Abstract
Serial catalog records must be interpreted and understood by a wide variety of catalog users: collection development, acquisitions, public services staff, patrons, and even catalogers themselves. This NASIG preconference workshop covered past and current cataloging rules as well as how the applications of these rules are recorded in the MARC 21 format. Key MARC fields were explained, and the presenters demonstrated how some OPACs handle serials-specific information to either the benefit or detriment of catalog users. Also discussed was the expanded definition of continuing resources and how the application of this definition affects cataloging decisions. New standards for next generation catalogs will need to take into account the FRBR (Functional Requirements for Bibliographic Records) conceptual model as well as the development of other metadata content and communication standards beyond both AACR2 and MARC 21.
THE TAO OF SERIALS
This lively preconference might just as easily have been entitled “Everything you ever wanted to know about serials,” so encompassing was its scope. The half-day session covered all major cataloging, collection management, and retrieval and access issues related to serial publications. Although aimed at non-catalogers, including vendors and library staff in collection development, acquisitions, public services, and systems, the workshop gave even seasoned catalogers a fresh perspective on the yin and yang of the new bibliographic universe.
WHAT IS A SERIAL (OR, WHAT IS NOT A SERIAL?)
With the publication of the 2002 revision of the Anglo-American Cataloguing Rules, 2nd edition (AACR2) came a new definition for serial-like publications. The former straightforward division of monograph or serial was no longer sufficient to characterize the types of resources encountered and cataloged in libraries. AACR2 now breaks down the bibliographic universe into finite and continuing resources, and within both finite and continuing resources is a newly defined category of material called the integrating resource. An integrating resource is one that is updated by means of integrating new or revised material into the whole, rather than by issuing successive, discrete parts. Although integrating resources have been around for many years in print form, for example loose-leaf publications in which individual pages are updated, reissued and filed into the base work, it was not until the emergence of the World Wide Web that defining this type of material became an issue for the broader cataloging world.
Before a record can be created for any given resource, it must first be categorized as a finite (single or multi-part) monograph (e.g., good old fashioned books, either single or in a set of volumes), a finite integrating resource (e.g., an online text that is intended to be updated but with a planned end point), a serial (e.g., a journal issued in discrete parts, whether print or online), or an ongoing integrating resource (e.g., a website that will be seamlessly updated and is intended to continue indefinitely).
To put it another way, how any given resource is cataloged depends not only on the nature of the resource itself (what it is), but also on the mode of issuance (the manner in which it is received). For example, an annual print directory published in successive volumes is cataloged as a serial, but when issued online as a continuously updated searchable database, it becomes an integrating resource. How the resource is cataloged also depends on the desired level of specificity or granularity. For example, The Environmental Protection Agency has a website with information about the Agency itself as well as links to online journals, electronic documents, and searchable databases. One could catalog the entire website (as an integrating resource), or any of its component parts (the journals as serials, the electronic documents as monographs, and the databases as monographs or integrating resources, depending on whether they are static or updating). The cataloger must know or determine (based on what the library has ordered or received) the specific resource and level of granularity that is being cataloged.
WHY THE DEFINITION OF SERIAL (OR NOT A SERIAL) MATTERS
What does the emergence of a new category of material mean for users of the catalog, and why do all these esoteric definitions matter? They matter because the coding of the online catalog record differs for a monograph, serial, or integrating resource and this coding can affect how users find, interpret, and even edit a catalog record. For OPAC users, the online catalog will index and display integrating resources as either monographs or serials depending on the specifications of the particular online system. This means that if you perform a “journal title” search, you may or may not retrieve many serial-like publications that have been cataloged as integrating resources. Also, the order of display for multiple records, and the labels used on individual records, can either aid or confuse.
