Lexical bundle use of Turkish and native English writers: A corpus-based study

In recent decades, English has been the lingua franca of the academic and students and scholars are expected to produce written works in this global language. However, this creates a pressure on non-native students and scholars since they need to have a native- like proficiency to be able to fulfill their studies and publish their work, and moreover they need to be familiar with the distinguishing features of academic discourse such as words or multi-word combinations. In this sense, many studies have been conducted to identify such type of combinations, one of which is the recurring multi-word expressions, mostly referred to as lexical bundles. Being composed of three or more words, lexical bundles such as on the other hand and as a result of are extremely common and important in shaping academic discourse. Moreover, lexical bundles vary across different disciplines meaning that successful use of lexical bundles typical of a specific academic discipline is important for writers and the absence of such bundles may not sound fluent and native-like. Recent studies (e.g. Adel & Erman, 2012; Chen & Baker, 2010) have revealed that non-native writers produce not only fewer types of lexical bundles, but also less varied ones. Furthermore, they also overuse a restricted number of bundles in their writing. However, they should have a successful and native-like control over the lexical bundles which are commonly used in their discipline. In this regard, there is a need for studies identifying the frequent lexical bundles in a particular discipline and examining non-native writers' texts to see to what extent they approximate native writers in terms of lexical bundle use. The primary aim of this study was to investigate Turkish and native English postgraduate students' and native scholars' use of lexical bundles in a specific academic discipline, that is foreign language teaching, in terms of frequency, functions and structures of bundles. For this aim, a corpus of 150 texts was collected containing Turkish and native English students' MA and PhD theses along with native scholars' published research articles. Four-word combinations occurring 25 times per million words and appearing in at least 5 texts or more were identified as lexical bundles. WordSmith Tools 6 was used to retrieve the lexical bundles in the research corpus. The results revealed that Turkish postgraduate students used far more lexical bundles in their texts compared to both native students and scholars. However, there was a redundancy in Turkish students' texts when the token frequencies were examined, meaning that Turkish students overused most of the lexical bundles. On the other hand, statistical analysis of the bundle lists revealed that Turkish postgraduate students employed different bundles from their native peers and scholars. Finally, the structural and functional categories did not show any statistically significant differences. Keywords: Lexical bundles, English academic discourse, Turkish postgraduate students, native English postgraduate student, native scholars