Kimiko Tsukada

Degree of foreign accent in English sentences produced by Korean children and adults

The purpose of this longitudinal study was to evaluate the influence of age (adult vs. child) and length of residence (LOR) in an L2-speaking country (3 vs. 5 years) on degree of foreign accent in a second language (L2). Korean adults and children living in North America, and age-matched groups of native English (NE) adults and children, recorded English sentences in sessions held 1.2 years apart (T1 vs. T2). NE-speaking listeners rated the sentences for overall degree of perceived foreign accent using a 9-point scale. The native Korean (NK) children received significantly higher ratings than the NK adults did, but lower ratings than the NE children. The NK children—even those who had arrived as young children and been enrolled in English-medium schools for an average of 4 years—spoke English with detectable foreign accents. The effects of LOR and the T1–T2 differences were non-significant for both the NK adults and the NK children. The findings were inconsistent with the hypothesis that adult–child differences in L2 speech learning are due to the passing of a critical period. The suggestion is made that the milder foreign accents observed for children than adults are due, at least in part, to the greater L2 input typically received by immigrant children than adults.

A developmental study of English vowel production and perception by native Korean adults and children

Abstract This study examined the production and perception of English vowels by native Korean (NK) learners of English on two occasions separated by about 1 year. A preliminary experiment revealed that NK adults classified some pairs of contrastive English vowels using two different Korean vowels whereas other pairs showed classification overlap, implying they would be difficult for Korean learners of English to discriminate. In two subsequent experiments, NK adults and children differing in length of residence in North America (3 vs. 5 years; 4 groups of 18 each) were compared to age-matched native English (NE) speakers. In Experiment 2, NK children were found to discriminate English vowels more accurately than NK adults but less accurately than NE children. In Experiment 3, English words containing /i ɪ e ɪ e ae ɑ ʌ/ were elicited using a picture-naming task. Some vowels produced by NK children were heard as intended significantly more often than vowels produced by NK adults. Acoustic analyses revealed that NK children produced significantly larger between-vowel contrasts than NK adults but did not differ from NE children. Taken together, the results suggested that although children are more successful than adults in learning the phonetic properties of second-language vowels, they might continue to differ from age-matched native speakers in certain respects as uncovered by the vowel discrimination test in Experiment 2.

Release Bursts in English Word-Final Voiceless Stops Produced by Native English and Korean Adults and Children

The aim of this study was to evaluate the acquisition of statistical properties of a second language (L2). Stop consonants are permitted in word-final position in both English and Korean, but they are variably released in English and invariably unreleased in Korean. Native Korean (K) adults and children living in North America and age-matched native English (E) speakers repeated English words ending in released tokens of /t/ and /k/ at two times separated by 1.2 years. The judgments of E-speaking listeners were used to determine if the stimuli were repeated with audible release bursts. Experiments 1 and 2 revealed fewer final releases for K than E adults, and fewer releases for /t/ (but not /k/) for K than E children. Nearly all /t/ and /k/ tokens were heard as intended in experiment 3, which evaluated intelligibility. However, the K adults’ /k/ tokens were identified with less certainty than the E adults’. Taken together, the results suggested that noncontrastive (i.e., statistical) properties of an L2 can be learned by children, and to a somewhat lesser extent by adults.

Effects of speaking rate and vowel length on formant frequency displacement in Japanese.

This study examined effects of phonemic vowel length and speaking rate, two factors that affect vowel duration, on the first and second formants of all vowels in Japanese. The aim was to delineate the aspects of formant displacement that are governed by the physiological proclivity of vowel production shared across languages, and the aspects that reveal language-specific phenomena. Acoustic analysis revealed that the phonemic long vowels occupied a more peripheral portion of the F1 × F2 vowel space than the phonemic short vowels (effect of vowel length), but effects of speaking rate were less clear. This was because of the significant interactions of the two effects: the formants of phonemic short vowels were more affected by speaking rates than the phonemic long vowels. Regression analyses between F2 and duration revealed that formant displacement occurs when vowels are less than 200 ms. Similarities and differences found for Japanese and English are discussed in terms of physiological proclivity of vowel production versus language-specific phonological encoding.