For collection development staff, the definitions matter because determining the percentage of the library budget to be spent on “serials” depends on what you choose to call a serial, especially when determining what percentage of funds to use for electronic resources. For acquisitions staff they matter because the establishment of standing orders and other types of subscriptions may depend on the type of bibliographic record created in the database. And of course for public services staff, knowing how to interpret OPAC search results and being able to guide patrons to more useful search strategies is essential.
YET MORE LAYERS TO THE DEFINITION OF SERIAL
To add to the complexity, serials cataloging has its gray areas, those publications that, according to AACR2 and relevant Library of Congress Rule Interpretations (LCRIs), can be treated either as monographs or serials, depending on their frequency of issuance as well as the stability of their titles. Some intriguing examples of these include conference proceedings, travel guides, and blogs. The Library of Congress only catalogs conference publications as serials if each edition does not have its own distinctive title. Although an LCRI allows for most travel guides to be cataloged as serials, because of the frequency of their editions, access to different authors of the various editions is usually lost. The United States ISSN Center will assign ISSNs to blogs that are topical (e.g., a blog devoted to autism), but not to personal blogs that simply share people's private lives. When resources are sometimes cataloged as monographs and sometimes as serials, this adds to the confusion for all users of the catalog.
Electronic serials, particularly online publications, add yet another layer of complexity to our current cataloging environment. According to national practice, catalogers can either add information about an online version to an existing print record (single-record approach) or create a new record for the online version (separate record approach). Either method is acceptable, and libraries often have a mix of both methods in their databases because records are created in-house as well as received from outside vendors. This, of course, adds to the confusion for catalog users when they retrieve more than one record for the same work, or do not realize that a serial is available online because it shows up in a results display as a print title.
There is no simple solution to the OPAC retrieval and display issues for electronic serials. One can argue that a user should be able to link to all versions of a serial title from one record. One could also argue that the catalog record should unambiguously describe each serial format, and that the OPAC should retrieve different records for the versions of the same title together in a useful display. The emergence of online serials has only added to the debate catalogers have long held over multiple versions and how best to treat them in our catalogs.
HOW ARE SERIALS CATALOGED?
As with all catalog records, those for serials represent a mix of cataloging conventions that have evolved over the history of bibliographic control. Current national-level serials cataloging follows AACR2, the LCRIs, and policies and practices outlined in the CONSER Editing Guide (CEG) and the CONSER Cataloging Manual (CCM). AACR2, the LCRIs and the CEG address description only, not subject headings or classification. (The CCM does have a section discussing assigning subject headings for serials.)
The cataloging rules and guidelines contained in the aforementioned publications govern what information about a serial is included in a catalog record, how it is recorded, and from where within the serial the information should be taken. The rules also govern how the searchable headings are formulated and how entries should sort in the catalog. Serials are described based on first or earliest available issue, employing the principle of transcription, not interpretation, of the data found. Key descriptive elements, including the title and publication information, reflect what appears in the earliest available issue, and in most cases variations found on later issues are recorded in notes embedded in the record. The amount of information contained in an original serial record depends very much upon how many issues of the title are available to the cataloger at the time the record is created.
Records created according to AACR2 also follow the convention of successive entry cataloging. This means that if a serial title undergoes a major change, the existing record is ceased (closed out) and a new, successive record for the major title change is created. Linking fields in both the old and the new records connect the titles to one another and allow for them to be retrieved together in most OPAC searches. Successive entry cataloging, in effect since the adoption of AACR2 in 1981, is in sharp contrast to latest entry cataloging, the convention used in the first edition of the Anglo-American Cataloging Rules (AACR).
Under latest entry cataloging, key descriptive elements were based on the latest available issue; if title, publisher name or place, or any other significant information changed as evidenced by later issues of a serial, catalogers revised the existing record to reflect the latest issue. Instead of a new, successive record for title changes, the title field in the catalog record was updated with the current title, and title added entries were created for any earlier titles. It is another ongoing debate among catalogers (as well as public services staff) as to whether successive or latest entry cataloging better meets the needs of most catalog users.
