PCC Resources for
Participants in the discussion: Everett Allgood, Jean Altschuler, Jeanne Baker, Pam Dunlop, Crystal Graham, Mary Grenci, Karen Hsu, Ed Jones, Jennifer O'Connell, Regina Reynolds, Ann Sitkin.
Consultants: Tad Downing, Guenter Franzmeier, Michael Gorman, Wayne Jones, Sally Strutt, Hugh Taylor
AACR Group 1 was charged with reviewing definitions and entry conventions. The group was initially charged with redefining 'serial' according to Model B in the Hirons/Graham paper. However, the problems of doing this led to a proposal for a "Modified Model C" in April. This revised model was used as the basis for definitions, and in the discussion of entry change conventions. The process has been an evolutionary one.
DEFINITIONS
RECOMMENDATION 1.
Expand AACR2 type of publication 'serial' to 'serial and serial-like entities.'
Rationale. AACR2 divides chapters by either the 'class of materials' (e.g., sound recording, map, etc.) or the 'type of publication' (AACR2 1.0A1.) There are currently two 'types of publication':
- Monograph: a nonserial item (i.e., an item either complete in one part, or intended to be completed, in a finite number of separate parts).
- Serial: a publication in any medium issued in successive parts bearing numeric or chronological designations and intended to be continued indefinitely.
Participants at the Toronto conference agreed that the current definitions do not adequately cover all types of bibliographic entities. Loose-leafs, electronic databases, web sites, and other publications of an ongoing nature that don't fit the definition of serial are among those not included.
This recommendation would add to AACR2 a category of materials never included, i.e, those that are not issued in discrete parts and are ongoing by nature. In order to explain what is meant by this category of "serials and serial-like entities," we recommend that introductory material be added to the code which would cover the category of type of publication and how the catalogers makes the determination as to which type is appropriate. It would also cover the application of the concept "serial-like" to a variety of ongoing publications. The placement of such explanation will depend on the ultimate structure of the code. If the code is reordered according to area of the record, the introduction would be at the beginning of the entire code. If not, the information could be added as an introduction to an expanded chapter 12.
Discussion. Much of the time between January and March was devoted to the redefinition of 'serial' and a working definition was achieved some time in March. In the process, it became clear that using a single term to cover the broad spectrum of bibliographic entities that are not finite would not only require needless distinctions for newer types of materials, but would also make the distinctions between traditional serials and monographs more difficult to determine. After a number of people voiced their desire to move more rapidly towards Model C from the Hirons/Graham paper, Hirons and Reynolds issued a proposal in April to adopt a "modified Model C" which removed some of the vagueness of the original model and called for a pragmatic approach to USMARC implementation in systems. The revised model used the term 'ongoing,' first used in the Hirons/Graham paper, as an umbrella to tie together publications for which the parts/updates are successively issued and those for which the parts/updates are "integrated"into the whole. It was aired widely at various venues and discussed with the JSC following the ALA annual meeting in early July 1998. The JSC gave Hirons the approval to pursue this new approach but suggested working within the terminology of the code and taking into consideration the "Logical Structure of the Anglo-American Cataloguing Rules," prepared by Tom Delsey (National Library of Canada).
As a resource person in the ISBD(S) Working Group, Hirons and Ed Swanson contributed a paper to that group recommending that ISBD(S) be revised to cover the scope of 'ongoing' entities. In November the group agreed to the expansion but indicated a reluctance to use the term 'ongoing' because of the difficulty of translating it into other languages. Reynolds reported that there had also been concern with the term at the ISSN Directors meeting in September. In response to the problems with the term 'ongoing,' Reynolds suggested the wording 'serials and serial- like" to inclusively represent both traditional serials and publications that are not finite but do not fit the definition of a serial. Most found this an appropriate alternative and thus, this terminology is used in this document.
Group1 also discussed what might constitute an appropriate noun. The Hirons/Graham paper used the term "publication" but some at the conference felt this was too print-centric. As alternatives, the group discussed the use of "resource," which is used in ISBD(ER) and "entity" and finally settled on the latter. This is not a serial issue alone and the ultimate decision on such terminology will have to be determined by the JSC and editors of the code.
