بررسی و دست کاری متغیرهای آکوستیکی جملات سوالی و تبدیل آنها به جمله خبری بر اساس مدل شیب

This paper investigates intonational characteristics of yes/no questions in contrast to statements in Persian. Based on Tilt model, acoustic variables including duration, pitch amplitude and tilt are compared between these two sentence types. Since studies (Eslami, ۲۰۰۰; Mahjani, ۲۰۰۳; Mahootiyan, ۱۹۹۷; Sadat Tehrani, ۲۰۰۷) have shown the difference between yes/no question and statement is in the last intonational event, the goal is to determine which acoustic parameters are conceptually important in identifying the type of the sentence. To achieve this, at first, in the production stage, we used statistical analyses in relation to different levels of duration, pitch amplitude and tilt in questions and statements. Then, in the perception stage, first, we manipulated   acoustic variables (duration, pitch amplitude and tilt), then perceptual tests were used to measure the success of work. In the production stage, ۴۰ yes/no questions and their corresponding and statements were read by five Persian speakers, and then through Praat software, the data were analyzed and the acoustic parameters measured. In the next step, by means of Student's t- test, we compared the mean of different variables in the questions and statements. The results of our statistical analyses showed that the average of duration and pitch amplitude except tilt variable were higher in final peak in questions. In the next stage, we manipulated the target points, including: edge f۰, final peak, rightmost valley and penultimate valley in order to change the question to statement. Perceptual salience of these acoustic cues was investigated through questionnaire identification tests. The perception tests indicate that only the utterance-final edge F۰ is a strong perceptual cue to make a distinction between question and statement. And simoltaneovs manipulation of all target points showed greatest effect on identifying the type of sentence.