Counting roots: an analysis of Portuguese cardinal numbersfrom the perspective of Distributed Morphology

Abstract

This article proposes an analysis of the derivation of cardinal numerals in Portuguese from the perspective of Distributed Morphology and argues in favor of the semantic content of roots. Based on the description of the data, the analysis maintains that the roots that form numerals have semantic content consisting of the cardinality they express. Thus, the meaning of derived words depends on the underspecified core meaning of the root and on the structure in which it is embedded. In Portuguese, numerals can function as arguments (“três é um número primo”) or as predicates (“Eu vi os dois meninos”). In this analysis, this contrast is explained by the categorization of the root, which may be carried out by a nominalizer or an adjectivizer. Both simple and complex numerals are analyzed. The former refers to those formed from a root and a categorizer (“nove, quarenta, cem”), whereas the latter refers to those formed by compounding (“dezesseis, duzentos”), coordination (“vinte e três”), and adjunction (“dois mil”). I propose that List 3 contains interpretive instructions for complex numerals that reflect encyclopedic knowledge about the functioning of the decimal system. Thus, restrictions on root selection are explained in terms of the semantic content of the root and encyclopedic knowledge.