A major linguistic study of nominal classification systems across a variety of languages, first published in 2000.

This book offers the first comprehensive survey of the study of gender and classifiers throughout the history of Western linguistics. Based on an analysis of over 200 genetically and typologically diverse languages, the author shows that these seemingly arbitrary and redundant categories play in fact a central role in the lexicon, grammar and the organization of discourse. As a result, the often contradictory approaches to their functionality and semantic motivation encapsulate the evolving conceptions of such issues as cognitive and cultural correlates of linguistic structure, the diverse functions of grammatical categories, linguistic complexity, agreement phenomena and the interplay between lexicon and grammar. The combination of a typological and historiographic perspective adopted here allows the reader to appreciate the detail and insight of earlier, supposedly ‘prescientific’ accounts in light of the data now available and to examine contemporary discussions in the context of prevailing conceptions in the study of language at different points in its history since antiquity.

Linguists have long been interested in systems of nominal classification due to their diverse functions as well as cognitive and cultural correlates. Among others, ongoing research has focused on semantic, functional and morphosyntactic properties of complex systems such as co-occurring gender and numeral classifiers. Such approaches have typically focused on the languages of north-western South America and Papua New Guinea. This volume proposes to fill in a gap in existing research by focusing on Asia, based on case studies from languages belonging to a wide range of families, i.e., Austroasiatic, Austronesian, Dravidian, Hmong-Mien, Indo-European, Mongolic, Sino-Tibetan and Tai-Kadai as well as the language isolate Nivkh. Gender and classifiers in these languages are approached within several different perspectives, i.e., functional, typological and diachronic, thus revealing complex patterns in their lexical and pragmatic functions as well as origin, development and loss. Describing and analysing such properties is a unique and innovative contribution of the volume.

Classification is a popular topic in typological, descriptive and theoretical linguistics. This volume is the first to deal specifically with the diachrony of linguistic systems of classification. It comprises original papers that examine the ways in which linguistic classification systems arise, change, and dissipate in both natural circumstances and in circumstances of attrition. The role of diffusion in such processes is explored, as well as the question of what can be diffused. The volume is not restricted to nominal systems of classification, but also includes papers dealing with the less well-known phenomenon of verbal classification. Languages from a wide spread of world regions are examined, including Africa, Amazonia, Australia, Eurasia, Oceania, and Mesoamerica. The volume will be of interest to linguistic typologists, descriptive linguists, historical linguists, and grammaticalization theorists.

There are two ways for a language to classify its nouns: either by means of classifiers, which specify the semantics of the classified noun, or by means of grammatical gender, which groups all nouns of a language into formal classes. This thesis investigates the common assumption that classifier systems may develop into grammatical gender systems. Because this diachronic phenomenon has not yet been documented for any language, the likeliness that such a development would occur is examined by means of a typological study of synchronic systems. In analyzing the data, this study adopts a new perspective on the development of nominal classification by separating how the means of formal expression develops from the development of those components that have to do with a system's semantic transparency. This twofold account for the data from a variety sample of 40 languages shows that there is indeed a number of systems that lie at the intersection of classifiers and gender systems, but that a direct shift from classifier to gender is not likely to occur.