Search

Studies in grammaticalization have established that processes of grammatical change frequently have lexical items as their source, and more specifically nouns and verbs. But there is at least one potential exception: As argued by Holger Diessel in a number of publications, demonstratives cannot be historically derived from other entities, such as lexical items. Based on a survey of languages from different parts of the world and belonging to different language families, the present paper argues that this generalization is in need of reconsideration.

Semantic maps are tools which represents cross-linguistic regularities in the mapping of meaning on form. Although they have become increasingly popular in semantic studies since the late 1990s, they have been associated with a number of methodological problems, including the reliance on the preferences and intuitions of the individual researcher. In this paper, we argue for the systematic use of large databases to improve the validity of semantic maps. We do so by presenting a study of instrumental and comitative markers based on a sample of 200 languages, and comparing the results with previous research in this area.