Revising Archival Descriptions: part 1 – the methodology

Increasing the accessibility of information about the collections in our care is a priority for all archive services. To achieve this we must consider the language we use as an integral and fundamental part of our collections management work.

The Inclusive Cataloguing Working Group of the Diversity and Inclusion Allies (ARA) has highlighted some aspects and some issues and we hope this contributes to the ongoing discussions. We need to be aware of the impact the language used in our online catalogues has on our users and that changes also have the potential to directly impact the discovery and use of the material.

Revising Archival Descriptions project

Archives and Records Council Wales (ARCW) has been awarded funding from the Welsh Government’s Anti-Racist Wales Action Plan to identify biased and offensive language in archival catalogues. The Revising Archival Descriptions project has adopted the methodology highlighted by the University of Leeds Archive Testbed project award from The National Archives in April 2021. This project saw the University Library collaborate with the Leeds Arts and Humanities Research Institute to look at what they described as ‘problematic language’, the barriers they represented to engagement, and how this could be supported through the use of AntConc open-source linguistic software.

Methodology

Revising Archival Descriptions is a project looking to apply the methodology to a significant sample of collections from across a range of archive services. The larger the sample the stronger the evidence for any conclusions and recommendations.

AntConc works with plain text files. We’ve primarily been using EAD finding aids from services or via the ArchivesHub, with MS Word catalogues being saved as XML and processed with no problems. The catalogues have been searched against a terms list, in our case the ‘Brotherton Full list’ which features over 1000 terms that are racist, sexist and misogynistic, ableist, homophobic, transphobic, trans-exclusionary, feature religious discrimination, and those that are generally offensive. The use of * wildcard significantly extends the breadth and scope of results returned which can be viewed within AntConc or saved as .csv file and viewed in MS Excel.

The work is very much in-progress but it is due to conclude at the end of February. A second blog will detail the process of reviewing the hits from AntConc, the headline results from the project, and possible next steps.

For more details on this project, please contact Vicky Jones: vicky.jones@llgc.org.uk