Toward articulatory-acoustic models for liquid approximants based on MRI and EPG data. Part II. The rhotics

Abstract

Magnetic resonance images of the vocal tract during sustained production of [symbol: see text] by four native American English talkers are employed for measuring vocal-tract dimensions and for morphological analysis of the 3D vocal tract and tongue shapes. Electropalatography contact profiles are used for studying inter- and intra-talker variabilities. The vocal tract during the production of [symbol: see text] appears to be characterized by three cavities due to the presence of two supraglottal constrictions: the primary one in the oral cavity, and a secondary one in the pharyngeal cavity. All subjects show a large volume anterior to the oral constriction, which results from an inward-drawn tongue body, an anterior tongue body that is characterized by convex cross sections, and a concave posterior tongue body shape. Inter-subject variabilities are observed in the oral-constriction location and the way the constriction is formed. No systematic differences are found between the 3-D vocal tract and tongue shapes of word-initial and syllabic [symbol: see text]s. Tongue-shaping mechanisms for these sounds and their acoustic implications are discussed.

Sex-specific fundamental and formant frequency patterns in a cross-sectional study

“…The distinctive 'dip' or lowered F3 values for the rhotacized vowel is likely to be the result of similar articulatory configurations that are typically reported for /‰/ (e.g. Alwan et al, 1997;Espy-Wilson et al, 2000) and replicates previously reported speaker sex differences in adults for the formant frequency values of /‰/ (Westbury et al, 1998).…”

“…Alwan et al, 1997;Espy-Wilson et al, 1997, and the creation of a sublingual cavity anterior between the palatal and lip constriction (e.g. Espy-Wilson and Boyce, 1999).…”

“…A number of acoustic-phonetic dimensions were investigated for both the peripheral (aa, ae, ao, eh, iy and uw) and non-peripheral (ah, er, ih, uh) vowels as a function of age group and sex. Details and the motivation for these examined (Traunmüller, 1988 Alwan et al, 1997;Dalston, 1975;Espy-Wilson et al, 2000). The range in formant frequencies were therefore examined for F1 to F3.…”

Articulation strategies for English liquids used by Japanese speakers

“…Speakers employ a wide variety of articulations that result in lowered F 3 [e.g. [8][9][10][11]. Variation in articulation strategies extends even to Japanese English speakers, as Masaki et al [12] discovered in an MRI study, finding 4 distinct strategies for differentiating /ô/ and /l/ employed by both native English speakers and Japanese learners.…”

A method of tooth superimposition on MRI data for accurate measurement of vocal tract shape and dimensions

“…Engwall [7] built polygon mesh data of the teeth and palate using a dental cast for the same purpose. Narayanan et al [8,9] and Alwan et al [10] measured casts using calipers, and sliced dental impressions to trace the outlines of the slices. This method was also adopted by Espy-Wilson et al [11].…”