A statistical method for the identification and aggregation of regional linguistic variation

Abstract

A B S T R A C T This paper introduces a method for the analysis of regional linguistic variation. The method identifies individual and common patterns of spatial clustering in a set of linguistic variables measured over a set of locations based on a combination of three statistical techniques: spatial autocorrelation, factor analysis, and cluster analysis. To demonstrate how to apply this method, it is used to analyze regional variation in the values of 40 continuously measured, high-frequency lexical alternation variables in a 26-million-word corpus of letters to the editor representing 206 cities from across the United States.

A multivariate spatial analysis of vowel formants in American English

“…In order to identify statistically significant patterns of regional variation in the values of the thirty-eight individual vowel formant variables, each variable was subjected to a spatial autocorrelation analysis (Grieve, 2011(Grieve, , 2012Grieve et al, 2011).…”

“…To calculate both spatial autocorrelation measures, it is necessary to define a spatial weighting function, which is a set of rules that assigns a weight to the comparison of every pair of locations so that comparisons between locations that are close together are given greater weight than comparisons between locations that are far apart (Odland, 1988;Grieve, 2011Grieve, , 2012Grieve et al, 2011). In this study, a reciprocal spatial weighting function was used, which assigns a weight to each pair of locations based on the reciprocal of the distance between the locations so that the weight decreases with distance.…”

“…see Prokić & Nerbonne, 2008;Nerbonne & Heeringa, 2009;Wieling & Nerbonne, 2010;Grieve et al, 2011). In particular, Ward's method for hierarchical clustering (Ward, 1963) was used because it tends to identify the clearest dialect regions and because it is one of the most common methods for hierarchical clustering in dialectometry.…”

Site-Restricted Web Searches for Data Collection in Regional Dialectology

“…Labov et al, 2006;see Grieve et al, 2013), or corpus-based methods (e.g. Grieve et al, 2011) are used for data collection.…”

“…The maps for each variant were therefore subjected to a Getis-Ord Gi local spatial autocorrelation analysis (Ord and Getis, 1995) in order to identify underlying patterns of regional variation (see also Grieve, 2011Grieve, , 2012Grieve et al, 2011Grieve et al, , 2013 . A Getis-Ord Gi analysis is a geostatistical technique that identifies significant patterns of spatial clustering in the values of a variable that is measured over a series of locations.…”