Discrimination of English and Thai words ending with voiceless stops by native Thai listeners differing in English experience

This study examines the discrimination of words ending with voiceless stops / p t k / in first language (L1) and second language (L2) by three groups of native Thai participants. These participants differed in their countries of residence and experience with L2 English in the formal education system. The first group (T1) was a group of 18 Thai listeners who were living in Australia. The second and third groups consisted of 12 university students (T2) and 12 high-school students (T3) living in Thailand. Eighteen Australian English (AusE) listeners were included as controls. English and Thai words minimally contrasting in the final stop (e.g. ‘cap’ vs. ‘cat’) were presented to the listeners to investigate whether L2 phonetic learning occurs even for the contrasts that are expected to be discriminated with high accuracy from the onset of L2 learning and if so, how it is influenced by L2 experience. All three Thai groups showed reasonably accurate discrimination for both English and Thai words, but only T1 showed discrimination accuracy comparable to AusE in English. Further, only T3 was clearly more accurate in discriminating unreleased Thai than English stop contrasts, most of which were accompanied with release bursts. These two findings are taken to be evidence for phonetic learning of specific aspects of L2 contrasts rather than positive L1 transfer.

Durational Characteristics of English Vowels Produced by Japanese and Thai Second Language (L2) Learners

This empirical study investigated durational characteristics of English vowels /i/ as in ‘heat’ and /I/ as in ‘hit’ produced by two groups of second language (L2) learners who differ in their first language (L1) backgrounds, i.e. Japanese and Thai. In these languages, vowel length is phonemic. Experiment 1 compared the /i/-/I/ contrast produced in monosyllabic words ending with /p t k/ by Australian English (OZ1, n=6) and Thai speakers (T, n=22). Experiment 2 compared the same vowel contrast in monosyllabic words ending with /t/ and /d/ by another group of Australian English speakers (OZ2, n = 12) and Japanese speakers (J, n =24). While both non-native groups showed a greater durational differentiation of the /i/-/I/ contrast than the OZ groups to which they were compared, J and OZ2 differed to a greater extent when the /i/-/I/ contrast occurred before /t/ than before /d/. This finding highlights the importance of taking language-specific phonetic factors such as allophonic variation into consideration in L2 speech acquisition research.

Cross‐language perception of word‐final stops: A comparison of Cantonese, Japanese, Korean, and Vietnamese listeners

Four groups of Asian language speakers [Cantonese (NC), Japanese (NJ), Korean (NK), Vietnamese (NV)] were compared for their accuracy to discriminate English and Thai word‐final stop contrasts (/p‐t/, /p‐k/, /t‐k/). Word‐final stops can be either audibly released or unreleased in English, whereas they are always unreleased in Thai. English (but not Thai) was familiar to all participants. In Cantonese, Korean, and Vietnamese, word‐final stops are unreleased as in Thai. Japanese, on the other hand, does not permit word‐final stops. All groups were more accurate in discriminating English than Thai contrasts. However, their response patterns differed according to their first languages (L1). NK was almost as accurate in Thai as in English while NJ was much better in English than in Thai. NC and NV were intermediate between NK and NJ. These findings suggest that experience with specific (i.e., unreleased) and native phonetic realization of sounds may be essential for accurate discrimination of word‐final stop contrasts. However, in order to explain the differences between NC, NK, and NV in their discrimination of unreleased Thai stops, it is necessary to examine the robustness of phonetic cues to word‐final stops in their L1s. The role of L1 experience in cross‐language perception is discussed.

