Abstract (summary)
This study compared the use of lexical bundles in academic writing in Applied Linguistics across three corpora: expert writers, native students and non-native students. The expert corpus consisted of articles published in Applied Linguistics journals; the native student corpus consisted of MA dissertations of native English students who did an Applied Linguistics Masters in English universities. The non-native student data consisted of Applied Linguistics MPhil dissertations of Pakistani students who did their MPhil in Pakistani universities. The size of the three corpora were as follows: native student corpus:312981, non-native student corpus:502945, expert writers' corpus:505958. The highly frequent bundles used in the three corpora were categorized into structural and functional categories (Hyland, 2008). These bundles were analyzed quantitatively as well as qualitatively. The findings revealed that the expert writers were different from native and non-native students in their use of structural and functional bundles. The expert writers used more Phrasal bundles and more bundles for organizing the text than the two student groups. The expert writers also showed better control of bundles for hedging. The students, on the other hand, used more bundles for describing research. Occasionally, they used vague and informal bundles, especially for quantifying. The non-native Pakistani students used far more bundles for describing the procedures of research and used far more bundle tokens than the other two groups. This might be due to the larger size of their dissertations. Interestingly, most of the differences between expert and student writers in their use of bundles applied to both sets of students. This suggests that the main challenge for all students is learning the conventions of academic writing, rather than any problems linked to non-nativeness. Therefore, the appropriate use of bundles in academic writing might need to be taught more explicitly to both native and non-native students.