885095946

3:15

7SP9. Three-dimensional geometry of the hard palate. Kenneth L. Watkin (Biomedical Eng., McGill Univ., Montreal, Quebec H3A 2B4, Ca nada)

The purpose of the present investigation was to determine the three-dimensional geometry of dental casts and in uioo measurements of the hard palate using a pulsed inductive magnetic coil sensing/digitization devicc connected to an IBM PC/AT. A specially constructed palatal tracing wand, parallel port interface, and software were developed data acquisition along with 3-D database software for real time projection. Comparisons of 3-D palatal morphology will be presented. Results will be discussed in light of current anthropomorphic Information on palatal shape and speech production models.

3:30

7SPI0. Three-dimensional reconstruction of the palate and tongue during the production of sustained vowels. Kenneth L. Watkin (Biomedical Eng., McGill Univ., Montreal, Quebec H3A 2B4, Canada)

The purpose of the present investigation was to reconstruct the three-dimensional (3-D) shape of the orał cavity during the production of sustained vowels. Using a specially developed palatal tracing wand and ultrasonic transducer localization device, real time convex phase array images of the tongue were digitized and stored on an IBM PC/AT along with the 3-D shape of the palate. The reconstructed 3-D shapes of the tongue surface relative to the 3-D shape of the palate will be presented. Data analysis strategies will be discussed along with implications for speech articulatory modeling.

3:45

7SP11. Articulatory compensation fn four-year-olds. Melanie M. Campbell (Dept. of Speech and Hearing Sci., City Univ. of New York Graduate Ctr., 33 W. 42nd St., New York, NY 10036), Richard S. McGowan, Nancy S. McGarr (Haskins Labs., New Haven, CT), and Katherine S. Harris (City Univ. of New York Graduate Ctr., NY)

The ability of six normal four-year-old children to compensate in their speech production for reduction of two types of feedback was examined. Subjects recorded repetitions of a carrier phrase containing a target word with one of thrcc vowels (/i/, /i/, /ae/) in mixed, ran-domly selected sequences in four conditions: normal, masking noise, biteblock, biteblock plus masking. The ability of each child to compensate was measured by comparison of the first three vowel formants of each test condition with those of the normal condition. Data from three subjects have been acoustically analyzed using LPC techniques. Prelim-inary results suggest that whilc four-year-old children morę easily over-come effects of noise, they do not fully compensate for presence of a biteblock. Biteblock effects appear to vary depending upon vowel height of the target. F2 values are lowered in biteblock production of /i/, whereas fl valucs are lowered in /ae/. Formant values for midvowel /i/ are less affected. Considerable individual differenccs are seen in ability to compensate for feedback reduction. (Work supported by NINCDS Grant DC00121-29]

4:00

7SP12. Perceptual eraluation of infant CV transitions. Nancye C. Roussel and Raymond G. DanilofT (Dept. of Commun. Disord., Louisiana State Univ., 163 M and DA, Baton Rouge, LA 70803)

The transition portion of the speech signal has been identified as critical to the perception of both consonants and vowels. Furui [J. Acoust. Soc. Am. 72,43-50 (1986)1 demonstrated that, for adult speak-ers, a 10-ms segment of the transition centered on the area of maximum spectra] movement contained the most critical information for joint consonant and vowel perception and that a period roughly 50 ms long

1980 J. Acoust. Soc. Am.. Vol. 89. No. 4. Pt 2, April 1991 (including the aforementioned 10-ms critical interval) was sufficient for CV syllable perception. An investigation of the perceptual cues contained within consonant-vowel (CV) transitions of infants ages 6-15 months was begun. CV-like syliables were selected from the vocaliza-tions of two infants, a małe and a female, and presented to listeners for identification. Truncated versions in which portions of the transition were systematically deleted were also presented for Identification and critical intervals for the perception of consonant, vowel, and syllable were calculated and compared with those obtained for adult speakers. Comparisons across age levels were also madę.

4:15

7SP13. Developcnental and linguistic effects on the coarticulation of frlcative-vowe! productions. .Kathleen A. Siren (Dept.. of Speech, Commun. Sci. and Theatre, St. Johns’s Univ., Jamaica. NY 11432) and Kim A. Wilcox (Univ. of Kansas, KS)

Previous work (Nittrouer, Studdert-Kennedy, and McGowan, 1989) supports the notion that children attack the acquisition of speech and languagc in larger-than-scgment units, as evidenced by greater coar-ticulatory interactions in young childrerTs speech that diminish with age. In this study, the experimental design from Nittrouer et al was expanded to include familiarity versus nonfamiliarity with a stimulus item. The notion of familiarity included two factors, meaningfulness of the stimulus item, and relative amount of motor practice. Results of acoustic measurements of children’s (3-, 5-, and 7-year-olds) and adult’s productions confirm that children do exhibit a greater effect of a following vowel (/i/ vs /u/) on the preceding fricative (/s/ vs ///) when compared to adults. In addition, relative meaningfulness of a stimulus item appears to decrease the degree of coarticulatory interac-tion between these segments, regardless of age of the individual. These results are discussed with regard to traditional models of coarticulation and traditional theories of speech and language acqui$ition.

4:30

7SP14. Acoustic typology of infant pre- and post-linguistic protosyllables. J. Piroli, R. DanilofT, and C. J. Miller (Commun. Disord., Louisiana State Univ._f 163 M and DA Bldg., Baton Rouge, LA 70803-2606)

Three hundred quasisyllabic utterances were excerpted from the vo-calizations of three infants at monthly intervals in the 4- to 14th-month epoch of deve!opment. Samples were recorded whilc at play with the mother in a toy-filled, sound-treated chamber. F\, FXfo and I& trajec-tories for the entire quasisyllab!e were extracted from digital FFT/ waveform displays. Syllable, consonant, vowel and transition durations, formant and I_Q cnvelope ve!ocities, and jitter ad shimmer measurement were madę. Natural-class transcriptions and articulatory “well-formedness" evaluations were madę for each utterance. A behavioral analysis checklist was available for each utterance. Analysis revealed that even perceptually labelable seemingly unexceptional tokens dis-played combinations of nonadult-like acoustic parameters. Tokens were sorted by clusters of variables into a typology of emergent syllabic forms based upon adult and nonadult-like valuesof the classificatory acoustic-perceptual variables. The syllable well-formedness related complexly to the infanfs attributed intent to communicate. The prelinguistic utterances were considerably morę “juvenile” acoustically than perception would lead one to bel»eve.

4:45

7SP15. Aging effects on the ability to use tern porał context in speech perception. Marta Tetzeli and Marleen T. Ochs (Vanderbilt Univ., Div. of Hearing and Speech Sci., 1114 19th Ave. S., Nashrille, TN 37212)

Elderly and young adults listened to a continuum of stimuli varying in closure duration from “rabid” to “rapid” presented at fast and normal speaking rates, with a carrier phrase and in isolation. Rates for the

121st Meeting; Acoustical Society of America 1980