Training humans in non-native phoneme perception using a monkey psychoacoustic procedure.

Area of Science:

• Psycholinguistics
• Psychoacoustics
• Phonetics

Background:

• Second language acquisition often involves challenges in perceiving non-native speech sounds.
• Previous research utilized animal psychoacoustic procedures for human phoneme training.

Purpose of the Study:

• To investigate the efficacy of an animal-based psychoacoustic training procedure for improving perception of non-native phonemes in humans.
• To compare the learnability of the Japanese /r-l/ contrast by English listeners and the Middle Eastern /d-D/ contrast by Japanese listeners.

Main Methods:

• A manual left/right lever response procedure with continuous stimuli and corrective feedback was employed.
• Japanese and English listeners underwent 32 daily training sessions to categorize specific phoneme contrasts.
• Stimuli consisted of consonant-vowel syllables from multiple talkers and vowels; native contrasts served as controls.

Main Results:

• Japanese listeners demonstrated nativelike perception of the English /r-l/ contrast after training, as evidenced by accuracy, response time, and vowel context effects.
• English listeners did not achieve nativelike perception of the Middle Eastern /d-D/ contrast under the same training conditions.
• Performance measures indicated that the /r-l/ contrast is more 'robust' than the /d-D/ contrast.

Conclusions:

• The psychoacoustic training procedure is effective for learning certain non-native phonetic contrasts, like the Japanese /r-l/ for English speakers.
• The learnability of phonetic contrasts may depend on their inherent 'robustness' or susceptibility to perceptual loss.
• Findings have implications for understanding second language phonological acquisition and the nature of phonetic perception.