Anis Farheen Naz, Umat Cila, Ahmad Kartini, Hamid Badrulzaman Abdul
a Centre For Rehabilitation and Special Needs, Faculty of Health Sciences , Universiti Kebangsaan Malaysia , Jalan Raja Muda Abdul Aziz 50300 , Kuala Lumpur , Malaysia.
b Institute of Ear, Hearing & Speech, Universiti Kebangsaan Malaysia , Kuala Lumpur , Malaysia.
Cochlear Implants Int. 2019 Jan;20(1):12-22. doi: 10.1080/14670100.2018.1530420. Epub 2018 Oct 6.
This study examined the patterns of recognition of Arabic consonants, via information transmission analysis for phonological features, in a group of Malay children with normal hearing (NH) and cochlear implants (CI).
A total of 336 and 616 acoustic tokens were collected from six CI and 11 NH Malay children, respectively. The groups were matched for hearing age and duration of exposure to Arabic sounds. All the 28 Arabic consonants in the form of consonant-vowel /a/ were presented randomly twice via a loudspeaker at approximately 65 dB SPL. The participants were asked to repeat verbally the stimulus heard in each presentation.
Within the native Malay perceptual space, the two groups responded differently to the Arabic consonants. The dispersed uncategorized assimilation in the CI group was distinct in the confusion matrix (CM), as compared to the NH children. Consonants /ħ/, /tˁ/, /sˁ/ and /ʁ/ were difficult for the CI children, while the most accurate item was /k/ (84%). The CI group transmitted significantly reduced information, especially for place feature transmission, then the NH group (p < 0.001). Significant interactions between place-hearing status and manner-hearing status were also obtained, suggesting there were information transmission differences in the pattern of consonants recognition between the study groups.
CI and NH Malay children may be using different acoustic cues to recognize Arabic sounds, which contribute to the different assimilation categories' patterns within the Malay perceptual space.
本研究通过对语音特征的信息传递分析，考察了一组听力正常的马来儿童和人工耳蜗植入儿童对阿拉伯辅音的识别模式。
分别从6名人工耳蜗植入儿童和11名听力正常的马来儿童中收集了336个和616个声学样本。两组儿童在听力年龄和接触阿拉伯语音的时长方面进行了匹配。所有28个以辅音-元音/a/形式呈现的阿拉伯辅音通过扬声器以约65 dB SPL的音量随机播放两次。要求参与者口头重复每次呈现中听到的刺激音。
在马来语母语感知空间内，两组儿童对阿拉伯辅音的反应不同。与听力正常的儿童相比，人工耳蜗植入组在混淆矩阵（CM）中分散的未分类同化现象明显。/ħ/、/tˁ/、/sˁ/和/ʁ/这几个辅音对人工耳蜗植入儿童来说比较困难，而识别最准确的是/k/（84%）。人工耳蜗植入组传递的信息显著减少，尤其是在位置特征传递方面，低于听力正常组（p < 0.001）。还获得了位置-听力状态和方式-听力状态之间的显著交互作用，表明研究组之间在辅音识别模式上存在信息传递差异。
人工耳蜗植入的马来儿童和听力正常的马来儿童可能使用不同的声学线索来识别阿拉伯语音，这导致了马来语感知空间内不同同化类别的模式。
