Same Difference: Identity and Diversity in Linked Open Cultural Data
Susan Wightman Brown · International Journal of Humanities and Arts Computing · 2022
Linked Open Data (LOD) was designed to respect heterogeneity in source datasets. However, the fundamental mechanisms of interlinking require sameness without nuance, so Linked Data is at risk of the problems associated with lack of diversity in big data generally. This article investigates the tension between difference and sameness specifically as it relates to asserting the identity of entities. It links ambiguity in natural language and cultural expression to Derrida's notion of différance and the foundation of Linked Data structure in Peircian semiotics. Representing entities so as to foreground rather than suppress subtle differences or ambiguity is a challenge given the lack of anything in between owl:sameAs and owl:differentFrom that can be supported by formal logic. The reuse of Uniform Resource Identifiers (URIs) recommended as a best practice for data interlinking is troubled by confusion over the relationship between URIs and representation, whereas owl:sameAs suffers from a range of forms of misuse. Despite these challenges to representing nuance and ambiguity, however, there are several ways in which humanities researchers and cultural institutions can pursue better means of representing diversity and difference using LOD, particularly through interdisciplinary and multisectoral collaboration.