International Journal of Learner Corpus Research - Volume 1, Issue 2, 2015

This study evaluates the impact of writing proficiency on native language identification (NLI), a topic that has important implications for the generalizability of NLI models and detection-based arguments for cross-linguistic influence (Jarvis 2010, 2012; CLI). The study uses multinomial logistic regression to classify the first language (L1) group membership of essays at two proficiency levels based on systematic lexical and phrasal choices made by members of five L1 groups. The results indicate that lower proficiency essays are significantly easier to classify than higher proficiency essays, suggesting that lower proficiency writers make lexical and phrasal choices that are more similar to other lower proficiency writers that share an L1 than higher proficiency writers that share an L1. A close analysis of the findings also indicates that the relationship between NLI accuracy and proficiency differed across L1 groups.

This paper studies longitudinal changes that occur in academic learner Finnish. The approach is data-driven, and it shares the usage-based view of language. The data are part of the Corpus of Advanced Learner Finnish and the studied features are determined statistically by tracing the morphological forms that show the clearest changes in their frequency during the observation period of 16 months. Features found to depict such changes are then studied in greater detail to find out the constructional nature and possible reason for the observed change. The results show that the use of the preterite tense constructions decreases while the present tense constructions increases and that the change is not due to any single lexical unit. The use also becomes quantitatively more native-like. The change is likely due to emergence of a new linguistic register — academic Finnish — and the results thus support the usage-based theory of language acquisition.

When assessing a second language speaker’s nativelikeness or accentedness, researchers often employ holistic judgments (Abrahamsson & Hyltenstam 2009, Ioup et al. 1994) or auditory analysis of specific segments (Rampton 2013). Acoustic analysis, which can help quantify minute details, can be quite time-consuming when large corpora are involved. This research note describes the Accents of Non-Native English (ANNE) learner corpus which employs the open-source Language Brain and Behaviour-Corpus Annotation Tool (LaBB-CAT; Fromont & Hay 2012) that allows researchers to automatically extract timing information about segments in the corpus and process them with Praat (Boersma & Weenink 2009), facilitating large-scale acoustic analysis.