Measuring Context Across Cultures: A Multidimensional Corpus-Based Approach

Cross-cultural research has long proposed systematic differences in the degree to which cultures and languages rely on context in communication, most prominently captured in Hall’s low-context/high-context distinction, which reflects the preference for explicit or implicit communication. Yet these classifications have rarely been subjected to large-scale, systematically comparable empirical testing. This interdisciplinary study proposes a corpus-based framework for quantifying implicitness across languages and communities, informed by state-of-the-art research in linguistic typology and corpus linguistics. We analyze the expression of Agents and Time reference of events in a manually annotated comparable corpus of news texts from nine languages representing five language families. Events were identified and annotated using comparative concepts designed for language comparison, and the variation between explicit and implicit expressions was analyzed using Bayesian mixed-effects logistic regression models. The results provide only partial support for the established cultural continua. While patterns of Agent expression broadly align with expectations, Time reference shows more divergence from the previous claims. These findings highlight the need for a multidimensional approach to cross-cultural variation in explicitness and implicitness, while raising broader questions about how culturally shaped communicative preferences interact with language-specific micro-conventions across different domains of meaning.