How Does Speech Timing Work?

The fellowship will support the completion of a book, titled 'How Does Speech Timing Work?'. Speakers manipulate speech sound durations for a variety of meaning-related purposes. This book will discuss the kinds of timing patterns people produce when they speak, and will evaluate theories of how speech articulation is controlled to produce these timing patterns. Because speech timing patterns are often termed rhythmic, it will also discuss available definitions of speech rhythm, and will evaluate rhythmicity claims for speech against available evidence. This book will be of interest to anyone interested in how speech production works, including linguists, psycholinguists, motor control specialists, speech technologists, and speech therapists.

This research has several beneficiaries outside of academia.

Models of speech production timing are relevant for speech technology applications, speech synthesis in particular. The discussion of the stretches of speech whose durations are affected by timing processes will be of most use; many of these findings are new, and summaries are few. Our book will provide an accessible path towards understanding this literature. Articulatory synthesis work will benefit from the discussion of motor control issues in speech timing.

Models of speech timing are also highly relevant for individuals with speech timing disorders, and for the clinicians who diagnose and treat them. Parkinsons' disease, apraxia, and stuttering all involve surface timing disorders. Understanding the normal control of timing is a pre-requisite to assessment and therapy of these disorders.

Our work may also interest musicians who wonder about the relationship between timing in music vs. timing in speech. In particular they may be interested to know that the primary purpose of timing in non-poetic, non-musical speech is to convey meaning. On the other hand, timing control in music is geared primarily towards maintaining the regular recurrence of rhythmic intervals. The rhythmic control of timing in music enables synchrony with other performers; synchrony with other talkers is not normally required in speech.

We expect that speech technologists and musicians will be reached through the publication of the book. Additional work will be required to ensure benefit by individuals with speech disorders. As a first step towards clinical relevance, we will present our work to speech therapists and pathologists. Feedback from these presentations will enable us to learn more about timing disorders. Clinicians and speech pathology researchers will learn about the range of normal speech timing patterns; this knowledge may be helpful in disorder diagnosis. In addition, they will learn about the range of available theories of timing control mechanisms. As research progresses that narrows down the set of plausible control mechanisms for normal speech timing, researchers will be able to determine the timing mechanisms that are not functioning properly in disordered individuals. Knowing which mechanisms are abnormal will increase the likelihood of successful therapeutic techniques.