The Interaction of the Bilingual’s Two Phonetic Systems: Differences in Early and Late Korean-English Bilinguals

One of the basic questions of bilingual research is to what extent the bilingual’s two phonetic systems influence each other, a question that has occupied a prominent place in bilingual research almost from the outset of the field (see, for example, Weinreich, 1953). Recent studies in bilingual first language acquisition (BFLA) demonstrate that, at least in simultaneous bilingual acquisition, infants exposed to two languages are developing what appear to be two phonetic/phonological systems from the beginning (Johnson and Wilson, 2002; Kehoe, 2002; Bosch and Sebastian-Galles, 2001, Vihman, 2002), although their two systems may differ from monolinguals of either language (Mack, 1989; Bosch, Costa, & Sebastian-Galles, 2000). In late bilingualism (i.e., when a second language is learned in adulthood), where one language system is already established when a second is learned, most likely a separate system for the second language does not develop in the beginning. In fact, most researchers agree, at least as initial stages of L2 learning, adults perceive and produce both native (L1) and second language (L2) vowels and consonants (“sounds” for short) as instances of similar L1 sounds. In fact, it is only after extensive exposure to the second language that separate long-term memory representations (categories) for L2 sounds are formed, if at all (Flege, Meador, & MacKay, 1999). The differences between these two types of bilingualism indicate that several factors most likely influence both the extent the two languages influence each other and the direction of the influence (Flege, 1995; Piske, Flege, McKay, & Meador, 2002; Khattab, 2002). The purpose of this study is to examine how two factors, the amount of similarity between native (L1) and second-language (L2) sounds, and age at the time of second language learning, determine how the L1 and L2 phonetic systems interact both at initial and more advanced stages of L2 learning. The first factor, the amount of similarity between L1 and L2 sounds (or cross-language similarity), at least in late L2 learning, heavily determines how much the L1 influences the L2 (Flege, MacKay, and Munro, 1999; Best, 1995; Flege, 1995). In particular, the more similar L2 sounds are to L1 sounds, the more likely the L1 will influence how L2 sounds are perceived and produced. For example, Flege, Bohn, & Jang (1997) showed that learners of English perceived and produced English vowels differently depending on their native language. In addition, Aoyama, Flege, Guion, Yamada, & Akhane-Yamada (2003) have determined that Japanese learners of English identify English /r/ as being more similar to Japanese // than this Japanese consonant is to English /l/. These same Japanese learners are more likely to produce English /r/ more accurately than English /l/. These and many other studies indicate that cross-language similarity is an important factor in how much the L1 and L2 interact in bilinguals (Best, 1995; Guion, Flege, Akhane-Yamada, & Pruitt, 2000). The second factor that may influence the organization of a bilinguals’ L1 and L2 phonetic systems is the learner’s age at the time of learning a second language. In particular, in early bilinguals (i.e., those that learn a second langauge in childhood), the native language seems to exert less of an influence on the perception and production of the L2, suggesting that cross-language similarity may play less of a role in early than in late bilinguals (Baker, Trofimovich, Mack, and Flege, 2001). . For example, Flege, MacKay and Meador (1999) found that early Italian-English bilinguals were more accurate in their perception and production of English vowels than were late Italian-English bilinguals,