A Comparative Study On Four-word Lexical Bundles In Applied Linguistics’ Academic Writing

Recently multi-word units have received wide attention from linguistics home and abroad. Piles of studies indicate that language is composed of multi-word units and the acquisition of multi-word units help second language learners reach fluent and native-like level. A type of multi-word units called lexical bundles are frequently recurrent and formally fixed in given registers. Many a corpus research show that lexical bundles are important building blocks in discourse which can differentiate registers(Biber et al., 1999; Biber & Barberia, 2007). Overseas studies have paid attention to lexical bundles used in academic texts, including the use of them in various disciplinary academic writing and the comparison of lexical bundle use among native-speaker students, academic experts and second language learners.However, domestic research mainly focuses on the use of lexical bundle in timed compositions written by undergraduates. There is a lack of studies on advanced learners’ use of lexical bundle in academic texts.The current study aims to compare four-word lexical bundles in Chinese English major postgraduates’ master theses of applied linguistics and that in international core journal articles of the same academic discipline. This study complies two corpora which are RA corpus(experts’ research articles) and MA corpus(Chinese learners’ Master of Arts theses). The research first examines the abundance and frequency distribution of lexical bundles in two corpora as well as the shared bundles and bundles not shared by learners and experts. Afterwards, the comparison of the two lists of lexical bundles is made in terms of their structures and functions. The followings are research results:The lexical bundles abiding by the preset criteria in RA corpus are 16% more than that in MA corpus. Among them, 20.21%(38 types) of the lexical bundles are shared in the two corpora while 81.91%(154 types) are not shared. There is an obviously inadequate use of hedges and passive voice constructions in learners’ data.On the other hand, lexical bundles explicitly used by learners embody a style of spoken register such as that is to say.According to Biber et al.(1991)’ s structural taxonomy, there are a total of 16 structural types concluded in the overall lexical bundles which can be categorized into5 major types. Both groups of academic writers use NP-based bundles as the largest structural category, orderly followed by PP-based bundles, VP-based bundles,Clause-based bundles and Others. But the comparison of sub-categories implies difference between the two groups. Learners have a less use of “NP with other post-modifier fragment”, “Other NP expressions” and some fixed PP frames such as“in/on/at + the + Noun of”. On the other hand, they use some structural patterns more specific to spoken register, i.e., “Personal pronoun + lexical VP fragment” and “VP with active verb” which do not exit in expert writings.According to Hyland(2008a)’ s functional framework, there is no significant difference between the general distribution of functional bundles used by learners and experts. The biggest functional categories are research-oriented and text-oriented bundles which account for nearly 90%. Participant-oriented bundles only take up a small part of 12% in both corpora. In contrast to expert writers, learners use more location bundles especially time-referent clusters, resultaive signals and stance bundles. On the other hand, learners use less descriptive bundles, transitional signals,structuring signals and engagement bundles. The reasons leading to these distinctions may be the following:(1) there are many overlapping bundles in certain categories excessively used by a group of writers;(2) the overuse or underuse of certain structural types may cause the corresponding usage of certain functions realized by the structures;(3) learners overuse some types of functional bundles or avoid employing other categories under the influence of cultural and instructional factors.In conclusion, the present study shows that Chinese English major postgraduates differ from experts in abundance and frequency of lexical bundles in applied linguistics academic writing and that they embody difference from experts in several structures and functions. Learners should pay special attention to the characteristic use of lexical bundles in norm academic texts. Academic writing teaching should also attach great importance to explicit instruction on lexical bundles.