Abstract
This thesis aims to identify some of the challenges Norwegian learners of English encounter when writing English for academic purposes. To this end, the linguistics and literature section of the Norwegian subcorpus of the Varieties of English for Specific Purposes dAtabase (VESPA) is investigated within the phraseological framework of Douglas Biber s Lexical Bundles methodology. This approach to phraseology aims to identify the most frequent recurrent expressions in a corpus, regardless of their idiomaticity or structural status, in order to identify the most productive phrases of the sample. These bundles are then sorted according to their function in discourse. The process is repeated for a similarly composed section of the British Academic Written English corpus, and the two sets of bundles are compared, with the BAWE sample as a native control group, focusing on three key areas: What bundles are used? What functions to they serve in discourse? Are any bundles over- or underused in the Norwegian sample? This analysis reveals that Norwegian university-level learners are in fact using lexical bundles in a largely native-like manner in terms of overall frequency and function, but also that certain areas continue to be problematic, owed mainly to their comparatively limited repertoir of bundles that express the writer s stance or contribute to the explicit ordering of discourse.