The MIT Press, 1999. — 606 pp.
The application of quantitative modeling methods to the acoustics of speech sound production underwent a major advance with the work of Gunnar Fant. His book, Acoustic Theory of Speech together with his continuing work and that of his students and colleagues in the 1960s, 1970s, and 1980s, has been a major stimulus to raising the field of acoustic phonetics toward the level of a quantitative science. This book attempts to build on this earlier work of Fant and others over these decades.
The aim of the book is to present a theory of speech sound generation in the human vocal system. The acoustic aspects of this theory are grounded in a representation of speech sound production as the creation of sound sources in the vocal tract and the filtering of these sources by the vocal tract airways. The sources can be of various kinds, including the quasi-periodic laryngeal source, noise due to turbulent airflow, and transient sounds. Since the articulators move with time, the sound sources and the filtering also change with time. Examination of the time-varying sources and filtering leads to the observation that some aspects of the transformation from articulation to sound are categorial. That is, the types of sound sources and the filtering of these sources can be organized into classes. These classes are closely related to the discrete linguistic categories or features that describe how words appear to be stored in the memory of a speaker or listener. The theme of this book is to explore these relations between the discrete linguistic features and their articulatory and acoustic manifestations.
The book begins with a review of the anatomy of the speech production system, and a discussion of some principles relating airflows and pressures in the vocal tract. The next four chapters describe mechanisms of sound source generation in the vocal tract, present theories of the vocal tract as an acoustic resonator exalted by these sources, review some principles of auditory psychophysics and auditory physiology as they may be relevant to auditory processing of speech, and present an introduction to phonological representations. With these five chapters as background, the remaining chapters are devoted to a detailed examination of the vowels (chapter 6), the consonants (chapters 7 to 9), and some examples of how speech sounds are influenced by context (chapter 10).
Little attention is given to the description and modeling of the production of sounds in languages other than English. The aim is not to be complete, even for the sounds of English, but rather to present an approach to modeling the production of speech Sounds in general. An attempt is made to show that, when reasonable assumptions are made about the physiological parameters involved in producing a sound sequence, acoustic theory can make predictions about the sound pattern, and these predictions agree well with the measured pattern. A goal for the future is to extend this modeling effort to a wider variety of speech events across languages, and to examine not only the broad acoustic characteristics of different sounds but also their vari- ability across speakers and across contexts.
The book is intended to be useful to speech scientists, speech pathologists, linguists with an interest in phonetics and phonology, psychologists who work in areas related to speech perception and speech production, and engineers who are concerned with speech processing applications. This book has evolved from notes for a graduate course in Speech Communication at the Massachusetts Institute of Technology. The course is taken by students in engineering and in a graduate program in Speech and Hearing Sciences. Some students in linguistics, cognitive sciences, and medical engineering also attend the course.
Anatomy and Physiology of Speech Production
Source Mechanisms
Basic Acoustics of Vocal Tract Resonators
Auditory Processing of Speechlike Sounds
Phonological Representation of Utterances
Vowels: Acoustic Events with a Relatively Open Vocal Tract
The Basic Stop Consonants: Bursts and Formant Transitions
Obstruent Consonants
Sonorant Consonants
Some Influence of Context on Speech Sound Production