In this study I analyze the effect of allo-repetition (repetition of another’s utterance (Tannen 1987)) on the use of giving verbs by an English-Japanese bilingual child. Previous research on child language acquisition (Moerk & Moerk 1979, Ochs & Schieffelin 1984) and the acquisition of Japanese giving-receiving verbs (Uyeno et al. 1978, Ishiguro 1985, Ogihara 2005) suggests that exposure to more uses of linguistic forms improves acquisition. For example, Ogihara (2005) demonstrated that a bilingual child acquired serial giving verbs (e.g., V-te ageru) before single giving verbs (ageru) because caretakers used serial verbs more frequently. However, an examination of the use of giving verbs in actual conversational interaction shows that exposure to these forms may not always have a positive influence on their acquisition.
The data for this study come from an 80-hour videotaped corpus of conversations between an English-Japanese bilingual child and Japanese children/caretakers. I analyze how repetition influenced Sandy’s (3;10) use and understanding of the verbs kasu ‘lend/give (it to me)’ and kaesu ‘return/give (it) back (to me)’ in a conversation involving a conflict over a toy with her Japanese cousin Mine (4;7), which contained 18 uses of these verbs. Following Iwasaki (2002:94), I analyze these verbs as giving verbs because they have [NP1 ga NP2 ni NP3 o Pred] argument structure. Although the data suggest that Sandy had acquired the verb kasu at an earlier age (2;11) and she pronounced the verb kasite with the correct LHH accent initially, she modified her pronunciation to be progressively more like Mine’s pronunciation of kaesite. Specifically, Sandy altered the [a] in kasite to a diphthong [ai], and as tension increased she changed the accent to HLLL. I conclude that cross-linguistic differences and situation specific factors including participants’ needs and desires in the interaction may be more important than simple frequency counts.