The digitization of historical material at information and cultural heritage institutions is taking place as histories of colonialism increasingly take center stage in the ways libraries and museums reconsider their own past. The intersection of these fields raises important questions about the historical and ethical dimensions of digitizing historical material today. As international bodies develop guidelines to operationalize principles and standards for digitization, debates within the digital humanities have taken up the issue and urge a critical reflection on the practices of digitizing historical material from colonial contexts.
The ordering of knowledge across European information and cultural heritage institutions has a long history. Modern systems and methods of organizing collections and objects long preceeded our digital moment. In the late eighteenth and early nineteenth century, museums, scientific societies and emerging academic disciplines, for example, began to organize knowledge and objects of the non-European world in new taxonomies and classifications. In doing so, they helped craft what was unknown or labelled different as objects of scientific inquiry, often casting regions and their peoples in exoticized and racialized terms that was used to justify colonial regimes of exploitation and oppression.
The digital re-ordering of our time may not resemble its nineteenth-century equivalent in all respects. There are, however, historical echoes that call for critical reflection. The digitization of historical collections creates new forms of data and with that new ways of describing data. In this context, the Research Data Alliance and the Global Indigenous Data Alliance released the FAIR (2016) and CARE (2019) principles to ensure accessible and ethically responsible data stewardship. (While FAIR data must be findable, accessible, interoperable, and reusable, CARE data includes collective benefit, authority to control, responsibility, and ethics as criteria.) As our project IN_CONTEXT: Colonial Histories and Digital Collections sets out to digitize historical material and to develop a framework for a virtual research environment, the project continuously confronts fundamental questions when recreating physical sources in the digital realm: How do we deal with historical classification schemes and the adoption of contemporary designations? How do we negotiate the tensions between requirements from funding agencies, institutional standards, and international guidelines developed to ensure FAIR and CAREful data stewardship? How do we make accessible exoticizing and racist source material? Are editorial layers across website interfaces sufficient? How can such practices be extended to the design of APIs, computational access points to retrieve data? And who ought to be at the table when making these decisions?
These questions were the subject of a panel discussion on “Digitizing Cultural Heritage and Postcolonial Perspectives” at the 9th annual meeting of the Association for the Digital Humanities in German-speaking Countries (DHd) at the University of Trier. IN_CONTEXT’s John Woitkowitz and Human.Machine.Culture project co-lead Clemens Neudecker joined colleagues to reflect on the ethical dimensions of digitization projects. In a wide-ranging discussion, the panel pointed to the opportunities of digitizing collections and applying computational methods to digitized material. Global access to historical material that relates to colonial histories is valuable for research communities but specifically so for those communities where material was removed and who continue to deal with the legacies of colonial violence and exploitation. Likewise, identifying bias and sensitive content with the help of artificial intelligence-assisted tools opens up new opportunities for data stewardship and research methodologies.
At the same time, significant challenges exist to comply meaningfully with the CARE principles when dealing with colonial histories. Co-curation of data and the co-development of information infrastructure, such as web-based repositories and research environments, are important to mitigate bias and to vest interpretive authority within source communities. Such work, however, cannot be short-term. Building trust-based relationships that sustain projects requires long-term commitments, a key aspect that often is complicated by precarious employment, funding timelines, and uncertainty over the sustainability of project work. Likewise, grant stipulations that require open access to digitized material may not always be compatible with accepted usage as prescribed by source communities. Diverse degrees of digital infrastructure further complicate the notion of place-agnostic access to digitized material, an aspect that calls for rethinking platform development and digitization processes.
Accessibility and collective benefit, two important FAIR and CARE principles, also touch on aspects of language. The DHd working group Multilingual has taken an active role in raising awareness and providing a platform to further advance critical perspectives in this field. For users from a diverse range of linguistic communities to be able to understand and benefit from material and resources digitization projects make available, multilinguality needs to expand as a standard in information institutions. It also needs to be included in design processes across data modelling, user interface development, and API design if digitized material is to cut across existing linguistic boundaries.
In many ways, going digital in a global information environment is reconnecting knowledge and collections with their histories and legacies. This digital moment, however, is not without its contradictions. Digitizing material that relate to colonial histories cannot be done without the collaboration of those affected by these very histories. Instead, it must include meaningful collaboration and the resources to sustain such work. As debates evolve and new standards emerge, there will be imperfect solutions along the way. But as we reshape the knowledge orders in our time, there is ample room to push for a more equitable model of information stewardship.