It should be noted that this approach is not a perfect solution. The original Model C in the Hirons/Graham paper made its primary distinction between publications that were Static (i.e., complete), from those that were Ongoing (i.e., not complete) with the understanding that those that were not complete could change over time and rules would be needed to accommodate such change. This was considered too drastic a change to introduce into our current organization of cataloging, however, and thus, the aspect of finiteness vs. intention to continue has been retained in the model. As a result, there are publications categorized as monographs that can also exhibit change over time, such as multi-part monographs and electronic resources that will be revised within a fixed time. Thus, the need to handle change is not limited to 'serials and serial-like entities' and rules for such change need to be included with monographic publications, or instructions could be included to consult the rules for serials and serial-like entities when change occurs.
A further problem with retaining the distinction between intention to continue and finite is that for some publications, such as electronic publications, it may be unclear what the intention is. It is hoped that a pragmatic approach can be taken to quickly determine whether a resource has been created with the intention to be added to/updated continually or whether it is essentially finite in nature.
The resulting model for type of publication would be:
Bibliographic entity _____________________________________ | | Monographs Serials and serial-like entities _______________________________ | | Successively-issued Integrating | | Serials (incl. Loose-leafs series) Databases Web sites
Impact.
AACR2: The expansion in AACR2 will create the need for new rules and the need to reexamine existing ones, as documented in other proposals. The term "monograph" could be retained.
Note: The proposals in this paper deal only with bibliographic entities that are intended to be ongoing. It is possible that monographic entities can also be successively-issued and integrating, such as multi-volume monographs and electronic resources that are clearly finite.
ISBD(S): The ISBD(S) Working Group has approved the proposal to broaden the scope of ISBD(S) to cover integrating as well as successively-issued publications.
ISSN: The ISSN Directors will have to determine which serial-like entities will be in scope for ISSN assignment. Unnumbered series are already assigned ISSN. At their fall 1998 meeting the Directors agreed to experiment with assigning ISSN to databases. It is possible that not all serial-like entities would have to be treated in the same manner. Thus, the ISSN network might decide to assign ISSN to some serial-like entities but not to others.
MARC: The introduction of 'serial-like' seems to offer more flexibility with coding. Some serial-like entities might be coded as serials (e.g., databases, electronic journals), while others could continue to be coded as monographs (e.g., loose- leafs). A MARBI discussion paper will be written in the spring discussing possible options.
RECOMMENDATION 2.
Add to or revise the following definitions in the glossary. The definitions are given below in hierarchical order, rather than alphabetical, to clarify their their relationship to other terms.
Bibliographic entity: A manifestation of a work that forms the basis for bibliographic description. A bibliographic entity may be in any medium or combination of media or may not be a physical entity.
[This definition is the basis for many that follow.]
Monograph. A bibliographic entity that is complete or is intended to be completed within a finite number of parts or within a fixed period of time.
[This the current definition with the addition of the last clause which enhances the sense of a publication being finite, regardless of how it is issued.]
Serials and serial like entities. Bibliographic entities that have no predetermined conclusion.
Successive issuance. A publication pattern that describes a bibliographic entity that is issued as a succession of discrete parts, each of which bears a chief source of title. Examples of successively-issued entities include printed or other tangible serials, some electronic journals, multi-volume monographs, and monographic series.
Integrating issuance. A publication pattern that describes bibliographic entities that are added to or changed by means of updates that do not remain discrete and are integrated into the whole. The updates may be separately issued (e.g., loose-leafs), or integrated by the publisher (e.g., an electronic database or Web site). Bibliographic entities that are integrating have a single chief source of title that may be replaced or changed over time.
Serial. A bibliographic entity that is issued in separate parts usually bearing numeric or chronological designations and that has no predetermined conclusion. Serials include journals, magazines, annuals, some conference proceedings, newspapers, and monographic series).
[Rationale: The definition is now more consistent with that used for ISSN. "Usually bearing numeric..." includes both unnumbered series and publications that lack a designation on the first issue but are known to be a serial.]
Series. 1. A group of separate bibliographic entities related to one another by the fact that each item bears a collective title applying to the group as a whole. The individual entities may or may not be numbered. [2. and 3. in the current definition will be retained.]
[Note: This definition will be further refined by Judy Kuhagen.]
Monographic series. [Definition to be supplied later by Judy Kuhagen.]
Serial-like entity. A bibliographic entity that is not finite but that lacks one or more of the characteristics required to meet the definition of a serial. A serial-like entity is usually issued in an integrating form (e.g., loose-leafs, databases, etc.).
