. He cites the program as the reason the library was able to eliminate a backlog while it was going through a period of staff cuts (Smith 1994,3). Smith believes that mentoring facilitates "the sharing and understanding of cataloguing policies, procedures, and functions among graduate students, and novice cataloguers" (Smith 1994,6). Inter also makes some interesting recommendations, including: training should be done by those who teach cataloguing; there should be a clarification of AACR2 rules to help cataloguers better understand elaborations and exceptions (this could be done in a training manual, given to each cataloguer); punctuation should be emphasized; and the encoding of the MARC fields stressed (Inter 1989,19-22).
Changing copy is time-consuming and it must be checked. Who will do the checking and have the authority to make changes? Who will decide when and how much work will be carried out online/offline? Procedures must be developed for how authority work will be done on the copied cataloguing (Taylor 1988,311). Regardless, there should be rules established so that each cataloguer knows the library's preferences. Arlene Taylor has specified the following to be taken into consideration: 1. how much experience a cataloguer has if changes are to be done online, since time means money, and 2. the location and format of the local authority file, as well as LC's (Taylor 1988,311). Cataloguers also need to know how to handle incomplete, inaccurate, or incorrect information on a record. Training helps to focus cataloguers' attention on the most common mistakes, such as missing 300 fields or inaccurate edition statements; it should also make them understand why these changes are important, and how they affect the overall authority and usefulness of a catalogue. Taylor recommends that training include a requirement to read a record thoroughly before any changes are made, concentrating on typographical errors (Taylor 1988,298). Policies should include whether to integrate new records by making old ones agree with them, and the possibility of leaving both unchanged by using see also references instead (Taylor 1988,331). All of these options impact the quality of the catalogue and how a library's records will be perceived by other libraries using the network. They are, therefore, extremely important.
In Cataloguing with Copy: A Decision Maker's Manual, Taylor states that the next step is putting the policies in writing, using clear and specific points that can be referred to on a continual basis. This relieves a supervisor or trainer from having to repeatedly give detailed instructions (Taylor 1988,334). Finally, a library's cataloguing unit needs to be supervised by experienced personnel to be successful (Taylor 1988,334).
Copy cataloguing is a very complicated process that needs to be addressed by all libraries. It is here to stay and is bringing us closer to the possibility of international cooperation in cataloguing. If libraries can get used to sharing from databases within North America, then maybe the leap could be taken toward an international agreement of sorts. The days when very esoteric cataloguing was done are over, and cataloguers have to change their mindset and accept the impending changes that will assuredly come. New cataloguers will be trained with the expectation that more cooperative cataloguing can only help to keep expenses down, as will the sharing of resources. Currently, the world is becoming interconnected, with globalization of information a reality. Networks like the Internet are bringing great opportunities for universal understanding and tolerance. Libraries should play a role by making information sources accessible in a very standard and concise fashion - let us get the needed information out there, instead of complaining about the glory days. Copy cataloguing has inherent problems that need to be addressed, but there is nothing insurmountable in the proposition, that if achieved, would be a real and tangible beginning for worldwide accessibility of information and knowledge.
Related Papers
A Look Inside the Dewey Decimal System