Search

This paper presents the results of an articulatory study of palatalized consonants in Polish, a language with a typologically rare concentration of two phonemic series of posterior sibilants, one inherently palatalized, and the other contextually (allophonically) palatalized. For both phonemic and allophonic palatalization in Polish, it was found that the most stable correlates of palatalization are the advancement of the tongue root and a combined effect of raising and fronting of the tongue body. The advancement of the tongue root can be interpreted as the driving force in palatalization, while the effect of tongue body fronting and raising can be seen as secondary, resulting from the movement of the tongue root and the characteristic of the tongue as a muscular hydrostat.

In an analytic-linguistic approach to teaching segmental pronunciation and articulatory setting, teachers and voice coaches give explicit instructions on tongue placement and movements. Instructors assume that learners can do exactly as instructed. This assumption was tested in research by Wilson and Horiguchi (2012, PSLLT), who showed that phonetically untrained participants were very poor at following explicit tongue movement instructions. In their study, both the magnitude and direction of movement of the tongue’s centre of gravity were calculated from 2D ultrasound images. However, by only measuring changes in the centre of gravity, it is possible that movements were found to be smaller than they really were, especially if participants focused on the front of the tongue rather than the whole tongue body. In this study, we reanalyzed the original data, this time focusing on the surface of the tongue, rather than the centre of gravity. We made tongue traces using EdgeTrak software (Li et al., 2005), and compared them using a Smoothing-Spline ANOVA method (Davidson, 2006). Results differed from the original study, showing that participants were more conscious of what the front of the tongue was doing rather than the whole tongue body. Implications for segmental pronunciation teaching will be discussed.In an analytic-linguistic approach to teaching segmental pronunciation and articulatory setting, teachers and voice coaches give explicit instructions on tongue placement and movements. Instructors assume that learners can do exactly as instructed. This assumption was tested in research by Wilson and Horiguchi (2012, PSLLT), who showed that phonetically untrained participants were very poor at following explicit tongue movement instructions. In their study, both the magnitude and direction of movement of the tongue’s centre of gravity were calculated from 2D ultrasound images. However, by only measuring changes in the centre of gravity, it is possible that movements were found to be smaller than they really were, especially if participants focused on the front of the tongue rather than the whole tongue body. In this study, we reanalyzed the original data, this time focusing on the surface of the tongue, rather than the centre of gravity. We made tongue traces using EdgeTrak software (Li et al., 2005), and...

This study examined the contributions of the tongue tip (TT), tongue body (TB), and tongue lateral (TL) sensors in the electromagnetic articulography (EMA) measurement of American English alveolar consonants. Thirteen adults produced /ɹ/, /l/, /z/, and /d/ in /ɑCɑ/ syllables while being recorded with an EMA system. According to statistical analysis of sensor movement and the results of a machine classification experiment, the TT sensor contributed most to consonant differences, followed by TB. The TL sensor played a complementary role, particularly for distinguishing /z/.

This study shows that the retraction of the tongue tip contributes to the raising of the tongue body in the articulation of high vowels /i/ and /u/ by using qualitative simulations of tongue deformation by using a biomechanical 3D tongue model of Artisynth (Lloyd et al. 2012) and the analysis of tongue configuration in X-ray microbeam data (Westbury 1994). The simulations qualitatively replicated the tongue shapes at the temporal mid-point of /i/ and /u/ in rtMRI data by manipulating the activation values of tongue muscles in Artisynth. The results show that the tongue tip retraction helps to raise the tongue body higher in both /i/ and /u/. In the simulations of /i/, the tongue tip retraction raises the tongue body even without moving the tongue body. The results of x-ray microbeam analysis confirm those findings of simulations. The mixed linear regression model of tongue configurations in the tasks of sentence and paragraph production shows that more tongue tip retraction (a shorter horizontal distance between two anterior tongue pellets) significantly correlates with a higher position of the tongue body (higher one between two posterior tongue pellets) in both /i/ and /u/ and the correlation is stronger in /i/ than in /u/.This study shows that the retraction of the tongue tip contributes to the raising of the tongue body in the articulation of high vowels /i/ and /u/ by using qualitative simulations of tongue deformation by using a biomechanical 3D tongue model of Artisynth (Lloyd et al. 2012) and the analysis of tongue configuration in X-ray microbeam data (Westbury 1994). The simulations qualitatively replicated the tongue shapes at the temporal mid-point of /i/ and /u/ in rtMRI data by manipulating the activation values of tongue muscles in Artisynth. The results show that the tongue tip retraction helps to raise the tongue body higher in both /i/ and /u/. In the simulations of /i/, the tongue tip retraction raises the tongue body even without moving the tongue body. The results of x-ray microbeam analysis confirm those findings of simulations. The mixed linear regression model of tongue configurations in the tasks of sentence and paragraph production shows that more tongue tip retraction (a shorter horizontal distance ...

