Mismatch Response to Native and Nonnative Vowels in Czech and Russian: Where's the Phoneme Effect?

Abstract

Purpose: The mismatch negativity (MMN), a neural index of speech sound discrimination, is reportedly stronger for phonemically relevant speech sound differences than for phonemically irrelevant differences, even if the former are acoustically smaller. Some prior studies failed to find language-specific phoneme-dependent modulation of the MMN, and only a handful of early studies tested how the MMN is affected by phoneme status versus acoustic distance. The present study tested whether the phoneme-over-acoustics effect is replicable with new sounds, new languages, and a design that considers the directionality of the contrast.

Method: Czech (n = 23) and Russian (n = 24) speakers listened passively to oddball blocks with an acoustically small Czech /i/-/ɪ/ and an acoustically larger Russian /i/-/ɨ/ contrast. MMN was calculated using two attested approaches: from physically identical stimuli across blocks and from physically different stimuli within blocks. Mixed-effects models tested whether the MMN amplitude and latency are affected by vowel contrast in interaction with language background, that is, by the language-specific phoneme status.

Results: The analyses failed to detect an interaction of vowel contrast and language background on MMN amplitude. Analyses of Bayes factors indicated a very strong support for the null interaction effect of language background by vowel contrast, that is, absence of the language-specific phoneme effect. Some directionality effects were detected in the within-block analysis.

Conclusions: The results point toward a lack of language-specific phoneme effect on the MMN amplitude. Although the literature mostly reports phoneme effects on the MMN, the present lack of an effect is in line with some prior studies that failed to find language-specific phoneme modulations of the MMN. This has implications for language learning and developmental research. Given the occasional lack of language-specific effects in the MMN of healthy adults, one should be cautious when interpreting MMN as an index of language maturation or competence in children or in atypical populations.