Identification Of English Consonants Research Articles

The impact of L2-learning experience and target dialect on predicting English vowel identification using Korean vowel categories

Park and de Jong (2008, 2017) reported that their L1-based model predicted the identification of English consonants with Korean analogues quite successfully while error rates for those without Korean analogues were systematically greater than the model predicted, and Lee and Cho (2018) obtained similar results for English vowels. However, Park and de Jong (2008, 2017) and Lee and Cho (2018) examined only L1-Korean inexperienced EFL listeners’ identification of English consonants or vowels. English vowels such as /æ-a/ (bat), /ɝ-ɜ/ (Burt), and /ɑ-ɒ/ (stop) exhibit dialect differences between North American English (NAE) and Standard Southern British English (SSBE), but few researchers have examined Korean second language (L2) listeners’ perceptions of English vowels in regional dialects other than NAE. The current study was an investigation of the degree to which the L1-based model predicts L1-Korean experienced L2 listeners’ identification of NAE and SSBE vowels based on the predictions that experienced L2 listeners are less likely to use L1 (first language) categories in L2 vowel identification than are inexperienced L2 listeners and that even experienced L2 listeners might use L1 categories when identifying L2 vowels in a less familiar English dialect. Korean listeners with different lengths of residence (LOR) in the United States (long: 11.1 years vs. short: 4 years) and with limited experience with British English completed English vowel identification and English-Korean vowel mapping tasks. The results show that the L1-based model predicted only the short-LOR listeners’ identification of NAE and SSBE vowels rather accurately, indicating the effect of L2-learning experience. The model also more accurately predicted the short-LOR listeners’ identification of SSBE vowels relative to NAE vowels, showing that the impact of the amount of L2 learning-experience decreased with a less familiar dialect. When the listeners’ goodness ratings of the category mapping were considered, the predictive power of the model weakened. Further, the new-similar distinction for L2 English vowels vis-à-vis L1-Korean may be gradient, and the new-similar distinction is variable depending on L2-learning experiences.

English Consonant Identification in Multi-Talker Babble: Effects of Chinese-Native Listeners’ English Experience

The identification of English consonants in quiet and multi-talker babble was examined for three groups of young adult listeners: Chinese in China, Chinese in the USA (CNU), and English-native listeners. As expected, native listeners outperformed non-native listeners. The two non-native groups had similar performance in quiet, whereas CNU listeners performed significantly better than Chinese in China listeners in babble. It is concluded that CNU listeners may benefit from English experience, for example, better use of temporal variation in noise and better capacity against informational masking, to perceive English consonants better in babble. Possible explanations regarding the differential noise effect on the three groups are discussed.

“If a perceptual problem exists, what are its roots?” Types of cross-language similarity, hypercorrection, and unexplored root canals

James Jenkins had a keen interest in the nature of perceptual problems of nonnative listeners and in the perceived similarity of speech sounds. This presentation reports on a series of experiments which examined how well four different types of similarity predict nonnative speech perception. The different types are ecphoric and perceptual cross-language similarity, perceived within-language similarity, and acoustic similarity. One set of experiments examined how well cross-language perceptual assimilation (English to Danish) and within- language similarity ratings (English-English) of English consonants by Danish listeners predict Danes’ identification of English consonants. Another set of experiments explored the roots of English listeners’ discrimination problems for the closely spaced Danish unrounded front vowels by relating these problems to two types of perceived cross-language similarity (ecphoric and perceptual) and to acoustic similarity. Results suggest that each of the four types of similarity accounts for some of the perceptual problems, but none does so exhaustively, probably because the root system of these problems is affected by additional factors including nonnative listeners’ hypercorrection and by perceptual asymmetries. The clear conclusion from these experiments is that Jim’s question will keep us busy for quite some time. [Work supported by Carlsberg Foundation and Inge Lehmanns Legat af 1983.]

日本語母語話者によるバブルノイズ下の母音間英語子音知覚 : 英語習熟度と子音知覚の相関関係(音響教育特集号)

日本語母語話者によるバブルノイズ下の母音間英語子音知覚 : 英語習熟度と子音知覚の相関関係(音響教育特集号)

Identification of English consonants in intervocalic contexts in multispeaker babble noise by Japanese listeners: Correlation between English proficiency and consonant identification ability

Identification of English consonants in intervocalic contexts in multispeaker babble noise by Japanese listeners: Correlation between English proficiency and consonant identification ability

Identification of English Consonants: Focusing on Labial, Coroal, and Velar Sounds

Identification of English Consonants: Focusing on Labial, Coroal, and Velar Sounds

