Doyle P C, Danhauer J L, Mendel L L
Vocal Function Laboratory, School of Human Communication Disorders, Dalhousie University, Halifax, Nova Scotia.
J Speech Hear Disord. 1990 Nov;55(4):756-60. doi: 10.1044/jshd.5504.756.
This study was concerned with the perceptual responses of normal-hearing listeners to consonants produced by esophageal and tracheoesophageal (TE) talkers and a single talker who was proficient in both of these alaryngeal speech modes. The listeners' perceptual responses were analyzed using Symmetric Individual Differences Scaling (SINDSCAL) to determine whether distinctive feature differences existed between these two methods of alaryngeal speech. This a posteriori analysis revealed that primary features retrieved for both speech methods included sibilant, affricate, dental, nasal, and sonorant. Although greater perceptual weightings were observed for TE speech, these productive/perceptual features were weighted similarly for both speech methods. Some secondary group-specific feature differences were also observed, but these features did not contribute substantially to the total amount of variance accounted for in the analysis. Thus, the SINDSCAL results showed that the groups did not use different feature systems. These results are discussed in regard to the unique alaryngeal speech production methods employed by esophageal and TE talkers and the relative limitations of the alaryngeal (esophageal) voicing source they use. General clinical implications of the data are discussed.
本研究关注听力正常的听众对食管发声者、气管食管（TE）发声者以及一位精通这两种无喉语音模式的单一发声者所发出辅音的感知反应。使用对称个体差异标度法（SINDSCAL）分析听众的感知反应，以确定这两种无喉语音方法之间是否存在独特特征差异。这一事后分析表明，两种语音方法提取的主要特征包括咝音、塞擦音、齿音、鼻音和响音。尽管观察到TE语音的感知权重更大，但这两种语音方法的这些发音/感知特征权重相似。还观察到一些次要的特定组特征差异，但这些特征对分析中解释的总方差量贡献不大。因此，SINDSCAL结果表明，这些组没有使用不同的特征系统。结合食管发声者和TE发声者所采用的独特无喉语音产生方法以及他们所使用的无喉（食管）发声源的相对局限性，对这些结果进行了讨论。还讨论了数据的一般临床意义。