This study addresses three issues that are relevant to coarticulation theory in speech production: whether the degree of articulatory constraint model (DAC model) accounts for patterns of the directionality of tongue dorsum coarticulatory influences; the extent to which those patterns in tongue dorsum coarticulatory direction are similar to those for the tongue tip; and whether speech motor control and phonemic planning use a fixed or a context-dependent temporal window. Tongue dorsum and tongue tip movement data on vowel-to-vowel coarticulation are reported for Catalan VCV sequences with vowels /i/, /a/, and /u/, and consonants /p/, /n/, dark /l/, /s/, /S/, alveolopalatal /n/ and /k/. Electromidsagittal articulometry recordings were carried out for three speakers using the Carstens articulograph. Trajectory data are presented for the vertical dimension for the tongue dorsum, and for the horizontal dimension for tongue dorsum and tip. In agreement with predictions of the DAC model, results show that directionality patterns of tongue dorsum coarticulation can be accounted for to a large extent based on the articulatory requirements on consonantal production. While dorsals exhibit analogous trends in coarticulatory direction for all articulators and articulatory dimensions, this is mostly so for the tongue dorsum and tip along the horizontal dimension in the case of lingual fricatives and apicolaminal consonants. This finding results from different articulatory strategies: while dorsal consonants are implemented through homogeneous tongue body activation, the tongue tip and tongue dorsum act more independently for more anterior consonantal productions. Discontinuous coarticulatory effects reported in the present investigation suggest that phonemic planning is adaptative rather than context independent.

Phonological theory distinguishes nasal and oral vowel counterparts by velopharyngeal port opening, neglecting other phonetic differences between phonemic and coarticulatory nasalization. Recent articulatory work provides evidence of oropharyngeal distinctions, in addition to velic lowering. This study (12 Brazilian Portuguese speakers) uses real-time MRI to investigate oropharyngeal differences between oral, phonemically nasal, and phonetically nasalized vowels /a, i, u/. Tissue boundaries in midsagittal vocal tract images were automatically detected to reveal each vowel repetition’s aperture function. Principal Components Analysis determined vocal tract regions responsible for the greatest variance in the data. Time-dynamic analyses of vocal tract area in these regions used smoothing spline ANOVA. Results show the tongue body and/or hyperpharynx as the most important articulators. For /a/, nasal vowels demonstrate wider hyperpharyngeal and narrower tongue body regions compared to oral vowels. For /u/, o...

Korean lenis stops have been reported to become voiced intervocalically (Kagaya, 1974) and are often reduced to sonorants in casual speech with different degrees of reduction depending on the place of articulation (Lee, 1995). Velar lenis stops are more often reduced than the labials and coronals, which makes sense from the point of view of articulatory phonology (Browman and Goldstein, 1990). The oral gesture for a velar stop is on the same vocal tract tier (tongue body tier) as the neighboring vowels, so it should be more affected by overlap and blending with the vowel gestures than the closure gestures for other stops. This study investigates in what vowel environments lenis stop closure shows the most reduction. /VCV/ tokens in carrier sentences were recorded and acoustic analysis was done. The results showed that generally, reducing lenis stop closure is most in the environment of preceding and following low vowels, and least in the environment of high vowels, where the tongue body gestures of neighb...

