Despite the prolific amount of research into the discourse structure of the research article (RA), comparisons between the different sections of this important register have been extremely rare. This study is an attempt to analyse the structure and form of three main sections of the RA (i.e., Introduction, Method and Results) through examining the frequency, structure and function of lexical bundles found within each. By comparing the three sections, a better understanding of in-text linguistic variation (i.e., differences in the use of lexical bundles across the three sections) can be achieved. For this purpose, a corpus of 200 research articles in applied linguistics, segmented into the three sections noted above, was compiled. A list of four-word lexical bundles occurring at least thirty times per million words in the overall, one million-word corpus was made. Subsequently, each of the three sub-corpora was searched for any of the bundles on the list. The findings reveal the Results to be the most densely formulaic part of the RA, both in terms of the type/token frequency of bundles and the number of unique bundles that were not seen in any of the other sections. This section of the RA also included longer sequences of bundles than any other part. Structural and functional differences were also discovered among the three sections; some of these differences, especially in the distribution of various functional categories, indicated the existence of a close relationship between the discourse aims of a given section and the lexical bundles that are most characteristic of that part of the RA.

Keywords