Serial administrative records – acquisition journals, loan registers, correspondence indices – have long occupied a marginal position in historical scholarship, treated as auxiliary sources rather than primary data. This paper argues that such documents constitute research data in their own right and proposes an event-centred modelling framework that makes them analytically accessible as structured relational data.

The empirical basis comprises approximately 1,800 pages of handwritten acquisition and loan journals from two Viennese herbaria, spanning 146 years. Rather than modelling entries as datasets about objects, the framework treats each journal entry as a historical event, capturing temporal and spatial references, participating actors, object-related information, and administrative context. Uncertainties are retained as explicit attributes; systematic omissions are interpreted as historically significant structural phenomena rather than mere data gaps.

The transferability of the modelling logic beyond serial journal material is tested on a structurally related yet functionally distinct administrative corpus: nine subsidy files documenting Heinrich Handel-Mazzetti's botanical expedition to China (1913–1919). Although differing in institutional context, these files share the same underlying administrative logic of recording transactions and decisions. Sequential linkage across 23 years reveals dynamics invisible when files are considered in isolation – approval timescales as indicators of institutional trust, non-linear temporality of expedition work, and the structural role of private funding.

The methodological contribution lies in repositioning routine administrative documents epistemically: not peripheral material, but independent infrastructure data. Event-centred relational modelling thus extends the methodological repertoire of Digital Humanities to a category of sources long treated as ancillary to historical inquiry.