Assessing conventionalized language in English learner essays by applying a method of "warming up" in Swedish L1

Abstract: The aim of this study is to look at the use of formulaic language, i.e. memorized and conventionalized combinations of words, in essays written by Swedish intermediate level students of English. Drawing on previous research (Cohen and Brooks -Carson 2001) this study will apply a method of “warming up” in Swedish (L1) before writing in English (L2). The primary material includes thirty essays written directly in English, and thirty essays starting from an outline in Swedish and then written in English, by the same students but on a different topic. Since there is evidence that native speakers always score the highest on amount of formulaic language in written discourse, the main hypothesis of this study is that by starting from an outline in Swedish and then writing in English the students will make use of more formulaic language than when writing directly in English. The first research question involves the quantity of formulaic language in the students’ essays. The second question concerns the distribution and use of the different categories of FSs (NPs, VPs etc) in the direct and indirect modes of writing. The last question addresses the erroneous attempts made by the students in forming FSs and the possible influence of transfer from L1 in the essays starting from an outline in Swedish compared to the essays written directly in English. The findings of this study lend support to the main hypothesis. Moreover, the analysis shows that there are significant differences in the distribution and use of the different categories of formulaic language in the direct and indirect modes of writing. Results are suggestive of a more conscious and less uncertain approach in the essays starting from an outline in Swedish, however the findings may be circumstantial. Furthermore, the results indicate that the method of “warming up” in Swedish may encourage the use of combinations of words that are more conventionalized and do not always interact with syntax in a regular way. The findings also indicate that transfer is not more salient in the essays starting from an outline in Swedish if compared to the essays written directly in English, at least not in relation to the erroneous FSs found in the students’ essays.