In 2002, a major revision of AACR2 introduced the concept of major and minor title changes. The intention was to reduce the number of cases in which a new record needs to be created under successive entry cataloging. If a change is determined to be minor, added entries are made to the existing catalog record for the title variants rather than a new record created for the change. Some of the more common minor title changes include: abbreviated forms of words that were spelled out, changes in grammatical form (e.g., singular vs. plural), and additions or variations of articles and prepositions. Also considered minor is the removal or addition of the name of a corporate body associated with the serial, as in the case of the ONS newsletter becoming simply Newsletter. The problem with the last example is that the typical college undergraduate using the online catalog may or may not recognize this as the correct publication when he or she searches ONS newsletter in the OPAC and finds fifty plus serials with the title Newsletter.
THE MARC FORMAT AND HOW IT (SORT OF) WORKS
AACR2 governs the content of catalog records, whereas the MARC 21 format governs the means by which that content is recorded and communicated in machine-readable form. The original MARC format was developed at the Library of Congress by Henriette Avram in the early 1960s, and has served as the metadata container for library materials for over forty years. It is important to be aware that MARC is only a data container and that there are other “containers” that are coming into use, for example, Dublin Core and MODS (Metadata Object Description Schema), especially for digital materials. Following is a discussion of some of the MARC tags commonly used in serials cataloging.
Field 130—Uniform title main entry: In AACR2 serials cataloging, uniform titles are used to distinguish serials with the same title from one another. This type of uniform title is known as a differentiating uniform title, as opposed to the collocating uniform title, which is designed to gather like works together, such as those uniform titles used for law, music, or religious works. Serial uniform titles consist of the title proper and a qualifier of some kind, usually place of publication, corporate body name, or format of the material. Many serial uniform titles pose potential problems for users of the catalog.
If a serial title is qualified by the place of publication and that place changes, the uniform title is not updated to reflect the fact. While a note elsewhere in the catalog record updates the publication information, in a results list in the OPAC, the uniform title no longer accurately reflects the current serial title. If a serial is qualified by a corporate body name, and that name undergoes a change requiring a new authority record, according to AACR2, a new catalog record must also be created, with a new 130 uniform title qualified by the new corporate body name. Problems with serial uniform titles do not reflect limitations of the MARC format as much as they do AACR2 and the LCRIs.
Field 246—Variant title(s): Variant titles, including titles not chosen as the title proper, are recorded in repeatable 246 fields. Minor title changes that appear on later issues are also recorded in this field. Because users of the catalog may be searching for a serial title under a later (variant) title, if the OPAC does not clearly label and display the 246 field(s), the desired title may go unrecognized, as in the case of the ONS newsletter discussed earlier.
Field 022—International Standard Serial Number (ISSN): Subfield $a contains the ISSN for the serial publication being cataloged. Of almost equal importance is the subfield $y, which contains the incorrect ISSN. The reason this field is important is because often when a serial changes title the publisher will continue to print the ISSN for the former title on current issues. Acquisitions check-in staff routinely search by ISSN, so it is essential that any numbers associated with a given serial title (whether correct or not) be searchable in the catalog record.
A recently developed new category of ISSN, called the linking ISSN (ISSN-L), hoped to be implemented in 2008, will provide for collocation among the different versions of the same continuing resource. The linking ISSN will be the first ISSN assigned to any version; newly defined subfields in the 022 ($l and $m) will contain the ISSN-L and any cancelled linking ISSNs, respectively.
Field 362—Numbering: This is a field unique to continuing resources and contains the designation and/or numbering system found on the first and/or last issue. In OPAC displays this field has sometimes been confused for the library's holdings.
7XX linking fields: Linking fields are very important in serials cataloging to express the many relationships titles often have to one another. The most common relationship is that of earlier and later titles in a succession of title changes, recorded in the 780 and 785 fields, respectively. Other relationships include other physical formats (776) and supplements (772). Most online catalogs can use the 7XX linking fields to retrieve related titles in an OPAC search, although few systems have found elegant ways to display them together in ways that clearly express their relationships.
