International Journal of the Linguistic Association of the Southwest


Texas Spanish has been characterized from a variety of perspectives, whether that be descriptively (Cardenas 1970), historically (Lipski 1988), and/or sociolinguistically (Chaston 1996, Martínez 2003, Bayley et al. 2012, Carter and Wolford 2016). However, to date, no study has sought to produce a dialectological account of Spanish discourse in Texas. By combining methods of quantitative and corpus linguistics, the present study seeks to investigate the dynamics of how Spanish discourse features, namely discourse markers and fillers, vary across multiple regions in Texas. To do so, this study utilizes a corpus of contemporary Texas Spanish (Bullock and Toribio 2013) and evaluates it via a dialectometric approach (Speelman and Geeraerts 2008), applying various computational methods to analyze this variation from a geolinguistic perspective. A battery of dialectometric analyses (aggregate linguistic distance, multidimensional scaling, cluster analysis) reveal considerable subdialectal variation among the usage of discourse markers and fillers. More specifically, the results suggest the existence of six dialect clusters corresponding to four larger dialect cluster divisions. Additionally, the use of loaned and code-switched English discourse markers does vary across different locations in Texas. While these markers appeared across all locations, the majority of them were utilized in locations near the Mexican border (El Paso, Edinburg, and Weslaco). Therefore, although somewhat curious, proximity to Mexico does not seem to correlate with less English transfer.Overall, the various computational methods provided consistent outcomes and validation tests indicate confidence in the given clustering models. However, the present study offers only a snapshot of the true nature of discursive variation in this context, and more data is necessary to fully understand the geolinguistic realities of Texas Spanish.