Effects of Multiple Inputs with an Accent Cue in Speech Segmentation Facilitated by Language Experiences

Past studies using artificial language speech streams showed that adults use statistics to correctly segment words. However, these studies mostly used only a single stream input and monolingual populations. Given the number of bilinguals and the prevalence of exposures to multiple languages, how do monolinguals and bilinguals compare in speech segmentation task given multiple inputs? Speech segmentation becomes challenging with multiple inputs when learners combine input across languages. The statistics of particular units that overlap different languages may change and hinder correct segmentation. Current study addresses this question by using two interleaved artificial language streams and an indexical accent cue. In the study, participants were asked to segment two artificial language streams with or without an accent cue. Our results indicated that in the absent of the accent cue, bilinguals and monolinguals performed similar, while monolinguals’ performance was weakened compared to those of bilinguals in the presence of the accent cue. Weakened performance can be due to two accounts: switching of accents and difficulty of languages. This study, expanding Weiss et al. (2009) and bilingualism research, informs us that the level of difficulty of languages and cues can play a role in segmenting multiple language streams.