Purpose of this policy
The collections held by University of Leeds Libraries’ Special Collections and Galleries cover a huge span of history and subject areas. We are aware that content in the collections and catalogues may therefore include sensitive, offensive or outdated language and content. We want to address these concerns in a practical and balanced way – maintaining the historical context of when the record was created but also ensuring that none of our audiences feel excluded or misrepresented.
This policy sets out our approaches to sensitive language and content, drawing on research carried out in Special Collections and Galleries as well as from the wider archives and heritage sector.
Sensitive language and content, and the scale of our collections
Special Collections and Galleries hold millions of individual items. This scale affects our approaches to sensitive language and content. Examining every item in every collection manually to make case-by-case decisions is not feasible given the resources available.
Instead, as a first stage of identifying sensitive language and content, we employ automated searching and flagging. We check the results of these searches as thoroughly as possible but we cannot ensure that all the records we flag will actually be sensitive, or that all sensitive records will be automatically identified and flagged. This means that the flagging of sensitive language and content on the catalogue can appear inconsistent.
When resources allow, particularly in response to external funding or in response to user feedback, we may undertake manual audits to retrospectively amend existing collections to improve and balance our catalogue. This approach means that some of our collections are flagged more than others – this should not be interpreted as these collections being any more or less sensitive or offensive than others that have not been audited.
New collections and current cataloguing projects will have approaches to sensitive language and content built in at the outset, which will ultimately lead to a consistently balanced catalogue.
We are aware that language is subjective and always changing, there is often no single term that can be used to represent everyone in a community and that the acceptability and meaning of words changes over times. We will review our sensitivity workflows and cataloguing practice periodically.
Our approaches when we find sensitive language or content
We approach all our collections with inclusive cataloguing in mind. We aim to describe our collections in a way that is accurate, inclusive and respectful to the people who created them, use them, and feel represented by them.
We acknowledge that this work is vitally important, and that the challenges of sensitive language and content need addressing, but we also acknowledge that there is no universally agreed way to achieve this. We therefore aim to be fully transparent about the workflows we follow, the review processes in place, and any changes that are made.
Below we set out our different approaches to sensitive language and content, explaining the context and reasoning behind each one.
When the content or subject area of an object or document is judged to be potentially sensitive or offensive, a sensitivity statement will be applied to warn users of the potentially upsetting or offensive content or language.
Maintaining original terminology
When sensitive terminology has been used by the creator in the original record and links to an important aspect of the record, this will be maintained. We will put this original language in quotation marks, as a way of separating historic terminology and opinions from the current values and beliefs of the University. If necessary, we will also provide context explaining the language. In order to warn our users of the potentially upsetting language that remains in the record, these records will be flagged with a sensitivity.
When we reference flagging a record, or attaching a sensitivity statement, we are referring to the standardised access statements used by Special Collections to mark-up records with any additional information useful to users.
When sensitive language has been added in later by a curatorial voice, but the term nonetheless links to an important aspect of the record, the language will be replaced with an alternative but equally meaningful and accurate term. If the term is judged to not link to a key aspect of the record, this language may be removed from the description. Any editing that has taken place will be clearly flagged on the record and the original description will be saved and be accessible on request.
Collection-level sensitivity warnings
Sensitivity warnings will be applied to collections where a detailed audit has not yet been possible, but where it is deemed likely that there are sensitive descriptions in the catalogue or sensitive content. Individual records will not be flagged, but this approach provides a general level of warning for users interacting with the catalogue.
We will provide a general content warning on the landing page of the Special Collections and Galleries webpage that explains the context of historic language, the possibility of sensitive or offensive language in our catalogue, and clearly signposts this publicly available policy.
We want to be transparent about our approach to sensitive language and content in our archive descriptions. As explained above, our processes cannot guarantee that all the records we flag will be sensitive, or that all sensitive records will be flagged.
To minimise this margin of error we encourage users to give feedback on our processes or how records have been flagged. You can do this by emailing Special Collections at email@example.com, where the enquiry will be passed to our archivist team.
This policy was approved in February 2023. We will review the policy after one year. This might be extended to a three year review period after the policy is fully established.