Articulatory Modeling and the Definition of Acoustic-perceptual Targets for Reference Vowels

The present paper proposes some hints to set up a number of prototypical vowels as acoustic-perceptual targets (APT). APTs are defined using theoretical considerations and the extensive use of an articulatory model for exploring the acoustic and perceptual space. This paper is a revisit of the notion of Cardinal Vowels, which were mainly described in articulatory terms. Our approach avoids the problem of the articulatory system's compensatory abilities to attain a given APT with different articulatory strategies. The availability of well-defined APTs makes it possible to characterize both acoustically and articulatorily subtle but audible deviations between the different renditions of some of the sounds. Moreover, the proposed notation has the potential to highlight the acoustic characteristics common to all phonemes (vowels, semi-vowels, consonants) that share a similar place of constriction, on the one hand, and coarticulatory effects (such as palatalization, labialization and velarization), on the other hand.