Abstract
Dialect variation is of considerable interest in linguistics and other social sciences. However, traditionally it has been studied using proxies (transcriptions) rather than acoustic recordings directly. We introduce novel statistical techniques to analyze geolocalized speech recordings and to explore the spatial variation of pronunciations continuously over the region of interest, as opposed to traditional isoglosses, which provide a discrete partition of the region. Data of this type require an explicit modeling of the variation in the mean and the covariance. Usual Euclidean metrics are not appropriate, and we therefore introduce the concept of d-covariance, which allows consistent estimation both in space and at individual locations. We then propose spatial smoothing for these objects which accounts for the possibly nonconvex geometry of the domain of interest. We apply the proposed method to data from the spoken part of the British National Corpus, deposited at the British Library, London, and we produce maps of the dialect variation over Great Britain. In addition, the methods allow for acoustic reconstruction across the domain of interest, allowing researchers to listen to the statistical analysis. Supplementary materials for this article are available online.
Supplementary Materials
Online supplement: Details on the data preprocessing, simulation study, and example illustrating the advantage of d-covariances.
“Class” dataset and related functions: Dataset and functions used in the illustration of our smoothing method in Section 4.
Sounds: Samples of the vowel sound data, effect of the PC loadings, and examples of reconstructed sounds.
Notes
1 The maps shown in this article do not include the Isle of Wight, located off the south coast of Great Britain, as there is no data present here, and as it is not simply connected to the rest of mainland UK, it is not possible to provide smooth estimates there.