Utterance and pause lengths and frequencies in oral readings by superior esophageal speakers and by normal speakers.

The study of utterance and pause durations and their frequencies of occurrence in human speech provides important insights into pathologic conditionstional analysis of speech produced by laryngectomized individuals, in particular , is useful in evaluating progress and efficacy of esophageal speech, a most frequently employed alternative method of speech production by laryngecto-mized patients. In esophageal speech the lower respiratory system, used as a power source of sound generation in normal processes, is disconnected from the process of speech production. Instead, the patient is taught to " inject " or " swallow " air into the esophagus and to compress the trapped air within the esophageal/pharyngeal walls to set a neoglottis into vibration and create basic esophageal sound for speech (often called " belching " sound). Obviously, the quantity of air trapped within the esophagus is extremely limited. This results in much shorter and more frequent utterances, longer and more frequent pauses, slower speaking rate, and lower voice intensity than normal speech (2. 6, 7, 15, 16). Clinical experience indicates that these characteristics of speech are more prevalent among poor esophageal speakers, especially those who have just begun to learn the technique. Superior esophageal speakers, on the other hand, demonstrate more efficient use of the limited air supply as evi-denced by longer utterance lengths and Iess frequent pauses than inferior esophageal speakers. It is, therefore, of theoretical and practical interest to delineate utterance and pause length as well as the characteristics of superior esophageal speakers and to compare them with normal speakers. There are some methodological concerns related to utterance and pause analysis. Traditional analog methods of analysis (oscillographs, spectrographs, and graphic level recorders) are expensive, time-consuming and laborious, and 446