Subject Analysis Committee Subcommittee on Authority Records for GSAFD

Final Report: ALA Midwinter Conference 2001

January 15, 2001

I. Background

The Subcommittee was created in April 2000, with the following charge:

“Investigate issues related to the creation, distribution and maintenance of a machine-readable file of authority records for headings in the GSAFD. Explore and define the potential need for an ongoing subcommittee to oversee maintenance of such a file and the potential relationship to future revisions of the print publication. An interim report should be made at Annual 2000 and a final report by Midwinter 2001.”

The Subcommittee was created following discussions in Subject Analysis Committee (SAC) meetings during the Midwinter 2000 conference in San Antonio. It was observed that, with the publication of the second print edition of the Guidelines for Subject Access to Individual Works of Fiction, Drama, Etc. (GSAFD), there might be demand for a machine-readable version of authority records derived from the section on genre access. Initial discussions raised issues in several areas, from the work involved in creating such a file, to the logistics of distribution and potential intellectual property concerns. It was also unclear, in these discussions, as to whether or not SAC should be involved in any process of further revision, either of the (potential) electronic file or the print publication.

The Subcommittee agreed to divide the issues to be considered into two parts. Prior to Annual 2000, the Subcommittee concerned itself with issues related to the creation and distribution of an electronic file, including assessing the potential demand for such a file. Issues relating to the future of GSAFD, whether in electronic or paper form, are the focus of this final report.

The recommendations included in the Subcommittee's interim report, as approved by the ALCTS Board of Directors, were as follows:

"ALA Editions should make the file available via ftp, for a one-time charge to individual customers not to exceed the cost of the print GSAFD. ALA Editions should negotiate an appropriate price with interested vendors (including the incorporation of GSAFD in products such as Cataloger’s Desktop). In addition, ALA Editions should investigate making the genre thesaurus available as a Web page. The ftp file should be made available for distribution as soon as the additional testing described above has been completed. The availability of the file should be publicized via discussion lists (including PubLib as well as cataloging-oriented lists), LC Cataloging Newsline, the ALCTS Newsletter, AN2, American Libraries, and Public Library Association publications."

Discussion with ALA Editions regarding these recommendations is still in progress as of the date of this report.

II. The Future of GSAFD

The first edition of the Guidelines for Subject Access to Individual Works of Fiction, Drama, Etc. was published by ALA in 1990. The much-awaited second edition was published in 2000. GSAFD was developed to meet the long-standing need for an authorized standard for subject and genre access to separately published literary works. Subject access is delineated to include topics, settings and characters. The most extensive section of GSAFD is a list of genre terms, including reference terms and scope notes, the latter particularly featured in the second edition.

It is reasonable to claim that the development of GSAFD, particularly as implemented through the OCLC/LC Fiction Project, has been the factor most responsible for mainstreaming the practice of subject and genre access to separately published literary works. While there are disagreements on matters of application – for example, on the issues of subject heading complexity or the use of subdivisions – the basic practice is now central, not peripheral, to the cataloging of literary works. Public, academic and school librarians who care about the catalog as a tool to assist patrons with “what to read next” owe an enormous debt of gratitude to our colleagues who worked for years on the two editions of GSAFD.

The present Subcommittee must, however, consider the future of GSAFD not only in terms of the circumstances that gave it birth and its present uses, but also with regard to changing contexts in the cataloging world at large. We must also consider the limited time and energy that any voluntary group of people has to devote to advancing cataloging practice. That is, our recommendations must take into account not only the most useful future for GSAFD as an entity and the impacts of potential changes on the practice of working librarians, but also the best uses of volunteer time given to ALCTS by our colleagues.

The main question before the Subcommittee can be put simply, as: “Should there be any further work done on GSAFD as a work in itself?” If so, should we aim for a standalone printed publication as the first two editions have been, or would we recommend maintaining a regularly-revised database of “GSAFD headings” to supplement the Guidelines proper? If not, the alternatives seem to be either the merger or alignment of GSAFD with some other standard, or no further work done on GSAFD at all.

The Subcommittee’s recommendation is that, following Midwinter 2001, a subcommittee be created to investigate an alignment, resulting in a merger, of GSAFD with the Library of Congress Subject Headings. This alignment involves both the application guidelines and the genre headings themselves. This recommendation is in response to two significant recent changes in the world of subject cataloging, neither of which was in place when GSAFD was first developed. The first is the Library of Congress’ recent decision to resume assigning subject and genre terms to individual works of fiction. LC’s policy statement of Summer 2000, as published in Cataloging Service Bulletin no. 89, is attached as an appendix to this report. Although LC’s own practice will be limited to current American and other English-language imprints, it seems less urgent to maintain a separate standard authorizing this practice. Indeed, it seems plausible that discussions can be pursued which will allow subject and genre access to individual works to be authorized under LC’s umbrella, regardless of the extent of the Library’s internal policies. There may be a valid parallel here with certain Library of Congress Rule Interpretations, which state that the Library will accept data elements or values in derived records, although it does not include those elements in its own cataloging.

