The effects of linguistic experience on the perception of [h

Abstract

Languages with onset‐[h] are thought to be more likely to develop coda‐[h] than languages with no [h] at all [D. Silverman, Phasing and Recoverability (Garland, New York, 1997)]. It is therefore expected that speakers of languages with onset‐[h] should be more likely to perceive coda‐[h] than speakers of a language with no [h]. The goal of this study is to test this hypothesis. Speakers of French (a language with no [h]) and English (which allows onset‐[h]) were instructed to monitor for [h] in strings of nonce words with [h] in onset or coda positions. If English speakers perform significantly better than French speakers on coda‐[h] perception, then the hypothesis is supported. Subjects also participated in a mispronunciation detection task, where they monitored for errors in words of their language. The stimuli were common English and French monosyllabic words ending in [s, f, ∫], with the fricatives spliced and replaced with [h], keeping the formant transitions intact. The stimuli were randomly placed in a string of native words with no contextual cues. If the hypothesis is supported, findings should show faster and more accurate mispronunciation detection for English speakers than French speakers, suggesting that English speakers are more likely to perceive coda‐[h].