The conceptual semantics of English ‘speak’ (and why it matters)
This study investigates the conceptual semantics of the English verb speak, using the Natural Semantic Metalanguage (NSM) approach to semantic analysis. The analysis is corpus-assisted, relying mainly on two corpora: one very large (enTenTen21, more than fifty-two billion words), the other small and purpose-built (about 500 ‘speak’ sentences from twelve recent novels). Using the traditional “definitional” criterion for polysemy, the analysis recognises four senses of English speak, which form a polysemic web or network of meanings and usages. A range of supporting lexicogrammatical evidence is adduced, such as the proposed polysemic meanings having distinctive morphosyntactic properties, collocational profiles, and associated derivatives and phraseology. The paper also raises an ethnoaxiological question for linguistics as a discipline: To what extent has the discourse of mainstream Western linguistics been guided or shaped by an Anglo-English metapragmatic model of speaking, whereby “saying” is framed as individual, purposeful, contentful, and oral/aural?
Publication history
The verb speak, together with its derivatives, plays a tremendous role in the Anglophone vocabulary of theoretical, descriptive and applied linguistics (cf. Levisen 2025). It is no exaggeration to say that one can hardly “do linguistics” in English without using the word speak. English language monographs, textbooks and research studies are peppered with phrases such as those in (1) and (2).