Abstract :
This corpus-based analysis compared the move structure and lexical bundles (LBs) of thesis abstracts written by native Persian (L1-Persian) and native English (L1-English) writers. The main aims of the study were to compare the move structures of L1-Persian and L1-English thesis abstracts, identify the most frequent lexical bundles per move in each language, expose significant structural patterns of LBs in both languages, and investigate their functional characteristics. With the use of qualitative techniques, moves were analyzed manually, while AntConc 4.3.1 software was used to identify lexical bundles, which were then structured and functionally categorized. Results indicated that L1-Persian MA and PhD thesis abstracts had more coherent organization, compared to their L1-English ones, with about 25% of L1-Persian abstracts including all five moves represented by Swales and Feak and almost half including four moves. In contrast, L1-English abstracts revealed a lower rate of full five-move patterns, while PhD abstracts were more organized than those of MA. Surprisingly, the L1-Persian corpus revealed a greater percentage of lexical bundles per move, reflecting greater structural coherence. Structural analysis found noun phrase-based and phrasal structures to predominate across both corpora, with no verb phrase-based structures appearing in L1-English abstracts. Functionally, the majority of LBs in both corpora were research-oriented. Furthermore, text-oriented LBs were more frequent in the L1-English corpus, whereas participant-oriented LBs were more frequent in the L1-Persian corpus. The results of this study could have significant implications for human’s knowledge about linguistic variations in academic writing, providing guidance for non-native English writers, guiding the design of academic writing courses, enhancing genre studies, and enhancing peer review and publication procedures.
