Improving Representation and Access Through Ethical Description

MLA session Friday, March 3, 2023

Summary by Chuck Peters, Indiana University

Program slides are available via Humanities Commons, “Improving Representation and Access Through Ethical Description (MLA 2023 session)”

Speakers: Allison McClanahan, Indiana University; Kristi Bergland, University of Minnesota; Treshani Perera, University of Kentucky

Members of the MLA community discussed ways that they have approached ethical description at their institutions:

Archives of Traditional Music, Indiana University, Bloomington

Allison McClanahan (Collections and Cataloging Librarian, Archives of Traditional Music, Indiana University), described areas of focus in improving patron access and use of materials in the library catalog. The ATM, one of the oldest and largest university-based ethnographic and sound archives in the US, includes vocal and instrumental music, linguistic materials, folklore, interviews, and oral history.

Ethical description at the ATM includes adding bilingual and dual script description. When possible, the script from the language of the resource is included, rather than providing only a transliteration. The ATM accomplishes this with help from international students who are fluent in the required languages. Many of them are also graduate students in the Department of Folklore and Ethnomusicology. The descriptive cataloging of sound recordings, photographs, and many other materials all benefit from the efforts of so many native speakers of languages from around the world.

Ethical description of archival material at the ATM also includes the application of subject analysis from diverse subject authorities . In addition to LCSH and other controlled vocabularies commonly used in descriptive cataloging, the ATM uses the Human Relations Area Files Outline of World Cultures Codes (HRAF). HRAF is a set of codes, identifying primarily the continent, region, and culture of origin of the resource, and the codes are searchable in the local online catalog. HRAF does not replace LSCH; but is meant to provide another means of access to the materials. With HRAF codes added to MARC records, a library user doing research from a perspective other than music would come across ATM materials when searching the catalog using the HRAF codes. Catalogers at the ATM also employ the American Folklore Society Ethnographic Thesaurus (AFSET), allowing them more descriptive precision than Library of Congress vocabularies.

The ATM performs reparative description as well. Actions taken might be proactive or based on patron interactions, especially BIPOC patrons. Reparative description is a way to build and maintain trust between the ATM and its diverse user population, and users are always encouraged to give feedback on what they see in the catalog.

Reparative description might include removing the perspective of colonialism from the bibliographic record. These examples are from cataloger’s notes that were changed:

From: “Numerous males and females from West Africa and Angola”

To: “Numerous unidentified men and women from West Africa and Angola”

From: “Natives dancing …”

To: “Music and dance of peoples native to West Africa …”

Some legacy bibliographic records contain cataloger-supplied titles that include offensive demographic terms. Field collections, for example, follow a consistent naming convention that includes Country; Region; Cultural group; Date(s); and this is constructed and supplied by the librarian or archivist. Given the long history of the ATM, there are now bibliographic records in existence that contain outdated terminology. Established titles may have been cited and still need to be found; however, it was desirable to find a way to update the language, displaying the updated title while suppressing the (still searchable) former version.

The solution was to use MARC field 247, which was primarily used for former titles of continuing resources and could be coded to suppress the display. Using field 247 locally allowed the ATM to keep harmful language from the patron display, while still indexing as a title, with no impact on discoverability.

McClanahan emphasized that there is no one catch-all approach to using ethical description at the ATM. Practicing ethical description depends on the nature of the material along with the possibilities offered by technology and staffing. She added, “do what you can with what you can, but keep a list of projects that you would like to see eventually, and be mindful of who and how you can collaborate or work with towards that goal.”

University of Minnesota Libraries

Kristi Bergland (Music Metadata Librarian, University of Minnesota), provided details of actions taken by UMN Libraries in response to “Illegal aliens” and related subject headings included in descriptions of university library materials. As may have been the case with many libraries, developing and completing the workflow for bibliographic and authority file maintenance was a months-long process.

The discussion of what to do about the problem included possible consequences of deviating from LC practice. The problem of managing competing priorities—technical issues, messy data, lack of awareness—among library projects had to be factored in as well.

Bergland said it became apparent that it wasn’t only the subject heading that needed to change, but the conversation around it as well. Several events contributed to the decision to begin the project: First, there was the ongoing “Drop the I-Word Campaign”(Race Forward: The Center for Racial Justice Innovation.) There were objections from Dartmouth students that captured national attention, prompting statements from ALA and the Library of Congress. Congress responded to the LC decision, further complicating the issue. ALA’s Subject Analysis Committee (SAC) created a working group which issued a report in 2020, “Report of the SAC Working Group on Alternatives to LCSH “Illegal aliens.” Finally, the knowledge that other libraries also were addressing the need to change the subject headings added to the urgency. When local showings of the film Change the Subject took place, Bergland said one takeaway was that “these words were hurting people, and continuing to use this language makes people distrust our work.”

A values statement was developed by UMN Libraries: Diversity, Equity, and Inclusion in Metadata and Description: Statement of Values and Principles (2019-2020). In 2020 came the murder of George Floyd in Minneapolis, and the civil actions that occurred across the US as a result. It was realized that the lack of a well-prepared plan of action for libraries was not a valid substitute for inaction, as inaction was hurting people and maintaining barriers to access. Bergland said this includes access to resources, scholarship, and our profession.

By spring 2021 the decision was made to begin to address the headings locally. A local solution was implemented after a conversation between many constituencies, including the Discovery to Access Steering Committee; the Library Enterprise Systems Governance Group; the Cataloging and Metadata Group; and Law Library staff.

