Crossparametric dependencies and Social Network Analysis

The dependency structure of syntactic parameters (Baker, 2001; Guardiano & Longobardi, 2017; Roberts, 2019) poses a challenge for phylogenetic inference, as standard methods (Green- hill et al., 2017) are not designed to accommodate character interdependencies. Building on computational work that represents crossparametric dependencies as network graphs (Kazakov et al., 2018), this study proposes a pilot implementation of Social Network Analysis (SNA) to model parameter systems as relational networks. While the use of SNA is not new in the study of linguistic interactions (Li, Li, & Gao, 2025), its potential for modeling abstract systems of language properties has remained largely unexplored. We pursue two goals: (i) assessing crosslinguistic similarity through network comparison, and (ii) identifying central nodes within parameter networks. Our results show that parameter systems can be characterized using canon- ical SNA metrics, enabling the quantification of node-level centrality. Systematic node removal yields measurable shifts in similarity scores across language pairs, indicating that parameters contribute asymmetrically to convergence and divergence patterns. These findings suggest that SNA methods may significantly enhance the historical study of syntactic variation by integrating dependency structures into parametric phylogenetic modeling.