Approaching lexical typology
Maria Koptjevskaja-Tamm To appear in: Vanhove, Martine (ed.), TOWARDS A TYPOLOGY OF LEXICAL SEMANTIC ASSOCIATIONS.
The aim of the present article is to situate the research direction presented in the volume within the larger domain of modern typological research in general. As witnessed by the title of the volume, “Towards typology of lexicalsemantic associations”, three key words are crucial here – typology, lexical and semantic (cf. the title. The paper starts by a short summary of what is meant by typological research in general, then discusses the relation between semantic and lexical typology and points out several different groups of questions asked within lexical-typological research, its different foci. Chapter 2 touches onthe general premises for lexicaltypological research – possible words, semantic generality vs. polysemy, and the meaning of ‘meaning’. Chapters 3–5 are devoted to the three main lexical-typological research foci introduced in Chapter 1 – what meanings can(not) be expressed by a single word, what different meanings can be expressed by one and the same lexeme or by words derivationally related toeach other, and what cross-linguistic patterns there are in lexicon-grammar interaction. Each of these chapters considers numerous examples of research and the various methodological and theoretical issues relevant for it. The whole paper ends up with the general discussion of the urgent methodological problems facing lexical typology as a field.
1. Typology, semantic and lexical typology
Theterm “typology”, as is well known, has many different uses. What primarily matters for the present volume is typology understood as “the study of linguistic patterns that are found cross-linguistically, in particular, patterns that can be discovered solely by cross-linguistic comparison” (Croft 1990: 1). Typology can also refer to typological classification of languages into (structural) types on thebasis of particular patterns for particular phenomena. Typological research is driven by the persuasion that the variation across attested (and, further, possible) human languages is severely restricted, and aims therefore at unveiling systematicity behind the whole huge complex of linguistic diversity. In pursuing their tasks, typologists raise – and often try to answer – important theoreticalquestions, such as • According to what parameters does a specific phenomenon vary across languages, in what patterns do these parameters (co-)occur? • What generalisations can be made about attested vs. possible patterns? • What is universal vs. language particular in a given phenomenon, what phenomena are frequent vs. rare? • How are various linguistic phenomena distributed across the languages ofthe world? • Which phenomena are genetically stable and which are subject to contactinduced change? • How can the attested distribution of the different patterns across languages be explained? • How can the attested cross-linguistic patterns / generalizations be explained?
2 The papers in the present volume do in fact focus on linguistic patterns that can be discovered only bycross-linguistic comparison – cross-linguistically recurrent patterns of polysemy, heterosemy and semantic change – and are therefore examples of typological research. The domain of research shared by the papers in the volume is, however, somewhat outside of the main interests of modern typological research, that has so far primarily focused on grammatical and, to a lesser degree, phonetic / phonologicalphenomena under the labels of “grammatical typology”, “syntactic typology”, “morphological typology”, “morphosyntactic typology” (or, quite often, just “typology”), “phonetic typology” and “phonological typology”. None of those would suit the direction of the volume. We are dealing here, first, with lexical and, second, with semantic phenomena – which are the primary objects of lexical vs. semantic...