Locus Equations As an Index of Coarticulation for Place of Articulation Distinctions in Children

Locus equations were investigated as a phonetic index for children's production of stop + vowel tokens. Locus equations are straight-line regression fits to data points formed by plotting onsets of F2 transitions along the ordinate and their corresponding midvowel nuclei along the abscissa. Such functions for adult speech have been found to be extremely linear with slope and y-intercept values contrastively distinctive across place of articulation. Sixteen children, aged 3–5 years, produced /bVt/, /dVt/, and /gVt/ tokens embedded in a carder phrase and repeated in randomized order a minimum of three times. Six medial vowel contexts were used [i, I, ae, , a, u]. Both individual and group mean scatterplots were extremely linear and highly remniscent of adult prototypes. While labial and velar slopes exhibited some degree of overlap, labial versus alveolar and alveolar versus velar slopes were significantly different. All y-intercepts as a function of place of articulation were significantly different. Compared to adult norms, intersubject variability of slope and y-intercept ranges were greater for children. Locus equations can provide a phonetic descriptor for a child's attainment of stop place categories seeking to achieve the adult standard of a balance between coarticulatory adjustments and contrastive distinctiveness.

KEY WORDS: coarticulation, phonetics, formant, F2 transitions, place of articulaton

Submitted on July 18, 1991
Accepted on December 30, 1991

CiteULike    Connotea    Del.icio.us    Digg    Facebook    Reddit    Technorati    Twitter    What's this?