Lexical bundles in TED talks: Topical variations and pedagogical implications

Abstract

As an important component of fluent linguistic production, lexical bundles have attracted considerable attention over the past two decades in corpus studies in EAP, especially in the study of written academic registers in English. Spoken texts, however, have been rather underexplored, hence examining differences across disciplines in spoken academic discourse remains open. Drawing on a one-million-word corpus of TED talks, a pseudo-academic register representing popularized science in the spoken mode, the present study investigates the quantitative and qualitative features of four-word lexical bundles, specifically their forms, functions, and topic-specificity of target bundles in five broad topical areas. The results showed that lexical bundles can not only reflect the generic features of discourse, but also serve as an important means of uncovering intra-register variations by topic. This paper concludes with a consideration of pedagogical implications of the findings for EAP listening and speaking. This article was published open access under a CC BY licence: https://creativecommons.org/licences/by/4.0 .