The global recurrence and variability of kinship terminology structure

The extent to which kinship terminology varies between linguistic groups is a long-debated but unresolved social and linguistic puzzle. Contemporary research shows that a six-category typology is overly simplistic, but no alternatives have reached broad acceptance. This paper takes a data-driven solution to this problem. Using data from the release of Kinbank, a global database of 1,156 kinship terminology, I quantitatively review the global diversity of kinship terminology to derive a more granular typology of kinship terminology. In a two-part analysis, I show that kinship terminology structure is more diverse than is often assumed across three metrics. Firstly, more than six types are needed to represent the global diversity of kinship terminology. Secondly, typological categories are not equally variable. Some categories may contain identically structured terminology. Others may contain languages that only share a single feature. Finally, different subsets of kin (e.g. cousins vs grandparents) show different levels of variability. In the second part of the analysis, I explore the global distribution of the new typological categories, identifying globally and locally recurring structures. My analysis demonstrates how data can carve this semantic domain at its joints to identify observed clusters of diversity.