A feature-based approach to nominal predicates

Abstract This article proposes a feature-based model to derive nominal predicates. I take the position that nouns denote abstract properties and can enter the derivation readily being definite via bearing uniqueness presuppositions. The countability feature [+Count] on the classifier turns abstract properties into countable atoms. The quantifying force [+/−Quant] on the numeral head (for the case of English) either creates set-memberships and formalises indefinite predicates or performs a Max function that does not alter the number of its complement and formalises definite predicates. The morphemes that are traditionally associated with D, i.e., the, a, -s ( Abbott 2006 i.a.) are formal results of feature checking. These features also successfully derive superlatives and weak definites in English and capture the major facts in Mandarin Chinese with variations in allocation and lexically endorsed specifications. Consequently, D is disassociated from introducing (in)definiteness and is reduced to a pure type-shifter with zero lexical exponents.