The present study was designed to test the validity of the L1-L2 Interaction Hypothesis within the theoretical framework of Flege's (1995, 2002) Speech Learning Model (SLM), which claims that L1 and L2 subsystems are in a common phonological space, in...
The present study was designed to test the validity of the L1-L2 Interaction Hypothesis within the theoretical framework of Flege's (1995, 2002) Speech Learning Model (SLM), which claims that L1 and L2 subsystems are in a common phonological space, in order to better understand adults’ second language speech learning. For this purpose, the perception and production of Korean vowels (monophthongs) by English and Japanese learners of Korean were examined.
To verify the SLM’s hypotheses, assessing the phonetic similarity is required prior to the perception and production tests. In Experiment 1 and 2, two experiments were manipulated: (1) the acoustic analysis of vowel formants of Korean, English, and Japanese, and (2) the perceived similarity rating test. In Exp. 1, the productions of 8 Korean vowels /i, e, ɛ, a, ʌ, o, u, ɨ/, 11 English vowels /i, ɪ, e, ɛ, ӕ, a, ɔ, o, ʊ, u, ʌ/, and 5 Japanese vowels /i, e, a, o, ɯ/ by Korean, English, and Japanese monolinguals were compared by using IPA and the formant quality. Results revealed that for native speakers of English, Korean /i/ is identical, /e/ɛ, a, ʌ, o, u/ are similar, and /ɨ/ is new to English vowel systems whereas for native Japanese speakers, Korean /E, a/ are identical, /i, o, u, ɨ/ are similar, and /ʌ/ is new to Japanese vowel systems. In Exp. 2, the experienced groups (American- & Japanese-Koreans) and inexperienced groups (near-monolingual English and Japanese speakers) participated in a cross-language mapping experiment in which they identified Korean vowels in terms of their L1 categories, then rated the identifications for goodness scores to the English and Japanese category, respectively. The findings are that the degree of L1 influences appeared to depend on the degree of acoustic similarity between L1 and L2 vowels and the length of their exposure to the L2. The results suggest that there is a surprising degree of uniformity in the ways that subjects with different language backgrounds perceive L2 sounds. In addition, /ʌ/ for English speakers and /i/ for Japanese speakers were redefined as new sounds in terms of assimilation patterns and goodness rating scores.
In Experiment 3 (discrimination test), English and Japanese learners of Korean were assigned into the inexperienced (EI & JI) and experienced group(EE & JE) based on their length of residence in Korea (0.6 vs. 5 yrs.), and a native Korean (NK) control group participated in discriminating L2 contrasting vowels. The results demonstrated that Japanese subjects with lip rounding as a cue in their L1 were more accurate at recognizing Korean /o-u/ pairs than were English subjects, which suggests that L1 feature weighting might result in different discrimination scores (A'). In most cases, all groups have great difficulty discriminating similar sounds to their L1s because instances of two contrastive Korean vowels tend to be perceptually assimilated by a single L1 vowel category; however, the experienced groups performed better for the new sounds.
In Experiment 4 (acoustic analysis for L2), nonsense words containing 8 Korean /i, e, ɛ, a, ʌ, o, u, ɨ/ were elicited. Acoustic analysis revealed EE and JE produced more accurately than EI and JI for the new sounds and did not statistically differ from NK. Moreover, the Euclidean distances of contrasting Korean vowel pairs containing new vowels produced by the experienced groups were even closer to NK than were the inexperienced group. The results indicated that bilinguals' L1 profoundly influenced the production accuracy.
In Experiment 5 (Intelligibility), EE and JE were found to discriminate Korean vowels more accurately than EI and JI. New vowels produced by EE and JE were heard as intended significantly more often than vowels produced by EI and JI.
In Experiment 6 (L1-L2 interaction), the objective of this experiment was to determine how bilinguals' L1 vowel system influenced the organization of their L2 system and vice versa. SLM proposes two specific mechanisms through which the phonetic categories making up the L1 and L2 phonetic subsystems will interact in a common phonological space: category assimilation and category dissimilation. Category assimilation operates in case of category formation for an L2 sound has been blocked. As a result, a merged category of L1 and L2 is developed for the L2 sound which is similar to L1. On the other hand, category dissimilation is thought to function when a new category has established for a sound that is found in the L2 but not in the L1. It is conceivable that the bilinguals' combined L1-L2 phonetic spaces tend to disperse to maintain phonetic contrast. In either case, the phonetic category formation is crucial and relates to the phonetic similarity in the L1 and L2. The results demonstrated that category assimilation was confirmed. That is, bilinguals produced merged acoustic realizations of L1 and L2 vowels for the similar sounds so that L2 learners used L1 norms for similar L2 sounds the value of formant(△Bark) were intermediate to formant values of L1 and L2. However, the result showed no evidence for phonetic category dissimilation, which means it did not cause a newly established L2 category and the nearest L1 speech category to shift away from one another in a phonetic space.
Taken together, these results suggest that L2 experience influenced the perception of the L2 vowels differently and the perceptual difference has an influence on the production of the Korean vowels; this was consistent with Flege's SLM. Specifically, although phonetic category formation is more successful for new L2 sounds than similar sounds in L1, category dissimilation was not found in the present study. However, the most important finding of this study was that the L1-L2 phonetic subsystems of a bilingual interact in a common phonological space so that a merged phonetic category is placed between L1 and L2 phonetic categories produced by monolinguals in each language when the category formation for L2 is blocked.
，韩语论文，韩语毕业论文 |
