Although the last 50 years have witnessed a rapid growth in the understanding of vowel articulation and acoustics, most contemporary theories of speech perception have concentrated on consonant perception. Authored by leading academic and industrial authorities, this volume is intended to balance such a bias. The authors propose a computational theory of auditory vowel perception that accounts for vowel identification in the face of acoustic differences between speaker, speaking rate, stress. Topics include: acoustic and auditory effects of articulation, vowel categorization, and vowel constancy, among others. This work lays the foundation for future experimental and computational studies of vowel perception. With its important applications to linguistics research and artificial intelligence, this book will be eagerly read by students and researchers in psychology, linguistics, and computer science.