A New Proposal for Phoneme Acquisition: Computing Speaker-Specific Distribution.

Mihye Choi,Mohinish Shukla

doi:10.3390/brainsci11020177

Abstract

Speech is an acoustically variable signal, and one of the sources of this variation is the presence of multiple speakers. Empirical evidence has suggested that adult listeners possess remarkably sensitive (and systematic) abilities to process speech signals, despite speaker variability. It includes not only a sensitivity to speaker-specific variation, but also an ability to utilize speaker variation with other sources of information for further processing. Recently, many studies also showed that young children seem to possess a similar capacity. This suggests continuity in the processing of speaker-dependent speech variability, and suggests that this ability could also be important for infants learning their native language. In the present paper, we review evidence for speaker variability and speech processing in adults, and speaker variability and speech processing in young children, with an emphasis on how they make use of speaker-specific information in word learning situations. Finally, we will build on these findings to make a novel proposal for the use of speaker-specific information processing in phoneme learning in infancy.

Highlights

The two studies mentioned above showed that phonetic category learning is constrained by contextual information. Armed with this evidence for infants’ sensitivity to speaker variability, we propose that infants are capable of computing speaker-specific distributions and further hypothesize that multiple relevant and contextual cues, which could make the learning situation more complicated, might instead make it easier for infants to find phonemes
There is a listener who computes whether a variety of /r/ pronunciation from one speaker is distributed unimodally or bimodally for that speaker and builds speaker-specific distributional representations
It can prevent the listener from inferring the wrong number of phonemic categories owing to speaker variability

Summary

Introduction

By 6 months of age, infants are capable of discriminating acoustic tokens corresponding to two different phonemes in a categorical manner. They exhibit such categorical perception of their native contrasts, and non-native speech contrasts. 6–8-month-olds appear to compute how frequently speech sounds are distributed (e.g., bimodally or unimodally) and categorize sounds into one or two groups based on the number of modes of the observed distribution over the input tokens While these two accounts have driven much other research in this field, there has been growing evidence that suggests the two accounts may not be sufficient. With the issues related to the current phonological acquisition hypotheses [20], we believe that this proposal will contribute to building an integrated framework to explain the mechanisms that underlie this developmental phenomenon in phonemic acquisition

Previous Hypotheses

A New Proposal

Evidence from Adults

Evidence from Toddlers

Conclusions