The purpose of this study is to: Connecting lexical bundles and moves in research article introductions

This article presents a group of lexical bundles identified in a corpus of research article introductions as the first step in the analysis of these expressions in the different sections of the research article. A one-million word corpus of research article introductions from various disciplines was compiled and the lexical bundles identified in it were classified grammatically and functionally. The findings of these analyses agreed with previous studies in the most frequent types of grammatical correlates for these bundles and the most frequent functions performed but showed several new qualities for these expressions (Biber, Johansson, Leech, Conrad, & Finegan, 1999; Biber & Conrad, 1999; Biber et al., 2003 and Biber et al., 2004). A further step in the analysis matched these lexical bundles to the moves and steps which are characteristic of research article introductions (Swales, 2004), discovering that a group of lexical bundles were exclusively linked to one move or step in a move while a second group occurred across several moves and steps. In addition, some of these expressions were used to trigger the steps that called for their use while others complemented other expressions and were used as comments.

► The study examined the relationship between bundles and moves in introductions. ► Corpus processing yielded lexical bundles longer than six words, uncommon in English. ► Bundles were classified structurally and functionally and in relation to move scheme. ► A contextual analysis showed that longer bundles act as move triggers or comments.