Database. An integrating serial-like entity, in the form of a searchable computer file, consisting of a set of interrelated files created and managed by a database management system. A database may be issued as a direct-access computer file on CD- ROM, diskette, etc. or as a remotely-accessed computer file accessed via dial-up methods or on the Internet.
Electronic journal. A serial or serial-like entity broadly defined as one which usually consists of articles and is delivered via a computer network. Electronic journals include journals, magazines, annuals, some conference publications, and newspapers. They may be issued in successive or integrating forms.
[It is important that it be clear that electronic journals can be either successive or integrating. The term is also rather broadly used and is not meant to be the electronic equivalent of what we might call a 'journal' in print.]
Loose-leaf publication. An integrating entity that consists of unbound pages intended to receive updates. The format permits the addition, substitution, or removal of pages, and allows for frequent addition or revision to materials that need to be kept current.
[This definition excludes successively issued parts that are issued with holes for filing in a binder but remain intact; these are treated as normal serials.]
Web site. A serial-like collection of dynamic documents, commonly referred to as "home pages," that are usually interconnected and located on the same server. Web pages are issued in an integrating format.
[This is based on the definition from ISBD(ER). It needs to be developed with others working more closely with Web sites. However, in cataloging electronic serials, our experience is that Web sites are closely related. There should definitely be a definition in AACR2.]
Impact
AACR2. The Glossary would be revised and rules would be revised to reflect this terminology.
ISBD(S) and ISSN. In order to harmonize practices, it will be important for AACR, ISBD(S) and the ISSN network to use the same terminology and definitions. Both ISBD(S) and ISSN may have to reconsider some definitions.
ENTRY CHANGE CONVENTIONS
RECOMMENDATION 3.
For serials and successively-issued serial- like entities, use multiple records according to successive entry conventions when a major change occurs in the title proper, uniform title, or main entry heading.
Rationale. The separate parts of bibliographic entities issued successively each have and retain their own title and thus, a record for each major change is warranted. This is the convention that has been followed since the adoption of AACR and changing the practice would result in massive recataloging, reshelving, and disruption to catalogs with little merit. Retaining successive entry for tangible serials facilitates shelving by title, keeps records shorter and cleaner, is better for shared databases, and is in keeping with ISSN and other international standards. Successive entry would also be used for electronic journals that are issued in a successive manner in which earlier titles are retained.
Discussion. Group 1 briefly discussed a proposal by Lucia Rather (via Hirons) that all serials be cataloged according to latest entry by using the latest successive record and changing it into a latest entry record. While this has appeal for those wishing to have a single record and keep records in synch with those for integrating counterparts, members of the group rejected it as too disruptive to current catalogs. Previous experience with latest entry, particularly the complexity of the records over time, left most members strongly in favor of continuing successive entry. One member, however, voiced strong approval for latest entry for all serials, but acknowledged the problems with existing records.
Impact: No real impact, other than maintaining two types of conventions if recommendation 4 is adopted.
RECOMMENDATION 4.
For integrating entities, use a single record, according to latest entry conventions, to account for major changes in the title proper, uniform title or main entry heading.
Rationale. A bibliographic entity that at any one time has only one chief source of title is best expressed on a single record. If the title (or other entry element) changes significantly, a title history note and added entries are sufficient to provide access to the earlier titles. Latest entry conventions are currently being used (though not called this) for loose-leaf publications, and by those cataloging databases and Web sites as monographs. By applying latest entry to electronic journals with a single title source as well, one form of entry change convention would be used for all integrating entities. This will make it unnecessary to distinguish different types of integrating entities from one another; the only distinction necessary will be to determine whether an entity has a single chief source of title or a succession of such sources. This is the simplest and most straightforward way to describe these publications. But it is not without its problems.
Discussion.
- Problems associated with latest entry. This has been one of the most difficult issues to deal with and has produced some creative thinking but no perfect solutions. The problems with using latest entry for electronic journals, databases and other types of integrating entities, particularly those with print counterparts, are:
- Links. Links to a latest entry record from a successive counterpart would be difficult to maintain if created under current rules. Not only is there a one to many relationship, but the change of title in the latest entry record would require changes to that title in the links on all the successive records.
One possible solution to this would be to omit the title and replace it with a statement of relationship, such as "online version," similar to the current practice for microforms. Since the record number would not change, the link would be stable and would not need maintenance. While this would work well with physical format relationships, it would be more problematic for other types of relationships. This solution has potential problems for some systems and is probably not something that should be specified in the code, but left to more local solutions.
