The Dawit Isaak Database of Censorship (DIDOC) is a collaborative digital humanities initiative designed to document and analyse censorship of literary and journalistic works. Developed by the Dawit Isaak Library (DIL) in Malmö together with the Gothenburg Research Infrastructure for Digital Humanities (GRIDH), Swedish PEN, and researchers at Lund University, the project brings together cultural heritage expertise, freedom-of-expression advocacy, and digital infrastructure to create a foundational resource for comparative censorship research.

The project originated in 2023 through a collaboration between DIL and literary scholars at Lund University, later joined by Swedish PEN, whose involvement in Swedish Banned Books Week and advocacy for persecuted writers brought essential political and legal expertise. In 2024, DIL also approached GRIDH due to its experience developing Queerlit, a bibliographic database based on linked data. Together, the partners initiated a pilot to explore how censorship could be systematically documented, ethically represented, and made publicly accessible.

DIDOC is developed in Omeka-S, an open-source platform well suited for semantically structured digital collections and linked open data (LOD). The choice of Omeka-S was driven by the need for a robust, extensible system that could be self-hosted, support multilingual metadata, and allow fine-grained access control given the sensitivity of censorship data. GRIDH funded and implemented the pilot as part of its strategic work on “Responsible DH,” enabling rapid prototyping through Omeka-S’s modular architecture and intuitive data curation interface.

The DIDOC data model is built around six entity types—work, event, concept, person, organization, place—allowing bans to be modelled as events in the life of a work. Each event records the type of censorship, stated reason, place, date, source, and certainty, enabling the annotation of statements such as “This book was removed from schools in 2021 due to political content.” By using established LOD vocabularies such as Dublin Core, FOAF, and SKOS, DIDOC ensures semantic interoperability and supports future integration with external datasets, including the now-defunct Norwegian Beacon for Freedom of Expression, which documents nearly 50,000 censored works.

The pilot interface, released in October 2024 during Banned Books Week, included 160 curated titles from DIL’s collection of 1,200 censored works (now expanded to 271). It offers faceted exploration of censorship events, contextual editorial material, and a unified search across metadata and narratives of suppression. Over its first 11 months, the site received more than 3,000 visitors, indicating public and scholarly interest in the material.

A central challenge of DIDOC is ethical navigation. Not all censored authors can be safely documented, and censorship data is often incomplete, contested, or politically sensitive. Through collaboration with Swedish PEN, the project has established workflows for reviewing and withholding records when documentation may endanger authors or reproduce harm. These tensions—between transparency and protection, documentation and silence—shape the design and interpretation of the database.

DIDOC demonstrates how digital humanities infrastructures can illuminate patterns of repression while foregrounding the ethical responsibilities of documenting censorship. By combining linked data technologies, interdisciplinary collaboration, and careful metadata practices, the project lays the groundwork for a comprehensive, multilingual, and globally relevant research resource.