Frequency Change Patterns across Proficiency Levels in Japanese EFL Learner Speech

This study investigated the overall patterns of variation across seven oral proficiency levels of 1,263 Japanese EFL learners and native English speakers. The methodological approach combined a learner corpus, language processing techniques, and multivariate statistical analyses to identify patterns of language use. The largest spoken learner corpus in Japan, the National Institute of Information and Communications Technology Japanese Learner English (NICT JLE) Corpus was used for the analysis. This corpus consists of transcribed interviews with 1,281 learners and contains over one million running words of spoken English. The transcriptions were compared with data gathered from 20 native English speakers who performed identical speaking tasks. In the present study, 58 linguistic features (e.g., grammatical features) were used from the original list of 67 linguistic features in Biber’s (1988) study. The following research questions were addressed. First, what linguistic features do and do not vary among Japanese EFL learners at different oral proficiency levels and native English speakers? Second, is computer-aided analysis of multiple linguistic features useful for determining which ones characterize particular oral proficiency groups? This study found interesting rising, flat, and falling frequency patterns in how several linguistic features are used in different oral proficiency levels. Some linguistic features (e.g., phrasal coordination and nouns) were frequently used by the learners at a low level. The frequencies of some others (e.g., contraction, pronoun it, and emphatics) increased as oral proficiency increased. The study identified a set of linguistic features that differentiate among second language oral proficiency groups as well as between non-native and native speakers of English