There were choices to be made before taking action: First, source records could be changed in Alma, or in the discovery layer (Primo), or both. The best choice of action seemed to be to change the discovery layer only, while waiting for LC to change the source records. Additionally, changes might be made to the display, to indexing, or both. Both seemed preferable. Finally, the replacement terms needed to be selected. UMN selected terms from the original 2016 plan from LC: “Noncitizens” and “Unauthorized immigration” would replace “Illegal aliens.” Subject headings containing the word “Alien” would now use the term “Noncitizen.”

More than 6,000 records contained one or more of the headings involved in the change. Although Alma includes an authority control process, it was not helpful because of the two-heading solution. The plan was for catalogers to evaluate records and make updates manually or in small batches. Then, in December 2021, OCLC applied the new headings to approximately 41,000 WorldCat records. Most of the 6,000 UMN records were in this group. While updated records from OCLC could be used to replace those in the local file, cataloging challenges remained. Many of these challenges had to do with free-floating subdivisions added to headings. Some subdivisions are allowed with all headings; others are applied to classes of persons and/or topics, but possibly not both. Subdivisions that may have been added to “Illegal aliens” previously now were unauthorized for use with “Noncitizens”; for example, $x Legal status, laws, etc. is authorized under classes of persons but not topics, and is not used with “Noncitizens.”

To assist with the decision-making process going forward, library staff is creating documentation for original catalogers to use when applying the new headings. It was also necessary to decide how heading updates would be managed in the future. Bergland pointed out how important it is to communicate these decisions to the broader UMN library community, because many people are not aware of how this kind of work happens.

The group continues to identify future needs, including other metadata terms that might need attention. It shares what has been learned through this process, particularly about specific content choices. And finally, it is necessary to understand how best to do the work and to make information on actions taken readily available, within UMN Libraries and to the broader community.

University of Kentucky Libraries

Treshani Perera (Head of Fine Arts Technical Services, Little Fine Arts Library, University of Kentucky Libraries), discussed local solutions to address the need for ethical description of library materials. Perera said the goal of the presentation was to point the viewer to related resources; and to share practical advice for creating local workflows to address problematic language and controlled vocabularies. The slides for this presentation (linked above) include links to resources that could be used in this work. In particular, Perera’s workshop “Critical Cataloging: Identifying and Dismantling Bias in Description” (ARSC Continuing Education Webinar Series, January 14, 2021) on identifying bias in descriptive cataloging remains available online, and she published about her research on inclusive description work and initiatives in the article, “Description Specialists and Inclusive Description Work and/or Initiatives—An Exploratory Study” (Cataloging & Classification Quarterly, 60:5, 355-386).

The presentation focused on a local project Perera initiated at the University of Kentucky, utilizing the expertise of the UK Libraries Subject Headings Interest Group. The goal was to prioritize local headings and decisions to make the library catalog more inclusive. Problematic subject headings would be removed from display, replaced with local headings with inclusive language.

The UK Libraries Subject Headings Interest Group is an informal group that began to meet in 2021. Group members represent archives, special collections, reference and instruction, law, cataloging, and systems. Meeting every other month, the group offers an opportunity to have discussions that help identify problems with LCSH. The group refers to the Problem LCSH List posted on the Cataloging Lab website, which is a collaborative wiki. Perera said the Cataloging Lab offers controlled vocabularies for underrepresented groups, ways to propose new subject headings or revisions to existing headings, and serves as a clearinghouse of statements of violence in description.

The UK group makes use of a proposal form and a decision-making process to prioritize projects. The process has been formalized and made available to any library employee. Projects are then approached through an established workflow: problematic LCSH are hidden; alternative headings are displayed from authority records; and/or local headings are added or displayed. An example was shown with records containing the LCSH “Problem children.” The heading was suppressed and replaced with a local heading “Children with disruptive behaviors.” Perera noted that LC headings are not removed from MARC records; rather, the changes are accomplished with the way records are displayed in the Primo discovery layer. Patrons using the search term “Problem Children” will see instead “Children with disruptive behaviors” in the bibliographic record. UK Libraries have the luxury of being able to make such a local decision to change the display, because they do not share their catalog with other institutions.

Another example of a local solution to a problematic subject heading was to suppress the authorized access point displayed in MARC records, and show variant access points instead. For example, instead of displaying “Illegal immigration,” the two variants “Unauthorized immigration” and “Undocumented immigration”—already present in the authority record—were displayed. Again, the undesired terms are not removed from MARC records; this solution is achieved visually though the discovery layer.

Perera urged us to work to create meaningful change: as her local committee has shown, ethical description is not isolated, or “siloed,” work; furthermore, we should have as a goal to recognize the value of DEIA in every function of our organization, and not just in the catalog.

Small group discussions

Following the presentation, attendees participated in small group discussions using the following prompts:

• What resources do you use (besides Music Cataloging Bulletin (MCB)) and Library of Congress (LC) lists) to keep up with changing terminology?
• Does your catalog index and/or display any of the 3xx fields related to creators and audiences?
• How can we as a community balance the desire to create access (e.g. a search to find works by female composers) with the need to be ethical in descriptive practice (e.g. not assigning gender because of the many issues surrounding gender)?
• How can we work together to share what we have learned? Can we build a community of practice within MLA around ethical music description and subject analysis? How can we collaborate to do it better?