Past research has demonstrated that infants recognize familiar voices from a very early age (DeCasper & Fifer, 1980). However, very little is known about infants' ability to discriminate unfamiliar voices. In the current study, we test infants' ability to discriminate voices in their native language as well as a non-native language. Three experiments using the Visual Fixation Procedure were carried out with Dutch-learning 7- month-olds. In the first two experiments, infants were habituated to three female voices in one language and then tested on a new female voice in the habituated language (Voice Change) and a new female voice in a new language (Language Change). In Experiment 1 (Dutch versus Japanese) and Experiment 2 (Japanese versus Italian), infants dishabituated to Language Change trials. But only Dutch-habituated infants in Experiment 1 dishabituated to Voice Change trials. These results suggest that infants may be more sensitive to voice changes in a familiar language than voice changes in an unfamiliar language. This language familiarity effect could be driven by attention and/or perceptual learning factors. However, an alternative explanation for these results is that the Dutch voices used in Experiment 1 were acoustically more distinct than the Italian and Japanese voices used in Experiments 1 and 2. In order to test this possibility, a third experiment was carried out using backwards speech. Reversing speech retains many acoustic factors that are important for voice identification (fundamental frequency, speech rate, breathiness, etc), but destroys the language-specific prosody that is so important for infants' ability to recognize a language as familiar (Van Lancker, Kreiman, & Emmorey, 1985). Thus, by repeating Experiment 1 with backwards speech, we were able to test the possibility that the infants in Experiment 1 discriminated the Dutch voices more readily than the Japanese voices simply because the Dutch voices were more distinct than the Japanese voices. If this were the case, then we expected the infants in Experiment 3 to dishabituate to Voice Change trials just as the infants in Experiment 1 did. However, the infants tested with backwards speech failed to dishabituate to Voice Change trials or Language Change trials. In combination, these results constitute the first evidence that voice discrimination in infants is affected by language familiarity, a finding with important implications for the study of voice encoding and identification.
Cite as: Johnson, E., Westrek, E., Nazzi, T. (2005) The role of language familiarity on early voice discrimination. Proc. ISCA Workshop on Plasticity in Speech Perception (PSP 2005), 227-230
@inproceedings{johnson05b_psp,
author={Elizabeth Johnson and Ellen Westrek and Thierry Nazzi},
title={{The role of language familiarity on early voice discrimination}},
year=2005,
booktitle={Proc. ISCA Workshop on Plasticity in Speech Perception (PSP 2005)},
pages={227--230}
}