Gendered L1 attrition and L2 acquisition of pitch range in Japanese-English sequential bilinguals

Abstract

This project investigated pitch range production in Japanese-English female and male sequential bilinguals. This language combination was chosen because high pitch level has been claimed to index femininity in Japanese (Ohara, 2019) whereas in English an increase in pitch level is used to index friendliness by both females and males (Loveday, 1981). Data were collected through a reading task and voicemail task from 19 Japanese-English bilinguals in London (UK), 21 Japanese-English bilinguals in Tokyo (JP), 15 Japanese monolinguals in Tokyo and 16 English monolinguals in London (49 females vs 21 males). In both tasks, speech was directed to an imaginary addressee varying in formality and sex. Of interest was to examine the extent to which Japanese and English influenced one another with regard to L1 attrition and L2 acquisition. The effect of individual gender identity on pitch range was investigated to assess whether, e.g., female bilinguals who closely identified with femininity would produce a high pitch level. In the reading task, Japanese female and male monolinguals produced significantly higher f0 maximum and wider pitch span than English monolinguals, irrespective of addressee. Regarding L1 attrition, bilingual males produced a significantly lower f0min in their Japanese than the Japanese monolingual males, suggesting a restructuring of the L1. Regarding L2 acquisition, English f0 mean and f0 maximum of the female bilinguals was significantly higher than that of the English females, suggesting an influence from Japanese on English. Additionally, English f0 mean was lower for both female bilinguals who rated themselves as more masculine and male bilinguals who rated themselves as more feminine on the English gender questionnaire. These results were not replicated in semi-spontaneous speech. Summarising, gender-specific patterns of L1 attrition and L2 acquisition were evidenced with regard to read speech, but not semi-spontaneous speech. This suggests that the formality of read speech might enhance the production of language and gender normative pitch range and that individual gender identity might have been expressed alternatively in semi-spontaneous speech.