The document discusses integrating digital epigraphic data from multiple sources. It notes that the same inscription may be referenced by different identifiers in different datasets. When integrating these datasets, temporary duplicate identifiers may be assigned that later need to be merged. Storing the data and relationships in a property graph database rather than a triple store allows for more flexibility in representing complex and changing relationships between editions and inscriptions over time. However, this comes at the cost of some RDF and SPARQL capabilities. Questions are raised about whether RDF is fully suitable for representing contingent humanities facts and relationships.