Variability and systematicity in individual learners’ Japanese lexical accent

This paper presents novel experimental production data to establish generalisations about the accent patterns produced by individual English learners of Japanese. 21 British English learners read aloud tokens varying in word type (two- and three-mora nouns and verbs) and speech environment (in isolation and preceding a function word) and native speakers identified the accent types produced. The results show both considerable between-learner variability and within-learner systematicity. The accent types produced are seen to vary with word type and speech environment, but both how they vary and which accent types are produced are shown to be individual to the learner. It is suggested that this combination of between-learner variability and within-learner systematicity may be the result of a difficulty in phonologically encoding Japanese lexical accent.