A multidimensional analysis of L1–L2 differences across three advanced levels. Issue 2 (25th June 2018)

• Record Type: Journal Article
• Authors: Pan, Fan
• Is Part Of: Southern African linguistics and applied language studies. Volume 36:Issue 2(2018)
• Page Start: 117
• Page End: 131
• Publication Date: 2018-06-25
• Subjects: Linguistics -- Periodicals; Linguistics -- Africa, Southern -- Periodicals; African languages -- Periodicals; Afrikaans language -- Periodicals; South Africa -- Languages -- Periodicals; Linguistique -- Périodiques; Linguistique -- Afrique australe -- Périodiques; Langues africaines -- Périodiques; Afrikaans (Langue) -- Périodiques; Afrique du Sud -- Langues -- Périodiques; African languages; Afrikaans language; Language and languages; Linguistics; Africa, Southern; South Africa; Periodicals; 410.968
• DOI: 10.2989/16073614.2018.1476162
• Languages: English
• ISSNs: 1607-3614

Abstract

Abstract: A new multidimensional analysis exploring linguistic variation in L1 and L2 academic writing across three advanced proficiency levels (Master's theses, PhD dissertations, and research articles) is conducted. Through an exploratory factor analysis, I identify and interpret four dimensions that capture the lexical and grammatical differences between L1 and L2 academic writing in applied linguistics. Two-way ANOVAs show that there are significant differences between L1 and L2 academic writing on three of the four dimensions. Compared with L1 academic writing, L2 academic writing is generally less attitudinal, narrative, and academically involved. Among the three proficiency levels, L2 research articles are closest to L1 research articles, while L2 Master's theses differ most from L1 Master's theses on three dimensions. ANOVA tests also show that proficiency levels have a main effect on Dimension 2, as well as having an interactive effect with language backgrounds on Dimensions 2 and 3. Our findings have important pedagogical and methodological implications. First, L2 writers need to increase their awareness of the importance of stance expressions in academic writing. Second, proficiency levels need to be given considerable attention in corpus design to minimise their influence on the results of L1–L2 comparisons.