Identification of English Consonants and Vowels in Sentence-Embedded Words by Professional Interpreters and Student Interpreters

Identification of English Consonants and Vowels in Sentence-Embedded Words by Professional Interpreters and Student Interpreters

Identification of English Consonants by Interpreting Students and Professional Interpreters

Identification of English Consonants by Interpreting Students and Professional Interpreters

Identification of English consonants by Korean students

Previous studies have shown that Korean students have difficulty identifying some English consonants that are not in the Korean sound inventory. The aim of this study was to examine the accuracy score of English consonants correctly identified by 130 college students in order to find out English consonants with listening problems for the Korean students. The subject’s task was to identify one of the minimal pairs played in a quiet laboratory classroom. One hundred minimal pairs consisted of syllables with various onsets or codas: stops, fricatives, affricates, liquids, and nasals. Results were as follows: First, types of the correctly identified English consonants varied according to the level of each student’s total score. Second, the lower the group level, the lower the accuracy score for each consonant group divided by its manner. Third, those consonants in coda position of the test words were more difficult for the students to identify than those in onset position. Fourth, the subjects showed lower accuracy in the perception of any contrast with a fricative onset or coda. Finally, the students still had problems telling voiced from voiceless consonants, especially in coda position. Future studies will focus on examining students’ production data to trace their listening problems.

Modeling the perception of features in the identification of English consonants

The perception of voicing, place, and manner was investigated in a series of gating experiments in which subjects identified eight English consonants in six different duration conditions [T. L. Doeleman, J. Acoust. Soc. Am. 101, 3111(A) (1997)]. Subjects heard either all eight stimuli or two mutually exclusive subsets such that certain feature information was given (e.g., manner given in sets [p, t, b, d] and [f, s, v, z]). Although all subjects listened to the same 48 stimuli, identification significantly improved for subjects given place information (i.e., sets of labials and alveolars). The results are modeled using a spreading-activation connectionist model, the Integration-Competition model [Spivey-Knowlton (1996)]. In this model, the phoneme and feature representations compete with one another such that featural distinctions correspond to different weights assigned to the connection between feature and phoneme layers. The target phoneme has high activation from all featural arrays, resulting in the highest activation in the phoneme array. The increase in identification scores across duration conditions is modeled by the increase in activation of the target phoneme with successive time steps. The model also predicts observed consonant confusion patterns in that distracter phonemes sharing more features with the target phoneme will have higher activation levels.

Non-natives’ identification of English consonants in noise

This study examined the identification of English consonants presented in noise. The subjects were English native (NE) speakers and four groups of Italian/English bilinguals who differed according to age of arrival (AOA) in Canada and/or self-estimated percentage daily use of Italian (%ITAL). The naturally produced stimuli, which differed according to their initial or final consonant (e.g., ‘‘bado,’’ ‘‘sado,’’ ‘‘hodab,’’ ‘‘hodas’’), were presented at decreasing S/N ratios (+12, +6, 0, −6 dB). Subjects identified each consonant by selecting one of five written alternatives, including the target and four likely foils. Percent correct identification scores decreased similarly across the four S/N ratios for all five groups. However, the bilinguals scored lower than did the NE subjects, despite their long residence in Canada (M=35.4 years). Their scores varied inversely with AOA and %ITAL, and were related to their performance on a test of phonological short-term memory. Which consonants were most difficult to identify varied across the five groups. Error analyses revealed that errors due to confusions of place and manner of articulation were similar across the five groups. Additional analyses will determine whether the influence of the L1 can be characterized in terms of finer-grained features.

Perception of features in the identification of English consonants

The perception of voicing, manner, and place of articulation was investigated in a series of gating experiments in which subjects identified English consonants in six different duration conditions. The stimuli were created from natural speech consonant tokens excised from C[a] syllables. The consonants were then gated to include 10% to 50% of the total consonant duration from the onset, to which vowel babble was appended in order to make listeners less sensitive to the absolute duration of the stimuli. In experiment 1, subjects were presented with eight consonants [p, t, b, d, f, s, v, z] and results were analyzed in terms of correct identification for the following features: voiced or voiceless, stop or fricative, and labial or alveolar. In experiment 2, subjects were presented with two mutually exclusive sets of stimuli (subsets of the stimuli from experiment 1) such that the information for a particular feature was given (e.g., voicing given in the sets [p, t, f, s] and [b, d, v, z]). A comparison of the results of the two experiments will show whether certain feature information facilitates the perception of any other feature. Results will be discussed in relation to a psychological model of speech perception.