Neural correlates of infant accent discrimination: an <scp>fNIRS</scp> study

Alejandrina Cristià, Yasuyo Minagawa‐Kawai, Natalia Egorova, Judit Gervain, Luca Filippin, D. Cabrol, Emmanuel DupouxDevelopmental 2014 · DOI

TL;DR

This study uses fNIRS to investigate how infants' brains process subtle differences in language, specifically distinguishing between two regional French accents. The findings reveal that by 5 months of age, infants show distinct neural activity in left-hemisphere language areas when differentiating these accents, suggesting the early development of a specialized language network.

What Problem Does This Paper Address?

How do infants learn to understand language, and when does their brain start to specialize for their native tongue? While it's known that infants become attuned to their native language's sounds, the neural mechanisms behind discriminating similar linguistic variations (like regional accents) are less understood. Previous research has shown that infants can distinguish very different languages from birth, but discriminating subtle variations requires more experience and is thought to emerge later. This paper aims to pinpoint the brain activity associated with this early accent discrimination, specifically investigating whether it relies on general auditory processing or a more specialized, left-hemisphere language network.

Methods

The researchers employed functional Near-Infrared Spectroscopy (fNIRS), a non-invasive neuroimaging technique that measures changes in blood oxygenation in the brain. This method is well-suited for studying awake, mobile infants. They used a "change detection" paradigm: infants were presented with blocks of auditory stimuli. In "Homogeneous" blocks, all speech came from talkers using the same accent (either Parisian or Quebecois French). In "Alternating" blocks, talkers switched between the two accents. Significantly greater brain activation in Alternating blocks compared to Homogeneous blocks was interpreted as evidence of accent discrimination. The fNIRS probes were strategically placed to cover bilateral perisylvian regions, areas known to be crucial for language processing.

Datasets

The study used audio-visual recordings of infant-directed speech from two Quebecois and two Parisian female talkers. These recordings featured brief passages of French. The stimuli were designed to highlight phonological differences between the accents, such as vowel diphthongization, consonant cluster simplification, and phrasal rhythm variations. No pre-existing datasets were used; the stimuli were specifically created for this experiment.

Key Insights and Ideas

  • Early specialization for subtle distinctions: The most striking finding is that 5-month-old infants, but not 3-month-olds, show neural evidence of discriminating between very similar regional accents. This suggests that by this age, infants are not just passively processing sounds but are beginning to engage specialized neural machinery for language.
  • Left-hemisphere dominance emerges: The neural activity associated with accent discrimination in 5-month-olds was primarily localized to the left perisylvian regions, including the middle superior temporal gyrus and inferior frontal gyrus. This aligns with the known lateralization of language processing in adults and points to the early development of this specialization.
  • Maturation and experience are key: The absence of discrimination in 3-month-olds, contrasted with the clear evidence in 5-month-olds, highlights the importance of both neural maturation and linguistic experience in developing the ability to parse subtle linguistic variations.

Key Results

  • 3-month-olds: No significant difference in brain activation was observed between Homogeneous and Alternating blocks in any channel for the 3-month-old group. This indicates a lack of accent discrimination at this age.
  • 5-month-olds: A significant difference was found in one channel (left deep Channel 4a) where responses were greater for Alternating blocks compared to Homogeneous blocks. This channel is thought to sample regions including the middle superior temporal gyrus and inferior frontal gyrus.
  • Age and Hemisphere Interaction: A statistical analysis confirmed a significant interaction between age group and hemisphere, indicating that the age-related differences in neural activation were more pronounced in the left hemisphere.

Limitations and Future Work

The authors acknowledge that fNIRS has limited spatial resolution, making precise localization of neural activity challenging. They also note that while the study provides evidence for accent discrimination, it doesn't definitively isolate which specific acoustic cues infants used. Future research could explore this by manipulating the stimuli (e.g., low-pass filtering audio, inverting visual signals) to neutralize certain cues and better understand the discrimination process. Additionally, the study did not investigate how infants might later assign social identity to accents, a phenomenon observed in older infants. Further work could also explore how this early discrimination ability relates to later language development and the acquisition of multiple languages.

Summary updated

Abstract

The present study investigated the neural correlates of infant discrimination of very similar linguistic varieties (Quebecois and Parisian French) using functional Near InfraRed Spectroscopy. In line with previous behavioral and electrophysiological data, there was no evidence that 3-month-olds discriminated the two regional accents, whereas 5-month-olds did, with the locus of discrimination in left anterior perisylvian regions. These neuroimaging results suggest that a developing language network relying crucially on left perisylvian cortices sustains infants' discrimination of similar linguistic varieties within this early period of infancy.
BibTeX
@inproceedings{cristia2014neural,
  author = {Alejandrina Cristià and Yasuyo Minagawa‐Kawai and Natalia Egorova and Judit Gervain and Luca Filippin and D. Cabrol and Emmanuel Dupoux},
  title = {{Neural correlates of infant accent discrimination: an <scp>fNIRS</scp> study}},
  booktitle = {Developmental},
  year = {2014},
  doi = {10.1111/desc.12160},
}
Figure