Abstract
FUL (Featurally Underspecified Lexicon) assumes that a handful of features will account for the phonological systems in the world’s languages. Such an assumption would not be unusual. However, FUL makes several other assumptions including the following: (i) consonants and vowels share place features which are not represented on separate tiers; (ii) features are monovalent; (iii) there are no feature dependencies; (iv) coronal and plosive are always underspecified in representation but present on the surface, which in turn presupposes that both these features must occur in all languages; (v) phonological activity is not the only way to determine feature contrast. These assumptions are based on synchronic, diachronic, and experimental evidence. Detailed case studies examine whether these hypotheses hold in instances where the opposite claims have been made.
