Doctoral Dissertation Research: The production and perception of breathy voice during nasal sounds

University Of California-San Diego, La Jolla CA

Investigators

Abstract

Language users often adjust or enhance their speech to be better understood in different contexts, such as when speaking in a loud room, or to a person with an accent, or when speaking on the phone. The study of speech enhancement strategies addresses important questions about human language, including how language is perceived by listeners and represented in the mind. Answering these questions is also vital to both speech therapy and speech technology. For speech therapy, this understanding can improve methods used in diagnosis and treatment of disordered speech. For speech technology, this understanding can improve speech recognition and synthesis, which are both used daily by the public, such as when calling a company's customer service. Knowledge of the way we enhance speech also helps researchers understand how and why languages change over time. This understanding is crucial for educating the public on linguistic diversity, change, and preservation. This research seeks to understand what enhancements occur when making nasal sounds, which are found in almost all languages of the world, including American English. Nasality is a common feature associated with an individual's voice: in any language, some speakers sound more nasal than others. This research focuses on how the voice is used to make sounds nasal. First, the researchers investigate whether speakers of American English and Brazilian Portuguese produce nasals with voice adjustments from the larynx. Second, they investigate how the perception of voice influences the perception of nasal sounds. These two languages differ in their use of nasal sounds, and so together they provide a better understanding of how different types of nasal sounds are made in different languages. This knowledge will help clarify how this particular type of sound is produced and perceived, which has the potential to improve techniques used to treat disordered speech as well as software used in speech recognition and synthesis.