Gibbon Fiona E, Lee Alice
Department of Speech and Hearing Sciences, University College Cork, Cork, Ireland.
Clin Linguist Phon. 2011 Nov;25(11-12):1014-21. doi: 10.3109/02699206.2011.601393. Epub 2011 Oct 3.
A recurring difficulty for researchers using electropalatography (EPG) is the wide variation in spatial patterns that occurs between speakers. High inter-speaker variability, combined with small numbers of participants, makes it problematic (1) to identify differences in tongue-palate contact across groups of speakers and (2) to define 'normal' patterns during visual feedback therapy. This article shows how graphing EPG data in terms of articulatory separation of phoneme contrasts reduces these two problems to some extent. The graphs emphasise the importance of establishing the presence and extent of separation, as revealed in the EPG data, for phoneme contrasts produced by speakers. Separation graphs for contrasts /i/ - /u/, /s/ - /ʃ/ and /t/ - /k/ are presented using EPG data from adults and children with typical speech and those with speech disorders. When used in conjunction with acoustic and auditory perceptual analyses, it is proposed that representing articulation data in terms of separation will prove useful for a range of clinical and research purposes.
对于使用电腭图(EPG)的研究人员来说,一个反复出现的困难是不同说话者之间出现的空间模式存在很大差异。说话者之间的高变异性,再加上参与者数量较少,使得(1)识别不同说话者群体之间舌腭接触的差异以及(2)在视觉反馈治疗期间定义“正常”模式变得有问题。本文展示了根据音素对比的发音分离来绘制EPG数据如何在一定程度上减少这两个问题。这些图表强调了确定EPG数据中所揭示的音素对比的分离存在和程度对于说话者所产生的音素对比的重要性。使用来自具有典型语音的成人和儿童以及有言语障碍的成人和儿童的EPG数据,展示了/i/ - /u/、/s/ - /ʃ/和/t/ - /k/对比的分离图。当与声学和听觉感知分析结合使用时,有人提出以分离的方式表示发音数据将被证明对一系列临床和研究目的有用。