Field 856—Electronic location and access: This field contains the URL of the electronic serial, the specific issues and/or dates of coverage, and any notes on use.
OPAC RETRIEVAL AND DISPLAY ISSUES
There are many concerns about the usefulness of serial records in current online catalogs. Some of these concerns have already been touched on, for example the ambiguity of serial uniform title qualifiers and the coexisting practices of creating single and separate records for print and online versions. Some of the difficulties with serial records, however, have less to do with the cataloging records and more to do with the way online systems perform. Some searching commands limit the MARC fields they will index, and many OPACs default to a brief display that may not show all the necessary bibliographic information to identify a serial title. Display labels in the OPAC also often contain cataloging terminology misunderstood by users.
Keyword searches that depend on relevancy ranking can yield widely varying results, as demonstrated by a comparison of searches performed in both the Library of Congress OPAC and Google. It is not apparent what ranking the different systems use and how much weight is placed on which elements in the bibliographic record. It is clear that more work needs to be done in our online catalogs to make better use of the rich metadata we are creating for serials.
THE COOPERATIVE ENVIRONMENT OF SERIAL RECORD CREATION AND MAINTENANCE
The only thing constant about serials is the fact that they change. As more issues of a serial title are received or accessed online, information will be added to an existing serial record to describe, for example, variant titles, changes in publisher, new issuing bodies, variations in the numbering system, and sometimes, the cessation of a title. The serials cataloging community, working in a shared database such as OCLC and under a commonly agreed-on set of standards, together contribute toward the accuracy and authoritativeness of the metadata.
Other information sources also contribute to the control of serial information, including the ISSN Portal, which is the official, authoritative database of ISSN records. The importance of the ISSN reaches far beyond the mere recording of a standard number in a bibliographic record; the ISSN is one of the linking mechanisms in online catalogs for related serial titles, it is often the key element in OpenURLs used for electronic resources, and it is embedded in digital object identifiers (DOIs) to connect to content in serial titles. ONIX, originally developed as a standardized format used by publishers to send metadata about books to vendors, is also used for information about price lists, subscriptions, and general product information for serials. Metadata beyond the confines of AACR2 and MARC 21 can enhance our databases as we look toward the future of what we currently know as the library catalog.
THE FUTURE OF SERIALS AND SERIAL CATALOGING
What does our future hold? We know that electronic resources will continue to proliferate, particularly for serials. In the online context, we are faced with a new set of collection development decisions: Do we continue to keep the print, in how many packages does a given title appear, and how do we prevent duplication? Can we print, download, or loan the online version? What are the archiving possibilities, and what does our particular license allow in terms of perpetual access to the material? Libraries will continue to pursue consortial buying agreements, and also begin to consider shared archiving responsibilities as well. Who will be the custodians of the non-tangible resources we no longer own but simply access from vendors?
We also know that library catalogs will increasingly rely on the acceptance of metadata from non-traditional cataloging sources such as publishers and vendors, further upstream in the acquisitions process. There will be more batch-processing of records and automated quality control. We must be open to using metadata creation standards beyond AACR2 and communications formats beyond MARC 21. Other metadata schema such as MODS and Dublin Core will become integrated into our systems. The American Library Association, the Library of Congress, and national libraries from the English-speaking cataloging world have attempted to address the future of content designation with the development of Resource Description and Access (RDA), the successor to AACR2.
STANDARDS DEVELOPMENT FOR NEXT GENERATION CATALOGS
RDA will be our new standard for description and access of materials, with the expectation that catalog records will be used in the digital environment. Although designed primarily for use in libraries, it promises to be flexible enough to adapt to many information communities' needs, and to be compatible with internationally established principles and standards. One of the underlying principles of RDA is the application of the Functional Requirements for Bibliographic Records (FRBR), developed by the International Federation of Library Associations (IFLA).
