Understanding infant speech perception: The role of speaker variability and speaker familiarity in phoneme acquisition – A systematic review and meta-analysis

Phoneme acquisition comes with challenges, as infants are faced with enormous acoustic variability and lack of invariant features corresponding to phonemes across speakers. Nevertheless, infants acquire the phoneme inventory of their native language(s) within the first year of life through mere exposure to their native language. This implies that infants already have a mechanism in place to deal with speaker variability. To understand how this mechanism might function, we will investigate how voice information, used to distinguish speakers, influences speech perception. Previous studies have reported two different ways of how voice information influences speech perception in infants and adults: The variability benefit holds that phonemes are learned better when the training has a higher degree of speaker variability, while the familiarity benefit holds that recognizing phonemes is easier when they are uttered by familiar speakers. Thus, the familiarity and the variability benefit seem to contradict each other: Listeners learn better when there is more voice variability in the signal, but also seem to benefit from familiar voices. We here propose that these mechanisms might in fact be compatible: Speaker variability may aid category formation during phoneme acquisition, whereas speaker familiarity may rather improve online acoustic processing of phonemes. We will test this proposal in a systematic review and meta-analysis and explore whether studies confirm that variability in the signal helps phoneme generalization, whereas familiarity supports phoneme recall. Moreover, we will evaluate whether these benefits change over the life span, reviewing findings in infants, children, and adults. Our study will thus provide insight into how voice and speech information are integrated to aid phoneme acquisition in infancy, and to enable effortless speech perception in the face of speaker variability in adulthood