Native language experience influences the topography of the mismatch negativity to speech.

Bruce D Mccandliss,Hia Datta,Jason D Zevin,Kara A Rosania,Urs Maurer

doi:10.3389/fnhum.2010.00212

Abstract

The ability to learn second language speech sound categories declines during development. We examined this phenomenon by studying the mismatch negativity (MMN) to the /r/ – /l/ distinction in native English speakers and learners of English as a second language who are native speakers of Japanese. Previous studies have suggested that the MMN is remarkably plastic when evaluated as a waveform at a central electrode. We replicated this finding: analyses of the MMN at a typical electrode location (Fz) revealed only small, non-significant differences between groups, despite large behavioral differences in the ability to discriminate these sounds from one another. Topographic analyses, however, revealed reliable differences in lateralization of the MMN, such that native English speakers’ responses were left-lateralized relative to native Japanese speakers’ responses.

Highlights

Part of learning to speak and understand one’s native language (L1) is the development of expertise in perceiving and categorizing sounds from the “phonetic inventory” of that language
A strong mismatch negativity (MMN) was observed for both groups, which was slightly larger for native English speakers, and had a slightly earlier peak for native Japanese speakers
When a more thorough topographic analyses topographic analysis was conducted, consideration of the full dense array of electrodes revealed small but consistent effects of language experience: MMN topographies suggested that the probable cortical sources for the English sounds /ra/ and /la/ were less left-lateralized for Japanese speakers than for native English speakers

Summary

Introduction

Part of learning to speak and understand one’s native language (L1) is the development of expertise in perceiving and categorizing sounds from the “phonetic inventory” of that language. Electro- and magnetoencephalographic measures of brain responses to speech have made extensive use of passive mismatch paradigms in which auditory stimuli are presented repeatedly, with one stimulus (the “standard”) having a much higher frequency of occurrence than another (the “deviant,” Winkler et al, 1990; Näätänen et al, 2001). The MMN to speech is influenced by language experience, such that responses to unfamiliar speech contrasts are weaker and less left-lateralized than responses to native speech sounds. Näätänen et al (1997) presented participants with stimuli drawn from a synthetic vowel continuum and found a larger MMNm (the equivalent of the MMN when measured with magnetoencephalography) for stimuli identified as members of different categories than for stimuli identified as members of the same category, even when the physical differences were larger for the within- than for the between-category contrast. Topographic analysis of the MMNm response revealed stronger activity on the left than the right for the native-language contrast, but a smaller difference between the two hemispheres for the non-native contrast

Methods

Results

Discussion

Conclusion