Phonetic convergence in temporal organization during shadowed speech

The goal of this study was to examine phonetic convergence in various measures of temporal organization during shadowed speech across different American English dialects. Participants from the Northern and Midland American English dialect regions, as well as “Mobile” participants who had lived in multiple region, read 72 sentences to establish a baseline for temporal organization, and then repeated the same 72 sentences after Northern, Midland, and Southern model talkers. Measures of temporal organization (i.e., %V, ΔC, ΔV, rPVI-C, and nPVI-V) were calculated for the read sentences, shadowed sentences, and model talker sentences. Statistical analysis of the differences in distance between the model talker sentences and the shadowers’ read and shadowed sentences, respectively, revealed significant convergence by all three shadowing groups toward the model dialects for ΔV, and significant divergence by Mobile talkers away from the model talkers for nPVI-V. Though the result of divergence by Mobile talkers was unexpected, both results provide evidence that support previous studies, which claim that social perception is a large contributing factor in convergence and divergence. These results are also consistent with previous findings demonstrating variation across dialects in temporal organization and in addition, provide evidence for variation across dialects in convergence in temporal organization.