Implicit and Explicit Language Learning: Conditions, Processes, and Knowledge in SLA and Bilingualism

13. Explicit Training and Implicit Learning of L2 Phonemic Contrasts

13 Explicit Training and Implicit Learning of L2 Phonemic Contrasts FRED R. ECKMAN University of Wisconsin–Milwaukee GREGORY K. IVERSON University of Wisconsin–Milwaukee and University of Maryland Center for Advanced Study of Language ROBERT ALLEN FOX AND EWA JACEWICZ The Ohio State University SUE ANN LEE Texas Tech University THE PURPOSE OF THIS CHAPTER is to report preliminary findings of an ongoing investigation into constraints on the acquisition of L2 phonemic contrasts. We elicited production and perception data in two of the three logically possible ways in which a NL and a TL can differ with respect to a two-way phonemic contrast, as listed in (1). (1) NL–TL Differences in a Two-Way Phonemic Contrast (a) The NL lacks sounds corresponding to either of the two TL phonemes. (b) The NL has sounds corresponding to one, but not both, of the two TL phonemes. (c) The NL has sounds corresponding to both of the TL phonemes, but in complementary distribution as allophones of the same phoneme. This chapter considers only the latter two language-contact situations, those depicted in (1b) and (1c), and reports on the elicitation of both production and perception data to investigate these two NL–TL combinations. A language-contact situation that illustrates (1b) as well as (1c) arises with respect to Korean as the NL and English as the TL. Thus, exemplifying (1b), Korean has [p], as does English, but Korean lacks [f] altogether, a sound that stands in phonemic contrast to /p/ in English. At the same time, both Korean and English have [s] and [š], but whereas these sounds contrast in English, they are in complementary distribution in Korean, because [š] occurs only before a (phonological) high front vowel 159 or glide, and [s] occurs elsewhere. In Korean these two sounds are related by a principle or rule such as that in (2). (2) Korean Allophonic Rule /s/ is realized as [š] before the high front vowel or glide, elsewhere as [s]. The task of a Korean learner of English in acquiring these two contrasts seems straightforward: on the one hand, the learner must acquire the phoneme /f/ to differentiate words such as pan and fan, and, on the other hand, the learner must suppress the application of the NL allophonic rule so as not to render see the same as she. If the early stages of the interlanguage (IL) grammar are tied closely to the NL phonological patterns, then the learner will err on TL words containing /f/, most likely substituting /p/ (fan ⫽ pan), the phonetically closest segment in the NL inventory, and the learner will also transfer (2) into the IL grammar, erring on TL words containing [s] before a high front vowel (see ⫽ she). However, general principles of phonology, to be discussed below, constrain the application of (2) in the IL and thus restrict the errors that the second-language learner makes, depending on whether the learner is at a stage in which [š] represents the phoneme /š/ rather than /s/ in the IL, at least for some words. Through the hypotheses developed in detail below, we predict that L2 learners who acquire a contrast such as that in (1c) in what we term morphologically derived environments will necessarily generalize that contrast to morphologically basic environments but not vice versa. Employing these general phonological principles as the basis for an intervention strategy, we attempted to manipulate the learning and generalization of the /s/–/š/ and /p/–/f/ contrasts. The remainder of the chapter is structured as follows. The next section outlines the background for the study, setting the theoretical basis for the work and connecting it to previous research in the area. The section concludes with the statement and rationale of the hypotheses. We then lay out the methodology used to elicit the data. The two sections following deal with the findings, reporting them in the results section and interpreting them in the discussion section. The final section concludes the chapter. Background This section sets the context for the study by reviewing the literature in the two areas where the findings impinge on previous work. The discussion is followed by a description of the theoretical grounding for, and the statement of, the hypotheses. The results of this research can be seen as a contribution to the discussion of two areas of L2 phonology, the first a long-standing issue dating back to the times of the Contrastive Analysis Hypothesis (CAH) (Lado 1957), the second a more-recent question...