Identification of dialect cues by dyslexic and non-dyslexic listeners

Abstract

Spoken language encodes two different forms of information: linguistic (related to the message) and indexical (e.g., speaker’s age, gender, and regional dialect). However, some speech-language impairments (such as dyslexia) can reduce a listener’s ability to process both linguistic and indexical speech cues. For example, Perrachione et al. (Science, 333, 2011) demonstrated that individuals with dyslexia were less able to identify new voices than were control listeners. This study examines the ability of listeners with and without dyslexia to identify speaker dialect. Eighty listeners—40 adults and 40 children (20 in each group were dyslexic, 20 were not; 40 were males and 40 were females)—listened to a set of 80 sentences produced by English speakers from Western North Carolina or central Ohio and were asked to identify which region the speaker came from. Results demonstrated that adult listeners were significantly better at dialect identification and that listeners with dyslexia were significantly poorer at dialect identification. More notably, there was a significant age by listener group interaction—the improvement in dialect identification between adults and children was significantly smaller in listeners with dyslexia. This indicates that an initial limitation in language learning can inhibit long-term development of speaker-specific phonetic representations.