Gandour J, Akamanon C, Dechongkit S, Khunadorn F, Boonklam R
Department of Audiology and Speech Sciences, Purdue University, West Lafayette, IN 47907-1353.
Brain Lang. 1994 Jan;46(1):69-95. doi: 10.1006/brln.1994.1005.
Sequences of phonemic approximations (SPAs) to a single target word on a picture naming task were investigated in a Thai-speaking conduction aphasic. A corpus of 207 SPAs was analyzed with an adaptation of Valdois, Joanette, and Nespoulous's (1989) methodology to a tone language. Measures included the length of SPAs, overall degree of approximation to the target of the last attempt and each of the preceding attempts, as well as degree of approximation to the target in terms of number of syllables, syllable structure, segments (consonants/vowels), and tones. Results indicated that there was a steady progression toward the target regardless of eventual outcome or length of sequence. Number of syllables of the target word was better preserved than syllable structure, and tones better than segments. In the course of SPAs, tonal information was accessed earlier than segmental, and syllabic length information earlier than syllabic structure. Within SPAs, tonal information was more stable than segmental across successive approximations. Findings are discussed in relation to a functional model of single word production and multidimensional phonological representations.
我们对一位讲泰语的传导性失语症患者在图片命名任务中对单个目标词的音素近似序列(SPAs)进行了研究。我们采用了瓦尔杜瓦、乔阿内特和内斯普卢斯(1989)的方法并将其应用于声调语言,对207个SPAs的语料库进行了分析。测量指标包括SPAs的长度、最后一次尝试以及之前每次尝试与目标的总体近似程度,以及在音节数量、音节结构、音段(辅音/元音)和声调方面与目标的近似程度。结果表明,无论最终结果或序列长度如何,都朝着目标稳步推进。目标词的音节数量比音节结构保留得更好,声调比音段保留得更好。在SPAs过程中,声调信息比音段信息更早被获取,音节长度信息比音节结构信息更早被获取。在SPAs中,连续近似过程中声调信息比音段信息更稳定。我们结合单词语产生的功能模型和多维语音表征对研究结果进行了讨论。