Exploring Language Variation Across Europe - A Web-based Tool for Computational Sociolinguistics

Exploring Language Variation Across Europe A Web-based Tool for Computational Sociolinguistics Dirk Hovy and Anders Johannsen Center for Language Technology, University of Copenhagen {dirk.hovy,ajohannsen}@hum.ku.dk Abstract Language varies not only between countries, but also along regional and socio-demographic lines. This variation is one of the driving factors behind language change. However, investigating language variation is a complex undertaking: the more factors we want to consider, the more data we need. Traditional qualitative methods are not well-suited to do this, and therefore restricted to isolated factors. This reduction limits the potential insights, and risks attributing undue importance to easily observed factors. While there is a large interest in linguistics to increase the quantitative aspect of such studies, it requires training in both variational linguistics and computational methods, a combination that is still not common. We take a first step here to alleviating the problem by providing an interface, www.languagevariation.com, to explore large-scale language variation along multiple socio-demographic factors – without programming knowledge. It makes use of large amounts of data and provides statistical analyses, maps, and interactive features that will enable scholars to explore language variation in a data-driven way. Keywords: language variation, social media, web explorer, computational sociolinguistics 1. Introduction Language varies. Not only over time, but also between peo- ple who live at the same time, and within the same person between different situations. We do not expect a teenager to speak the same way as a pensioner, and we would not talk the same way at a dinner party as we would at a scientific conference, or at a six-year-old’s birthday party. We use language, sometimes actively, more often subconsciously, to mark our membership in a group, defined by socio- demographic factors: age (Barke, 2000; Barbieri, 2008; Rickford and Price, 2013), gender (Holmes, 1997; Rick- ford and Price, 2013), regional origin (Schmidt and Her- rgen, 2001; Nerbonne, 2003; Wieling et al., 2011), social class (Labov, 1964; Milroy and Milroy, 1992; Macaulay, 2001; Macaulay, 2002), ethnicity (Carter, 2013; Rickford and Price, 2013), and many more. At the same time, we use language to distinguish ourselves from other groups within the same socio-demographic category (Silverstein, 2003; Agha, 2005): young vs. old, men vs. women, town vs. country. Since everyone is at the intersection of mul- tiple such socio-demographic groups, language is affected by the interaction of many of these competing factors. Interactions between demographic factors can help explain similarities between otherwise distinct groups, say between age and gender, if we observe the same phenomenon in young women and elder men. So while interactions provide us with more explanatory power, they come at a cost: the more factors we consider, the more ways to interact there are. For two factors (say, age and gender) there is only one possible interaction, but this number grows exponentially, as a binomial coefficient of the number of factors. Tripling the number of factors to six increases the number of poten- tial pairwise interactions to 15. And so, while language variation is an active research area in …

Content truncated to 500 words