A pilot project to test the AI-assisted transcription, translation and visualisation of provenance data from colonial military campaigns in German New Guinea
A great many historical documents that are of vital importance to the study of colonial museum collections are currently not easily accessible: they are written in old German script, are difficult to read, and are only available in German – something that makes not only scientific analysis and international dialogue more difficult. Now, for the first time ever, a new pilot project conducted at the Ethnologisches Museum is employing artificial intelligence in an attempt to overcome these hurdles. The aim of the project is to transcribe the documents in question, to translate them into several languages, and to render their content accessible via a series of digital visualisations.
For several years now, there has been growing public debate about how to tackle our shared history of colonialism and its ongoing ramifications, with the collection of the Ethnologisches Museum in Berlin serving as a primary focus of attention. Colonisation, the acquisition of artefacts, and the establishment of ethnological collections in Berlin were all intimately linked. To this day, the Ethnologisches Museum is still home to a considerable number of objects that made their way to Berlin in the context of colonial military campaigns. The key aim of provenance research is to investigate and expose the violent history of these kinds of acquisitions.
As part of this project, innovative digital methods are being trialled for assisting and facilitating provenance research. Building on a previous museum project funded by the German Research Foundation (DFG) to digitise the historical archive of the Ethnologisches Museum (1830–1947), researchers now aim to determine how these documents can be transcribed and subsequently translated into several languages. In addition to this, the visualisation of the recorded data will be tested in order to provide content-related support to provenance research in this area, as well as to gain insight for future projects. The documents that have been selected for the project pertain to naval activities conducted in the colony of German New Guinea and provide contextual information about colonial military campaigns and the museum artefacts that were acquired within this context.
The project is a preliminary test run: What kinds of technologies can actually be employed for provenance research? Just how much human labour is required in order to obtain correct and usable results when using the AI-based technologies that are currently available to us? What are the limits and limitations of automated processes? The insights gleaned here will hopefully contribute to the ongoing evolution of digital tools for researching colonial collections.
Project management: Ilja Labischinski, Hendryk Ortlieb
Research associate: N. N.
Student assistant: N. N.
Institutions: Ethnologisches Museum, Zentralarchiv
Project sponsors: Federal Government Commissioner for Culture and the Media
Duration: 2025 to 2027