Translating Core Legal and Business Term Families in Chinese-English Legal and Regulatory Texts: A Clause-Level Corpus-Based Study

Abstract

This article examines the translation of core legal-business term families in official bilingual legal and regulatory texts issued by China’s Ministry of Commerce and its affiliated official channels. Combining a corpus-based approach with a clause-sensitive analytical framework, it takes the term token as the primary unit of analysis and the clause or provision as the auxiliary unit. The study focuses on four concept families—liability, breach, termination, and indemnity—and evaluates how they are rendered across different legal-functional environments, especially obligation or prohibition clauses (OBL) and limitation or exemption of liability clauses (LIM). The completed coded dataset contains 218 tokens drawn from a parallel corpus of authoritative Chinese legal-regulatory texts and their official English reference versions. The results show that literal translation (LT) is the dominant strategy in the corpus, followed by normative modulation (NM), while free translation (FT), communicative restructuring (CR), and mixed FT + CR occur far less frequently. At the clause-function level, OBL clauses are strongly LT-dominant, whereas LIM clauses display a more mixed profile in which LT, NM, and FT all remain salient. The article supplements the descriptive counts with representative qualitative examples in order to clarify how strategy labels are operationalized in context. The findings suggest that, in this corpus, translators rely primarily on direct lexical correspondence and conventionalized legal-administrative formulations, with freer reformulation used selectively in functionally sensitive contexts. The article argues that legal accuracy in this corpus is achieved mainly through stable terminological transfer and normative conventionalization rather than through extensive adaptive reformulation, while also stressing that the results remain descriptive and corpus-specific.