The perception of Mandarin lexical tones by listeners from different linguistic backgrounds

Abstract Two groups of non-native adult learners of Mandarin in Australia were directly compared in their ability to perceive monosyllabic Mandarin words contrasting in lexical tones. They differed in their linguistic experience (non-heritage (n=10), heritage (n=12)). A group of eight native Mandarin speakers and a group of ten functionally monolingual speakers of Australian English were included as controls. All non-native learners used English as their primary language of communication. However, the heritage learners were able to communicate in Cantonese as well as English. The primary question of interest was whether heritage learners’ knowledge of contrastive tone in Cantonese might give them an advantage over English-speaking learners in perceiving tone contrasts in Mandarin. In general, there were more similarities than differences between the two groups of learners in their response patterns. Of the six tone contrasts examined (T1-T2, T1-T3, T1-T4, T2-T3, T2-T4, T3-T4), the two groups significantly differed only on T1-T4. The heritage learners were less accurate on T1-T4 than the non-heritage learners who are monolingual speakers of Australian English. On the other hand, the non-heritage learners were more accurate than Australian English speakers with no prior experience with Mandarin on all tone contrasts. Thus, we conclude that simply having an exposure to and functional knowledge of another tonal language since early childhood does not guarantee accurate perception of Mandarin tones in comparison with adult learners without prior experience with tonal languages.

Vowel length categorization in Arabic and Japanese: Comparison of native and non-native Japanese perception

Abstract This study examined the perception of short and long vowels in Arabic and Japanese by native (NJ) and non-native (NNJ) speakers of Japanese. Both Arabic and Japanese use duration contrastively. The aim was to examine whether and how the listeners’ differential knowledge of Japanese might influence the processing of short and long vowels in an unknown language, Arabic, as well as Japanese. Of particular interest was the effect of NNJs emerging length categories resulting from second/foreign language learning experience on their cross-language perception. Ten Japanese vowels (five short, five long) were used as response categories. On average, the NJ listeners categorized the Japanese vowels more accurately than did the NNJ listeners (91 vs. 84%), in particular, long vowels (94% for NJ and 82% for NNJ), but the between-group difference did not reach statistical significance. However, qualitative differences were observed between the two groups in the way they responded to Arabic and Japanese vowels. This may be related to the distributional characteristics of Japanese vowels, which is that long vowels are much less frequent than short vowels, prompting the NNJ, but not the NJ, listeners to select the short category over the long counterpart when in doubt. As a result of this distributional imbalance, the NNJ listeners may have developed a more robust representation for short than for long vowels in Japanese. Both NJ and NNJ listeners assimilated Arabic vowels to a wide range of Japanese vowels, suggesting perceptual confusion.

Non-Native Japanese Listeners' Perception of Vowel Length Contrasts in Japanese and Modern Standard Arabic (MSA).

This study aimed to compare the perception of short vs. long vowel contrasts in Japanese and Modern Standard Arabic (MSA) by four groups of listeners differing in their linguistic backgrounds: native Arabic (NA), native Japanese (NJ), non-native Japanese (NNJ) and Australian English (OZ) speakers. The NNJ and OZ groups shared the first language (L1), but differed in their familiarity with Japanese. In both Japanese and MSA, vowel length is phonemic. In contrast, vowel duration plays a more limited (although not insignificant) role in English. Of interest was the discrimination accuracy of NNJ listeners who learned Japanese as a second (L2) or foreign language in adulthood. As expected, the NA and NJ groups discriminated their L1 contrasts more accurately than all the other groups, but the NNJ listeners showed a significant shift in their perceptual behaviour and outperformed the OZ listeners who have no knowledge of Japanese in discriminating the Japanese vowel length contrasts. Furthermore, NNJ was the only group who did not differ in their discrimination accuracy for the Arabic and Japanese stimuli. Taken together, the results obtained in this study suggest that NNJ learned to discriminate Japanese vowel length contrasts to some extent, but the learning did not carry over cross-linguistically to the processing of vowel length contrasts in an unknown language.