Exploring metalinguistic patterns with large-scale corpora: A case of Chinese notions of face and politeness

Scholars have repeatedly emphasized that clearly and explicitly defining a concept underscores good research. Drawing on an emic approach to firstorder concepts, metapragmatics offers researchers a viable method to assess the difference between their theoretical definitions and laypersons’ everyday interpretations. Powered by corpus linguistic methods, researchers can automatically extract patterns of metalinguistic expressions from large-scale corpora. Using Sketch Engine (Kilgarriff et al. 2014), the metalinguistic patterns are identified using frequency analysis, word sketch analysis, and distributional thesaurus analysis. To illustrate this approach, this chapter provides a case study of Chinese notions of face (mianzi) and politeness (limao). The results show that Chinese speakers perceive face and politeness as distinct concepts in everyday interactions. While politeness is mostly represented by people’s language choice, face covers a wider range of social behaviors. Moreover, the mixed semantic prosody of mianzi suggests that the concept needs to be studied in specific interactions. In contrast, the consistently positive semantic prosody of limao indicates that norms of politeness are shared among community members. In conclusion, this chapter demonstrates the utility of a metapragmatic approach to concept explication. The case study analyzed in this chapter extends corpus-based metapragmatic research to the Chinese context which has received relatively little attention (Haugh 2018).
© 2026 the author(s)