- ISSN. The ISSN is assigned to each new key title and the centers must report separate records each time a key title changes. Multiple titles and ISSN can be handled in the MARC record by including the ISSN in $x of the 247 field; thus, this is not a problem for cataloging per se. We are very conscious, however, of the importance of the ISSN and the need to stay in sync. But the needs of ISSN cannot direct the cataloging rules. It also seems possible that there will be a desire for a stable ISSN for a Web address, regardless of title changes, if the ISSN is to be used in such standards as the URN and DOI.
- Record maintenance. Latest entry records would require more maintenance. At present, only CONSER catalogers can change most serial records on OCLC. We know that incorrect loose-leaf records have proliferated on OCLC and we also predict that separate records will be created for new titles rather than those titles being added to the latest entry record because the relationship is not always clear. However, this seems to be an issue for PCC and other cooperative programs, rather than one of rules. We need to improve awareness, get more people involved in maintenance, and also encourage publishers to provide clear information on their publications, where ever possible.
- Record clutter. Over time, latest entry records can become cluttered with data that on electronic publications may no longer exist. Whether or not this will be as much of a problem as it was on serials cataloged under earlier rules remains to be seen. We may at some point need to consider policies for removal of data no longer appearing on the publication.
- Electronic journals. While latest entry can be easily applied to databases, Web sites, and loose-leafs, its application to electronic journals is more complicated. To begin with, while the other types just mentioned are always integrating, an electronic journal can be issued either successively or integrating. Thus, we will probably have to decree that unless obviously otherwise, an electronic journal is assumed to be successively-issued when first cataloged. It may require subsequent issues or the first title change to really determine how the publication will be formatted. Additionally, the publisher may change from one format to another. Clearly, policies will be needed for such changes. A further complication is that the articles contained in an electronic journal are indexed by indexing and abstracting services under the title associated with them at the time of publication and may also be cited under this title.
Because of the problems associated with electronic journals, the group spent significant time exploring various alternatives to a single record.
- Links. Links to a latest entry record from a successive counterpart would be difficult to maintain if created under current rules. Not only is there a one to many relationship, but the change of title in the latest entry record would require changes to that title in the links on all the successive records.
- Proposals for alternatives. Two proposals were explored during the past year. They were both proposed specifically for electronic journals; their applicability to all integrating entities is questionable.
- Incorporating entry. This idea was proposed by Sara Shatford Layne in May. She suggested that since the articles within the journal may retain their earlier title, and even if they don't, would continue to be cited under that title, these earlier issues (or clusters of articles) form analytics. She suggested that a new record be created each time the title changed with analytical added entries for the earlier titles and a link between the records. She also suggested that the description for each new record be based only on the information present at the time of cataloging. A library could opt to retain only the latest record in its catalog, while the full set of records would exist on shared databases. The proposal was widely discussed and during the fall, Group1 looked at the specifics of its implementation. After a series of discussions, it seemed that the concept was difficult to communicate and understand and perhaps too complicated for widespread implementation.
- Succession of latest entry records. After examining the results of the Group 1 discussion, Hirons, Reynolds, and Judy Kuhagen decided that it would be easier to apply existing latest entry concepts rather than introduce a new one, but retain Layne's suggestion for a succession of records to deal with the problems of latest entry given above. Most of the group found this to be a better solution because it was based on known principles and voiced their approval for the proposal. When the proposal was shared with other groups, however, a number of respondents found it too had its complications and voiced strong concerns.
The recommendation in this report is based on the belief of the chair and others that latest entry, despite its drawbacks, is the best approach because it eliminates the need for multiple records and confusing links, keeps all integrating entities cataloged in the same way, and is already being applied by those who catalog many of the integrating publications. However, the problems of electronic journals warrant serious consideration and it is recommended that the alternative above for a succession of latest entry records be included in the final report to the JSC so that others can consider and provide their comments.
Impact:
AACR2. A rule or rules would be added to chapter 21 (or elsewhere in a reordered code) that could cover entry changes for integrating entities. It would also need to cover the situation of a serial or serial-like entity that changed its form.
ISBD(S). While ISBD(S) does not cover entry per se, it does state when to make a new description, and since title changes would be handled within the description, ISBD(S) would need to include such rules.
ISSN. The adoption of latest entry will have serious ramifications for ISSN assignment. Will ISSN be assigned to databases and other integrating entities? Will the ISSN network make changes to accommodate these types of publication or will they be excluded?
