Consonants, vowels, and nasality

Abstract

AbstractContrastive features establish phonological contrasts. FUL assumes that a small number of features, defined by articulatory and acoustic correlates, suffice to represent all phonological contrasts. A fundamental aspect of the model is that place features are identical for consonants and vowels. There are no dependencies between features except for class features. Focusing on features for nasal vowels and consonants, their sufficiency to account for phonotactic constraints and assimilations, particularly for place, is established. Evidence comes from synchronic and diachronic analyses (including loan adaptation) and experimental evidence for processing and representation. Evidence comes largely from Bengali, which has an equal number of nasal and oral vowels and five places of articulation, providing a rich repertoire of neutralization contexts and representational constraints. The lack of dependent features, shared features for vowels and consonants, and underspecification appears to be supported from phonotactic constraints and assimilations of nasal–obstruent sequences as well as experimental evidence.