Brooks P L, Frost B J, Mason J L, Gibson D M
J Speech Hear Res. 1987 Mar;30(1):137-41. doi: 10.1044/jshr.3001.137.
The experiments described are part of an ongoing evaluation of the Queen's University Tactile Vocoder, a device that allows the acoustic waveform to be felt as a vibrational pattern on the skin. Two prelingually profoundly deaf teenagers reached criterion on a 50-word vocabulary (live voice, single speaker) using information obtained solely from the tactile vocoder with 28.5 and 24.0 hours of training. Immediately following word-learning experiments, subjects were asked to place 16 CVs into five phonemic categories (voiced & unvoiced stops, voiced & unvoiced fricatives, approximants). Average accuracy was 84.5%. Similar performance (89.6%) was obtained for placement of 12 VCs into four phonemic categories. Subjects were able to acquire some general rules about voicing and manner of articulation cues.
所描述的实验是对女王大学触觉声码器持续评估的一部分，该设备能让声学波形以振动模式在皮肤上被感知。两名语前极重度失聪青少年仅通过触觉声码器获取的信息，经过28.5小时和24.0小时的训练，在50个单词的词汇量测试（真人语音，单一说话者）中达到了标准。在单词学习实验之后，受试者被要求将16个辅音-元音组合归入五个音位类别（浊音和清音塞音、浊音和清音擦音、通音）。平均准确率为84.5%。将12个元音-辅音组合归入四个音位类别时也获得了类似的表现（89.6%）。受试者能够掌握一些关于发声和发音方式线索的一般规则。
