SPEECH SEGMENTATION IS ADAPTIVE EVEN IN ADULTHOOD: ROLE OF THE LINGUISTIC ENVIRONMENT

Abstract

In this paper, we show that adult listeners who speak the same native language but live in different linguistic environments differ in their use of prosodic cues that signal word boundaries in the native language. Non-utterance-final word-final syllables have higher fundamental frequency in French. Adult native French listeners living in France or in the US completed an artificial-language segmentation task where fundamental frequency cued word-final boundaries (experimental). Other native French listeners living in France completed the corresponding task without prosodic cues (control). Results showed that France French listeners outperformed US French listeners and control French listeners, but US French listeners did not outperform control French listeners. The poorer performance of US French listeners is attributed to their regular exposure to (and thus interference from) English, a language where fundamental frequency signals word-initial boundaries. This suggests speech segmentation is adaptive, with listeners tuning in to the prosody of their linguistic environment.