The study of lexical bundles, known as fixed phrases, chunks, clusters and multi-word expressions, has attracted considerable attention. While there has been much research on lexical bundles across different registers and a number of disciplines, their deployment in some special sections of research articles as the most high-stakes genre has not yet been well explored. Accordingly, the present study aimed at identifying 4-word lexical bundles by analyzing the data obtained from a collection of the abstract and conclusion sections of 1000 English research articles written by L1-persian and L1-English writers in AL and IT , as published between 2015 and 2019. The researchers used Antconc software to analyze the data composed of about 600,000 words; then, the functional analysis was carried out based on Hyland's (2008) framework. Overall, the analysis revealed that AL writers outweighed their IT counterparts in their use of lexical bundles. Also, L1-persian writers used more lexical bundles in the abstract section; despite this, both writers used the same number of bundles in the conclusion section. In addition, both AL and IT writers had a similar use of the three main functional categories; however, there were substantial differences and similarities in regard to these two parts of research articles. The findings of this study can help writing instructors to improve students' academic writing. They can enhance their abilities to better comprehend the role of lexical bundles in different genres and sub-genres.
