Abstract
This article seeks to contribute to the body of research on the formulaicity of academic written English by adopting a text-internal perspective on the use of lexical bundles across four main rhetorical sections (i.e. Introduction, Method, Results, Discussion) of research articles in public health. The corpus linguistic approach was adopted to explore a section-coded corpus of 200 research papers published in high impact journals with the aim of identifying salient shared and section-specific 4-word bundles, calculating their respective frequencies as well as classifying them structurally and functionally. The results show that the Method is the most formulaic section, containing the greatest number of bundles, including those most frequent and section-specific ones. The sections also differ in the extent to which they exploit the distinct structural and functional categories of shared bundles to fulfill their individual communicative needs.