2009/05/14

Li, F., Edwards, J., & Beckman, M. E. (2009). Contrast and covert contrast: The phonetic development of voiceless sibilant fricatives in English and Japanese toddlers. Journal of Phonetics, 37, 111–124.

Presentation: Sarah

Summary: Shelly

Previous studies have found that pronouncing sibilant fricatives are difficult for both English and Japanese young children. Acquisition of these fricatives is usually protracted, and children usually have a hard time distinguishing them. For example, at the age of three, English children often substitute /ʃ/ with /s/, which is a “fronting” error, while Japanese children substitute /s/ with /ɕ/, which is a “backing” error. However, it is possible that children actually differentiate these sounds by making covert contrasts, which is not perceivable by adults’ ears. Therefore, the present study measured acoustically children’s production of these sibilants, aiming to see whether there existed covert contrasts. For both English and Japanese groups, ten children under three and five adults were recruited. Stimuli were words with the target sibilants at initial position, and the variety of vowels following the sibilants was balanced. Both children and adults were asked to repeat after female native speakers saying those stimuli, and their production was recorded and transcribed by trained phoneticians, who were also native speakers. Results of transcription analysis showed that fronting errors (/ʃ/ ® /s/) predominated in English children, while backing errors (/s/ ® /ɕ/) predominated in Japanese, which corresponded to the findings of previous studies. However, as they further examined some acoustic parameters of these sibilants (F₂ frequency at the onset of the following vowel, and also fricative spectrum moments, including centroid frequency, standard deviation, skewness, and kurtosis), it was found that children who seemed to merge the two sibilants actually contrasted them in a couple of the above cues, and the reason why adult transcribers did not recognize these contrasts may be because the cues used by these children were different from those used by adults. Analyses on adults’ production showed that English adults differentiated /ʃ/ and /s/ by centroid frequency, and Japanese differentiate /s/ and /ɕ/ by onset F₂ frequency as well as centroid frequency, while error-producing children were found to use some other minor cues, such as skewness or kurtosis, which were covert to adults. From the above results, one can conclude that there indeed exist covert contrasts in children’s production. How these covert contrasts are perceived is worth studying in future study.