Conceptualization in current approaches of language typology

Abstract

In linguistics, one of the central issues with regard to methodology is conceptualization. Since linguistic analysis consists to a great part of applying concepts to data, it is important to be conscious of what our concepts mean and how we define them. The aim of this paper is to reflect on concept formation in linguistics, with a special focus on language typology. After sketching theories of conceptualization and discussing central issues regarding the definition of concepts, I compare two recent approaches: Multivariate Typology (Bickel and colleagues) and Canonical Typology (Corbett and colleagues). These approaches differ in some respects (e.g., for Multivariate Typology, frequency and statistical analysis are inherent parts of typology and can be used to check our concepts against real data, while Canonical Typologies rely on our ability to a priori identify canonical instances that can be rare or even non-existent). But they also share some aspects such as the idea to define concepts for cross-linguistic comparison through comprehensive lists of properties that can have various values. Language-specific concepts are then understood as having particular values of the properties.