Author:

Keywords:

Social Sciences, Linguistics, Language & Linguistics, linguistic methodology, quantitative corpus analysis, variation studies, FIXED-EFFECT FALLACY, LANGUAGE, WORD

Abstract:

© 2008 by Walter de Gruyter GmbH & Co. KG. All rights reserved. A current trend in Cognitive Linguistics is the growing interest in empirical methods for linguistic analysis. Two groups of researchers, centred around Gries and Stefanowitsch (G & S) and around Geeraerts, Speelman and Grondelaers (QLVL), have relatively independently from each other tried to develop a methodology for empirical research in Cognitive Linguistics that is based on thorough quantitative analysis of corpus data. This article discusses a number of case studies into syntactic variation of both approaches. It points out some of the methodological differences between them and compares how they deal with the inherent challenges posed by the spontaneous usage represented in corpora: Both approaches use multivariate statistics to deal with the simultaneous effect of multiple factors, but G & S have additionally developed methods to study the interaction between lexical variability and syntactic variation. On the other hand, G & S take a strongly psycholinguistic perspective on syntactic variation that tends to disregard the offline nature and sociolinguistic heterogeneity of corpus data. These latter two properties are a central concern of QLVL.