This project seeks to explain the rarity of clicks as consonants in the languages of the world. In most consonants, the source of sound is generated on air coming from the lungs as speakers breathe out. The source of sound is separate from the filtering action of the tongue and lips. The separation of the source of sound from the filter underlies current theories of speech production, and is assumed in all computer generated models of speech. However, in some speech sounds, the separation of the source of sound from the filtering action of the vocal tract is less clear.
For example, for sounds like 'h' that are made in the throat, the source of sound is changed due to the proximity of the tongue constriction to the vocal folds. In click consonants, speakers produce sound by creating a cavity between two parts of the tongue and the palate. The sound is generated by creating a vacuum between the tongue and the palate, which allows air to flow into the vacuum, and is simultaneously filtered by the tongue's shape and position. This project will consist of careful analyses of these complex speech gestures and their resulting sounds. The research team will collect high frame-rate ultrasound data of the tongue during click production. The ultrasound method yields an image of the tongue every 8 thousandths of a second, which is aligned with the audio signal.
The volume of the cavity formed between the tongue and the palate in the production of click consonants will be measured and correlated with the frequency of the sound waves that are produced. These and other analyses will be used to support Dr. Miller's petition to the International Phonetic Association to add two newly documented click sounds to the International Phonetic Alphabet. The results of this project will also contribute to computer speech generation by providing valuable data for future models that can handle source-filter interactions.
