Toddlers' sensitivity to segmental and suprasegmental mispronunciations of familiar words

Recent research has shown that children as young as 19 months demonstrate graded sensitivity to mispronunciations in consonant onsets and vowels in word recognition tasks. This is evident in their progressively diminishing attention to relevant objects (e.g., a dog) as mispronunciations increasingly deviate from the correct word form (such as /dog/ changing to /gog/, /kog/, or /sog/). Despite these sensitivities, uncertainties remain about their broad generalizability, especially regarding the differences between word onsets and codas, and between lexical segmental (consonants and vowels) and supra-segmental (e.g., lexical stress and tones) elements. The present study aimed to fill these gaps. Using the intermodal preferential paradigm, we conducted two experiments to evaluate toddlers' responses to coda and lexical tone mispronunciations. Our results revealed a linear decline in toddlers' attention to familiar objects as mispronunciations became more severe, suggesting that by 19-20 months, infants' lexical representations encompass detailed phonetic information of both segmental and supra-segmental categories. Moreover, our results indicate that toddlers utilize these details in lexical processing. Such findings offer a more comprehensive understanding of the phonetic structures within toddlers' early lexical representations, sheding light on the mechanisms toddlers use in processing various word positions, across different acoustic dimensions, and in multiple languages.