Foreign accent identification, prototypicality, and lectometric methods

Abstract

Lects constitute prototype categories (Gitte Kristiansen 2003). This implies that central/prototypical speakers are more easily identified as members of a specific lect (Kristiansen et al. 2018). The prototypicality of elements within natural categories have generally been measured through direct questions, indirect questions, or reaction time. Nonetheless, abstract categories (e.g., foreign accents) may be difficult to examine via questions. Following previous research that analyzed the use of the Levenshtein Distance (LD) (Vladimir I. Levenshtein 1965) to predict foreign-accentedness (Martijn Benjamin Wieling et al. 2014) or the intelligibility of foreign accents (Jurado-Bravo 2021), this study explores the use of LD as a predictor of prototypicality of Spanish-accented English. The LD between 50 Spanish speakers of English and different prototype benchmarks were calculated. These recordings were used as speech stimuli in an accent identification test. Reaction time measures were collected and correlated to the calculated LD. Results suggest that the LD to the stereotypical prototype can partly predict the prototypicality of Spanish-accented English.