The second major change is the existence of SACO, which has allowed librarians around the globe a mainstreamed means of influencing the development of LCSH. It is plausible that the proposed subcommittee could oversee a “gsafd funnel project,” devoted to harmonizing differences in terms and usages between GSAFD and LCSH. Pursuing such a project would mean, among other things, that the intellectual work invested in the editions of GSAFD would not go to waste, but would be carried into a broader arena via SACO. In a sense, this undertaking is already underway. Over 60% of GSAFD headings are established in LCSH. A group of LC headings such as Fantastic fiction was changed to Fantasy fiction, via a SACO proposal from the British Library in anticipation of GSAFD’s second edition. Coincidentally, as the Subcommittee was coming to its recommendation, Joanna Fountain’s comparison of LCSH with GSAFD, Sears, and AC headings was posted on the "gsafd" discussion list. That comparison is attached as an appendix, with permission. It should be noted that this comparison involves headings only, not reference terms or scope notes. A “gsafd funnel” would obviously need to oversee harmonization of all authority data elements.

Our recommendation does not imply that GSAFD itself will, or should, disappear or be deprecated in the short to medium term. A merger with LCSH, if carefully carried out to address several concerns noted below, would not likely be accomplished within a couple of years, and the publication itself could continue to stand as an authorized source of terminology for an indefinite period. There is a parallel with Moving-Image Materials: Genre Terms, which is still authorized and used as a source of form and genre headings for films and videos, although it was published in 1988 and never went into a second edition. What we envision in this case is that LCSH, both in terminology and application, would eventually emerge as the natural and preferred source for subject and genre analysis of individual literary works.

The Subcommittee came to its initial decision by early November 2000. On November 27, we posted a short initial description of our recommendation to the “gsafd” discussion list for comment. We chose this list, with a subscriber base of about 280, as the one with a ready-made constituency of interested parties, who would most likely have the best informed collective opinion as to GSAFD’s future. Following the initial posting, fourteen comments were received from nine individuals. This included five replies from three committee members, and two comments sent privately. These few responses brought some issues to light that have been incorporated in this report. It should be noted that most of the discussion centered on the list of genre terms, rather than the Guidelines in its entirety as a work.

III. Benefits and Drawbacks

Benefits. The primary benefit of a merger with LCSH would be the creation of a single source of genre and form headings for the most popular types of literature. As things stand, there are two competing sources of similar terminology for these materials. If GSAFD, as an alternative source, concerned itself with specialized materials not intended to be addressed by LCSH, it would potentially make sense for ALCTS to invest its resources in a “niche” thesaurus. But since the pertinent LCSH can easily be assigned as genre/form headings, and coded as such in MARC21 (a point that is often not understood), it should be clear that the two sources address the same materials and similar if not identical constituencies.

The conundrums presented by the use of multiple thesauri to catalogers, authority control librarians, reference librarians and the public have been regularly addressed in the literature and conference discussions. While few would argue that there could possibly be only one source of controlled vocabulary, it seems more than appropriate to remove sources of needless complexity where these can be identified. The “choice” between using LCSH and GSAFD for genre/form terms may not be worth the cost of multiple local policies and added levels of verification, considering that the vocabularies and audiences are so similar.

Stephen Hearn, in a message to the "gsafd" discussion list sent prior to the posting of our recommendation, warned against some of the consequences of using multiple thesauri as convenient sources of headings to "mix and match." His message read, in part:

LCSH and GSAFD are self-referential systems, not just collections of "authorized" terms in some abstract sense. Their headings' authority, meaning, and relationships to other terms depend on the system they reside in. If the boundaries of the original system are not regarded, their authorization, meaning, and relationships are all cast in doubt. Within a given catalog, one can cobble together self-consistent hybrids employing terms authorized by multiple systems (it takes work, but one can do it); but the inevitable redefinitions and realignments which this [local] merging process entails will make it that much harder to integrate one's catalog with other catalogs for outside users doing broadcast searching. It will also leave one's local users less able to carry a search successfully beyond the confines of their own catalog into a larger searching arena. (November 16, 2000)

The continuing existence of two highly similar thesauri, addressing the same types of materials and similar user bases, encourages the process of local mixing and matching, without evident benefits for national and international record sharing.

