This book deals with the traditional problem of the classification of linguistic units, with a primary focus on word classes. The approach is descriptive rather than theoretical, and is based on the use of distinctive features analogous to the ones used in phonology, which entails a radical reworking of the traditional classification. The first part presents some basic notions such as the use of distinctive features and the role of word classes in grammar; classification by prototypes; and the use of world knowledge as a resource to assign thematic relations to constituents in the sentence. In the second part, some descriptive problems are examined, namely the classification of verbs according to valency; connectives, adverbs, and the internal constituents of the NP; and the classification of units larger than words. This book will be of use as a guide for linguists working on the description of natural languages, as well as a resource for students on courses in linguistic theory and description.

The elaboration of linguistic theories depends on the existence of adequate descriptions of particular languages; otherwise theories will be poorly grounded on empirical data. This book starts from theoretical points of wide acceptance among linguists and goes on to present a descriptive metalanguage, able to express the facts of verb valency, which constitute one of the core areas in linguistic description. Most of the data come from an extensive survey under way of the valency of Portuguese verbs; but the present work's relevance goes well beyond that, and incorporates a proposal applicable to other European languages, illustrated by the wealth of English examples included in the exposition. Among the topics discussed are the syntactic component of constructions (following here a proposal recently published in Culicover and Jackendoff's Simpler Syntax); delimitation and definition of semantic roles; the role of linking rules and their relation to prototypes; and the connection between linguistic expressions and cognitive units such as frames and schemata. The result is a notational system flexible and robust enough to describe all aspects of verb valency.

This book deals with the traditional problem of the classification of linguistic units, with a primary focus on word classes. The approach is descriptive rather than theoretical, and is based on the use of distinctive features analogous to the ones used in phonology, which entails a radical reworking of the traditional classification. The first part presents some basic notions such as the use of distinctive features and the role of word classes in grammar; classification by prototypes; and the use of world knowledge as a resource to assign thematic relations to constituents in the sentence. In the second part, some descriptive problems are examined, namely the classification of verbs according to valency; connectives, adverbs, and the internal constituents of the NP; and the classification of units larger than words. This book will be of use as a guide for linguists working on the description of natural languages, as well as a resource for students on courses in linguistic theory and description.

The elaboration of linguistic theories depends on the existence of adequate descriptions of particular languages; otherwise theories will be poorly grounded on empirical data. This book starts from theoretical points of wide acceptance among linguists and goes on to present a descriptive metalanguage, able to express the facts of verb valency, which constitute one of the core areas in linguistic description. Most of the data come from an extensive survey under way of the valency of Portuguese verbs; but the present work's relevance goes well beyond that, and incorporates a proposal applicable to other European languages, illustrated by the wealth of English examples included in the exposition. Among the topics discussed are the syntactic component of constructions (following here a proposal recently published in Culicover and Jackendoff's Simpler Syntax); delimitation and definition of semantic roles; the role of linking rules and their relation to prototypes; and the connection between linguistic expressions and cognitive units such as frames and schemata. The result is a notational system flexible and robust enough to describe all aspects of verb valency.