I. Introduction
In Winter 2021, the University of Washington (UW) Libraries launched the CritCat group, composed of five staff members with backgrounds in cataloging and archival work. This group, with input from colleagues across the Libraries, developed the UW Libraries Critical Cataloging and Archival Description Statement, addressing harmful language in catalog records and finding aids. The statement outlines our efforts to remediate harmful language and provides an explanation of the complexities involved and technical limitations in the remediation process. The statement page also features links to a feedback form and an email address for users to request changes to metadata. Through the pilot project described below, the CritCat group has evolved into the Critical Cataloging, Archival Description, and Metadata Working Group (CritCat), comprising catalogers, metadata specialists, archivists, and a systems specialist.
In February 2023, the UW Libraries Metadata Implementation Group (MIG) initiated a pilot harmful-language remediation project to remediate the harmful term ‘Eskimo(s)’ used in the CONTENTdm digital collection ‘Harriman Alaska Expedition of 1899.’ The pilot project also planned to establish a model for upcoming metadata remediation initiatives at the Libraries. Running through December 2023, the pilot project engaged staff members with backgrounds in archival work, cataloging, and systems management. Outcomes include the development of a checklist for future remediation projects and correction of the derogatory term ‘Eskimo(s)’ in metadata across various platforms including CONTENTdm, OCLC, and Archives West, as well as in digital-collection landing pages.
It is noteworthy that the project played a pivotal role in reshaping the CritCat group at the University of Washington Libraries and establishing a new workflow for executing these metadata remediation projects.
II. Project scope
The impetus for the pilot project was a submission of the UW Libraries Harmful Language in Descriptive Resources feedback form in December 2022 pointing out use of the harmful term ‘Eskimo(s)’ in the digital collection ‘Hegg (Eric A.) Photographs of Alaska and the Klondike, 1897-1901.’ The term was replaced in the Hegg collection, but searches yielded more than 700 results from a variety of other digital collections. The UW Libraries MIG discussed these results in February 2023, and members voiced interest in forming a project group to start replacing this term where it appeared in digital-collections metadata.
A working group was assembled and selected the digital collection ‘Harriman Alaska Expedition of 1899’ as a test collection for developing a remediation workflow. It was selected due to the relatively limited number of items which included the term in metadata descriptions–23 in total.
During initial discussion, members had pointed out frequent overlap between item titles in metadata for digital objects and those in finding aids describing archival visual resources used to create digital collections. The finding aid for the archival collection associated with the digital collection ‘Harriman Alaska Expedition of 1899’ exemplified this, with item titles in the finding aid used as titles for corresponding digital images in the online collection.
Because using different titles and other metadata values for a physical object and its digital surrogate would be confusing to users, the working group decided to expand the project scope to include the archival finding aid. This made it clear that the working group needed participation from staff in Special Collections responsible for the archival collection.
The project scope was expanded further to include OCLC records following review of MARC descriptions for the archival and digital collections, which also contained the harmful term. Finally, Special Collections staff pointed out that the term was also used in a locally maintained Microsoft Access database, and this was also added to the project scope. This database is used to provide accompanying information for print reproductions of images from digital collections.
III. Workflow and communication
The pilot project was a collaborative effort involving three departments at the University of Washington Libraries: Cataloging and Metadata Services (CAMS), Special Collections (SC), and Information Technology Services and Digital Strategies (ITS/DS).
The initial team, comprised of six members of MIG with backgrounds in cataloging, metadata, and systems, initiated the project and selected the ‘Harriman Alaska Expedition of 1899’ for this pilot project. The group identified a need to edit metadata values in multiple fields in the digital collection, expanded the project scope to address metadata in various formats, and asked staff responsible for the archival collection and finding aid to join the project working group. The working group also met with the University of Washington Libraries principal cataloger Adam Schiff to propose and review Library of Congress Subject Headings (LCSH) for use in controlled fields in the digital collection. Review of LCSH for the project resulted in three new LCSH proposals, described below in section IV.
Editing tasks included the following:
- Edit item-level metadata for digital collections (CAMS, SC)
- Make initial edits to the finding aid in Archives West (SC): Align corresponding item descriptions with new digital-collections metadata
- Edit MARC records for digital and physical collections in OCLC (CAMS): Update LCSH and other headings where harmful language was present
- Make subsequent edits to the finding aid in Archives West (SC): Bring updated LCSH from MARC into the finding aid
- Edit copy-negative database (SC)
- Edit digital collection landing pages (ITSDS)
IV. Solutions
In March 2023, the Library of Congress (LC) announced its commitment to updating headings for Indigenous peoples living in the United States. LC’s interim guidelines for subject- and demographic-term proposals covering Indigenous peoples indicate they may hold LCSH proposals until a broader plan is developed.
In accordance with the limited allowable cases outlined in the interim guidelines, the group successfully proposed the new LCSH term ‘Inupiat children,’ aligning it with the existing term pattern for ‘Inupiat women.’ Two other LCSH terms were also successfully proposed as a result of the project, in collaboration with the principal cataloger: ‘Tupiks,’ and ‘Providence Bay (Siberia).’
A note on replacement terms: For the uncontrolled fields ‘Title’ and ‘Notes’ we used the terms ‘Siberian Yupik’ and ‘Iñupiat’ to describe members of two Indigenous groups depicted. For the controlled ‘Subjects (LCSH)’ field, we used ‘Indigenous peoples’ and ‘Inupiat’ (both subdivided by geographic location).
Bringing together necessary staff in the project working group and identifying the need for cross-unit collaboration in the future were significant project achievements. Metadata describing both a digital and a physical collection would be remediated, and changes needed to be made in collaboration with staff having knowledge of and responsibility for each. Metadata in a variety of formats needed to be changed, requiring staff with working knowledge of metadata in the digital-collections platform, EAD finding aids, MARC metadata, and a locally maintained Microsoft Access database. Staff with knowledge of the collection’s context and willingness to perform research to identify new terms was essential. Metadata for this project required values from LCSH and other controlled vocabularies, requiring staff who could perform subject analysis and construct valid headings.
The process of assembling the working group for this pilot project also serves as a template for future work, as group members felt it helpful to identify overlapping metadata descriptions and define a project scope at the outset. In this example, we found metadata in four locations which described the same or closely related information resources, all of which included the harmful term targeted for remediation.
V. Challenges and next steps
Tracking changes in descriptive metadata is a remaining challenge for this work. Storing deprecated harmful terms in a hidden, searchable field allows us to provide relevant results when the term is used for searching without displaying harmful language, but recording values in this field is labor-intensive. This approach is expected to be cost-prohibitive in large projects. (For details regarding implementation of a hidden field in the pilot project see the collection’s metadata application profile.)
For staff, deprecated metadata values provide a record of work which has taken place to date and may be used to facilitate future metadata improvement. The harmful term ‘Yuit Eskimos’–which is still valid LCSH–had been applied to photographs depicting Indigenous peoples of Siberia. Because no equivalently specific non-harmful alternative is available within LCSH, the broader term ‘Indigenous peoples–Russia (Federation)’ was applied. If LCSH publishes a more desirable specific descriptive heading in the future, searching for the deprecated would facilitate implementing the new term.
The project work highlighted additional challenges for metadata staff at our institution. The working group recognized that metadata for the digital collection needs cleanup beyond remediation of harmful terms, but this fell outside the scope of the pilot project.
VI. Further information
For additional information about this and other UW Libraries critical cataloging, archival description, and metadata projects, visit https://github.com/uwlib-mig/critcat.