Non-native phonetic learning is destabilized by exposure to phonological variability before and after training

Area of Science:

• Second language acquisition
• Phonetics and phonology
• Cognitive psychology

Background:

• Phonological variability is often used in phonetic training to improve non-native speech sound perception.
• The universal benefit of variability, especially during testing and sleep consolidation, remains under investigation.

Purpose of the Study:

• To investigate the impact of phonological variability during testing on non-native phonetic learning.
• To examine the interaction between variability and sleep-mediated consolidation in second language phonetics.

Main Methods:

• Participants were trained on a non-native vowel contrast in one context (/u/).
• Two groups were tested: one on the trained context, the other on both trained (/u/) and untrained (/i/) contexts.
• Perceptual accuracy and overnight changes were assessed.

Main Results:

• Participants tested with variability (trained and untrained contexts) showed reduced accuracy in perceiving the trained speech sounds.
• This group also exhibited no significant improvement in performance after a period of sleep.
• Conversely, the group tested only on the trained context may have benefited from consolidation.

Conclusions:

• Introducing phonological variability, even during the testing phase, can negatively impact phonetic learning.
• Variability during testing may interfere with memory consolidation processes, hindering performance gains.
• Learners may benefit from focused testing on trained contrasts to facilitate sleep-dependent improvements in non-native phonetic perception.