FRBR is a conceptual model for looking at bibliographic information that focuses on what users want or need, and how databases can successfully fulfill those requests. Primary user tasks identified by FRBR are: find, identify, select and obtain; ideally, we should define a set of mandatory elements in every bibliographic record that allows for these tasks to be accomplished. FRBR also breaks down a bibliographic entity, or “work” into four elements: (1) the work itself, which is the pure abstract concept (e.g., Shakespeare's Hamlet); (2) the expression of that work (a translation in French); (3) the manifestation, which is the embodiment of the expression, or the actual publication (the printed text of the French translation); and (4) the item, which is one example of a manifestation (the copy of the text you have on your library shelves). Databases should be retrieving and arranging together the four different types of entities in ways that highlight their relationships to one another.
It is actually very difficult to apply FRBR entity concepts to serials, because the definition of a serial “work” is open to much debate. Is each title change a new work, or is the history of a title, including the earlier and later publications, one entire work? Nevertheless, the two FRBR goals of meeting user task needs and identifying groups of entities that should be related and displayed together in the OPAC will certainly inform future metadata record requirements as well as the design of next generation catalog displays.
QUESTIONS AND ANSWERS
Q: What do those <>'s in serials cataloging records mean?
A: They are for recording uncertain data in your catalog record. For example, you have an issue for 2005 in hand and the place of publication is different than what is recorded in the 260 field of the existing catalog record. Also according to the existing record, the latest issue consulted was 2002. Your note should state: “Issues for <2005> published: Chicago, IL,” which means: “Issues for at least since 2005 …”
Q: Are all online serials qualified by “online”?
A: No. The uniform title is applied only to subsequent titles which conflict with other existing titles in the database. There are cases in which the online version of an otherwise uniquely titled serial is cataloged before the print version, and therefore needs no uniform title qualifier. But when the print version is later cataloged, it must be given a uniform title with the qualifier “(Print).” Confusing? Certainly.
Q: What is the $3 in the MARC field 856?
A: This subfield is for the “materials specified,” i.e., the part of the materials described in the catalog record to which the 856 field applies. For serials this usually means the issues/dates of coverage for the online access, e.g., “v. 66-” or “Current issues only.”
Q: Why do I sometimes see “online” and sometimes see “electronic resource” in the online catalog?
A: The “online” label is usually associated with the 130 uniform title qualifier for the electronic version of a serial; the “electronic resource” label is the medium (general material designation) found in the subfield $h of the 245 title field. Depending on how your OPAC searches and displays title lists and single-title records, these two terms may both appear (somewhere!) when associated with electronic serial titles.
Q: What is an e-ISSN?
A: Technically, there is no such thing as an “e-ISSN”; while there can be a unique ISSN assigned to the electronic version of a serial, it is still just a regular ISSN. Publishers have created the label “e-ISSN” to identify the ISSN that has specifically been assigned to the electronic version.
Q: Is there anything that cannot be put in a 9XX field?
A: No—this is a locally defined field. Go wild!
CONCLUSION
The subtitle of the preconference workshop was: “101 Things Non-Catalogers Should Know …”, and certainly our presenters Regina Reynolds and Marla Whitney Chesler delivered at least that many “things” in terms of trends, issues, and concepts. The half-day workshop provided useful history and background as well as updates on current developments in serials cataloging and control. A good portion of time was devoted to questions and discussion, and we certainly have much to ponder as we look to the future of serials acquisitions, control, and access.
CONTRIBUTOR NOTES
Regina Reynolds is the Head of the National Serials Data Program, the U.S. ISSN Center of the Library of Congress.
Marla Whitney Chesler is FEDLINK Network Librarian at the Library of Congress-FLICC/FEDLINK.
Melissa Beck is the Senior Cataloging Librarian at the UCLA Law Library.