A second benefit of the proposed merger would be that the development of genre/form vocabulary would proceed under an established institutional umbrella, rather than being primarily dependent on the efforts of volunteer committee members. The SACO process allows for broad collegial involvement, combined with LC’s institutional stability.

A third benefit would be that the energies of SAC members and colleagues, rather than being devoted to continual maintenance of a controlled vocabulary, would be available to address other, more pressing, issues.

Drawbacks. The primary drawback of a merger, judging by the online discussion, is that LCSH is, as a tool, more difficult to consult than GSAFD. One of the great virtues of the latter is that it provides a simple and quick way for librarians and patrons to find appropriate headings, as compared with sorting through the five-volume LCSH. Concern was also expressed that some GSAFD headings, which differ from LCSH and are preferred by many, may be lost through a merger. Similarly, it is possible that differing, preferred uses for identical terms may also be lost.

IV. Next Steps

A. With this report, the present Subcommittee has fulfilled its charge. We request that SAC dissolve this Subcommittee and form another to carry out our recommendations. The charge for the next Subcommittee would center on investigating a merger of GSAFD with LCSH, in a manner that carries forward the work already done by SAC and addresses the concerns that have been expressed. The specific issues that seem to us most critical to address are given in the following outline, although the next Subcommittee may discover others.

B. Issues to examine:

1. Examine LC’s developing guidelines regarding application of subject and genre/form headings to individual works, compare them with GSAFD and engage in discussion to reconcile differences.

2. Examine differences between LCSH and GSAFD regarding established headings and other authority data, including reference terms and scope notes.

a. Consider creating a "gsafd funnel," analogous to existing funnel projects (such as the African American funnel), to address the differences.

3. Work with LC to develop a process to extract a concise list of the most common genre/form headings, similar to GSAFD in simplicity. This tool should be easy for catalogers, public service librarians, and patrons to consult, and should be easily and regularly updated.

4. Examine and promulgate guidelines for establishing or revising genre/form headings. The issues involved include clarifying literary warrant as it relates to genre/form, and revisiting pseudo-genre headings which might be better established as [Topic]—[Form].

5. Resolve GSAFD's incomplete syndetic structure via the LCSH merger.

C. Additional considerations

1. One respondent to the online discussion, Ella-Fay Zalezsak, suggested that it may be useful to create a discussion list where librarians who have been locally juggling multiple thesauri can share their experiences. It may be that the present gsafd list will serve that purpose, but the list would need to be re-publicized if its focus changes in this way.

2. The next Subcommittee may want to consider hosting educational or discussion forums to address issues of application. There apparently remains confusion among librarians between sources of controlled vocabulary, the use of those sources for subject or genre/form headings, and appropriate MARC21 coding of headings from those sources. One respondent to the online discussion described a situation where paraprofessionals may assign genre headings in 655 fields, but not subject headings in 600-651 fields. She is concerned that if GSAFD disappears, the paraprofessionals will no longer be able to assign genre terms (since LCSH is considered to “equal” 600-651). Another potential forum would be a discussion of standards for establishing genre headings, and distinguishing genres from topics expressed via literary forms. This might productively include a presentation from experts in the field on how genres emerge in literature (and other media) and how they change over time.

Respectfully submitted,

David Miller (Curry College), chair

Ruth Bogan (Warren-Newport Public Library)

Shannon Hoffman (Brigham Young University)

Ingrid Mifflin (Washington State University)

Dale Swensen (Brigham Young University)

Mary Dabney Wilson (Texas A&M University)

Tony Olson (Northwestern University) (ex officio)

Appendix I. SUBJECT HEADINGS FOR INDIVIDUAL WORKS OF FICTION

The Library of Congress has participated in several pilot projects and co-operative projects that provide enhanced subject access to individual works of fiction. The Library will now be assigning additional subject headings to selected works of individual fiction as part of its normal cataloging practice. Complete instructions on this cataloging practice will appear in a future update to the Subject Cataloging Manual: Subject Headings. The additional subject headings will appear on bibliographic records for current acquisitions of American literature and other English-language literatures. Genre headings will also be assigned from the Guidelines on Subject Access to Individual Works of Fiction, Drama, Etc., or from Library of Congress Subject Headings.

– Cataloging Service Bulletin, no. 89 (Summer 2000), p. 25

Appendix II. Selected Genre Terms (MARC field 655), compiled by Joanna F. Fountain (attached).