Ultrasound and acoustic analysis of sibilant fricatives in preadolescents and adults

Abstract

This study describes the production of sibilant fricatives /s/ and /∫/, comparing Scottish English speaking preadolescent children with adults. The materials were the sequences /əCa/ and /əCi/ produced by 15 adults and 15 children aged between 10 and 12 years old. Quantitative analyses were carried out on both spectral information and on ultrasound imaging data on tongue shape, taken from nine successive time points during the fricative. The two groups of speakers were very similar to each other in the articulatory and acoustic characteristics distinguishing the two fricatives. Age-related differences in the fricative centroid measure occurred at consonant-vowel boundaries, with lower values in the preadolescents. Within-speaker variability was mostly similar across age groups, with the exception of the fricative centroid for /∫/, which was significantly more variable in preadolescents than in adults. Throughout the consonant duration, both groups consistently differentiated between the two consonants in both the fricative centroid and in one measure of tongue shape.

Using ultrasound tongue imaging to identify covert contrasts in children’s speech

“…LOC a-i , the ratio of the two perpendiculars, confirms this observation, returning higher LOC a-i values for both tokens of /ʃ/ than for the token of /s/ (the values are printed on each panel). This pattern, similar to that reported for older children and adults (Zharkova, 2016), illustrates the fact that the child raises the tongue predorsum in order to produce the required constriction for /ʃ/, as opposed to the production of /s/, where the constriction is formed by the tongue tip at the lower teeth, with no need to raise the predorsum in the context of a following low vowel. While the typically developing child differentiates between /s/ and both tokens of /ʃ/, the child who produces a dental stop for both fricatives has a somewhat higher value of the LOC a-i index for the token of /ʃ/ displayed in the right panel than for /s/, but not for the other token of /ʃ/ (middle panel), whose index value is nearly the same as for /s/ (see the bottom row of Figure 3).…”

An Ultrasound Study of the Development of Lingual Coarticulation during Childhood

“…In the right panel of Figure 2, curvature degree is a ratio of the line CD (a perpendicular from the point on the tongue curve furthest away from AB, the straight line between two ends of the tongue curve) to the line AB. Curvature degree was preferred in this study to another measure of tongue bunching, the dorsum excursion index, used in some of our previous studies (e.g., Zharkova et al, 2015b;Zharkova, 2016), because the latter measure was shown in Zharkova et al (2015a) to be affected by head-to-transducer stabilisation in the case of /ʃ/, while curvature degree had similar results for /ʃ/ across the 2 stabilisation conditions.…”

“…Specifically, the speaker needs to be able to separately control the tip/blade of the tongue, which is making the constriction for the consonant, and the rest of the tongue, which is involved in achieving the raised and advanced position required for the following vowel gesture. In consonant-/i/ syllables, substantial midsagittal tongue shape adjustments have been documented for adults and for pre-adolescent children in fricative consonants, with gradual changes towards the vowel tongue shape (Zharkova, 2016). Particularly, between mid-/s/ and the consonant offset, there was a noticeable change on a measure reflecting progressive raising and advancement of the tongue predorsum towards the high front vowel (cf.…”

“…To measure tongue differentiation, LOC a-i values were calculated for each token of /t/ from /ti/ at the end of the closure, and then compared with those from the middle of the closure. LOC a-i was chosen to quantify tongue differentiation because this measure has been previously shown to capture the raising and advancement of the tongue predorsum towards /i/ during the consonant articulated with the tip/blade of the tongue (Zharkova, 2016). Variability was measured by calculating coefficient of variation values on the 2 indices of tongue shape at mid-consonant, separately for each speaker, and including only the coefficient of variation values based on 5 repetitions of a given target, the total of 211 and 218 observations in the contexts of /a/ and /i/, respectively.…”

A Diagnostic Marker to Discriminate Childhood Apraxia of Speech From Speech Delay: I. Development and Description of the Pause Marker

“…The class of SSD termed SE includes children who have histories of feature-limited speech sound distortions, generally on the most challenging manner classes of speech sounds in a language (e.g., fricatives, affricates, liquids), which for some speakers may persist for a lifetime with or without treatment (Boyce, 2015;Flipsen, 2015;Klein, Byun, Davidson, & Grigos, 2013;Preston & Edwards, 2007, 2009Preston, Hull, & Edwards, 2013;Shriberg, Gruber, & Kwiatkowski, 1994;Sjolie, Leece, & Preston, 2016;Van Borsel, Van Rentergem, & Verhaeghe, 2007;Wren, Miller, Peters, Emond, & Roulstone, 2016;Wren, Roulstone, & Miller, 2012;Zharkova, 2016). There is research support for sociodemographic differences between children who have experienced the across-the-board, age-inappropriate deletion and substitution errors that define SD (e.g., differences in sex; Shriberg, 2010) and children whose speech errors include only the feature-limited distortion errors that define SE.…”