HALF A CENTURY OF MODELING MEDIEVAL MEANING. What Endures When Technologies Change?

This contribution reflects on the long-term development of the Mittelhochdeutsche Begriffsdatenbank (MHDBDB), one of the oldest Digital Humanities projects in the German-speaking world, as a case study in modeling medieval meaning under changing technological conditions. Starting from the premise that modeling meaning is a central task of Digital Humanities, the paper traces how different technological paradigms—from early database systems to Semantic Web infrastructures—have shaped, constrained, and at times obscured access to the data itself. Particular attention is paid to the challenges arising from highly interconnected RDF-based models, including issues of scalability, maintainability, and human readability within small research teams. The contribution argues that increasing technical complexity does not necessarily lead to better knowledge production. Instead, long-term usability depends on factors such as transparency, maintainability, and the ability of domain experts to understand and actively work with their own data. Against this background, the recent transformation of the MHDBDB towards a TEI-based, human-readable data structure is presented as a strategic shift toward greater autonomy. This includes not only independence from proprietary systems and heavy infrastructure, but also a reconfiguration of workflows that foreground interpretability and sustainability. A further focus lies on the integration of Large Language Models (LLMs) into the redevelopment process. Rather than relying on automation, the project employs controlled, expert-driven workflows, including prompt-based reconstruction, iterative validation, and human review at every stage. The case of the MHDBDB demonstrates that what endures across technological change is not a specific data model, but a set of principles: the need to model meaning explicitly, the central role of human expertise, and the responsibility inherent in every act of modeling. Ultimately, the paper argues for a more cautious, reflective approach to digital infrastructures—one that prioritizes clarity, control, and sustainability over maximal complexity.