In the context of the arts and humanities, heterogeneity largely corresponds to the variety of disciplines, their research questions and communities. Resulting from the diversity of the application domain, the analysis of overall requirements and the subsequent derivation of appropriate unifying schemata is prevented by the complexity and size of the domain. The approach presented in this paper is based on the hypothesis that data integration problems in the arts and humanities can be solved on the theoretical foundation of formal languages. In applying a theoretically substantiated framework, integrative solutions on the formal basis of language specifications can be tailored to specific and individual research needs—abstracting from reoccurring technical difficulties and leading the focus of domain experts on semantic aspects.