Previous studies have suggested two different hypotheses concerning the articulation of stress contrasts in English. Examination of low vowels surrounded by labial consonants generally shows lower jaw in accented as opposed to unaccented syllables, a result originally interpreted as a lower and hence louder vowel [Edwards et al., J. Acoust. Soc. Am. 89, 369–382 (1991)]. On the other hand, a study of nonlow back vowels in alveolar contexts showed higher and backer tongue body in accented syllables, suggesting a more peripheral vowel [de Jong, J. Acoust. Soc. Am. 97, 491–504 (1995)]. This paper reexamines the two hypotheses by looking at both jaw and tongue for both phonemically low and high vowels. Five speakers of Australian English read 20 tokens of two short discourses which placed the name ‘‘Babber’’ or ‘‘Beaber’’ twice each in accented and deaccented positions. Jaw and tongue positions were recorded simultaneously using a magnetometer. On average, the jaw was lower in all accented syllables, although ...

C-V coarticulation in monosyllabic words containing initial clicks and /i:/ vowels is investigated in Mangetti Dune !Xung with 114 fps lingual ultrasound and acoustic data collected using the CHAUSA method (Miller and Finch 2011). The 114 fps rate yields an image of the tongue every 9 ms (+/-4.5 ms). Vowels following clicks have three lingual gestures involving the tongue tip/blade (TT), tongue body (TB) and tongue root (TR). TT and TB constrictions carried over from the clicks merge into a single vowel constriction at consonant specific rates. The second formant (F2) distinguishes each word type through the vowel midpoint. In regression analyses, TBCL and TRCL best predict F2 for alveolar click initial words, while TTCL best predicts F2 for dental / palatal click initial words. The more open constriction is acoustically inert. In the palatal click initial word, both constrictions are equally close for some speakers, and the gestures undergo blending (Browman and Goldstein 1990). I argue that these patter...

Magnetic resonance images of the vocal tract during sustained production of the fricatives /s, sh, f, θ, z, yog, v, edh/ by four subjects are analyzed. Measurements of vocal‐tract lengths and area functions, and morphological analyses of the vocal tract and tongue shapes for these sounds are presented. Interspeaker differences in area functions are found to be greater in the pharyngeal cavity than in the buccal cavity with the nonstrident fricatives exhibiting greater differences than the strident ones. The anterior tongue body of the alveolar stridents exhibit concave cross‐sectional shapes while that of the postalveolars show a relatively raised tongue body with flat or slightly convex cross‐sectional shapes. The concave tongue shapes of the alveolars result in a more abrupt area function behind the constriction when compared to that of the postalveolars. Laminality or apicality of articulation is found to be speaker dependent. Moreover, a greater degree of anterior medial grooving and lateral lingua‐pa...

To improve the skill of speaking a second language (L2), one good way may be to be aware of the underlying tongue position for a language. We focused on such underlying position differences between English and French, particularly when pausing for a short time between speaking; something called inter-speech posture (ISP). In past research, Wilson and Gick investigated ISP between English and French spoken by bilinguals. In that research, bilinguals had distinct articulatory settings for each language, mostly in the lips. However, their tongue data was for only 4 points of articulatory settings: distance from the ultrasound probe to tongue root, tongue dorsum, tongue body, and tongue tip, but not overall shape. Furthermore, to measure tongue tip position, past research relied on the alveolar ridge, which is unclear to see: possibly making the results inaccurate for tongue tip. In this study, we analyzed the whole shape of the tongue and made models of them using SS-ANOVA in R so that we could compare the d...

Upper Sorbian has been influenced tremendously by German, causing it to adopting the uvular rhotic. The current study examines the uvular rhotic and its palatalized counterpart in three vocalic environments (word initial, intervocalic, and word final) with the vowels /a, e, o, i/ using ultrasound. The preliminary results indicate a striking difference between the two phonemes. The plain uvular rhotic was characterized by a retracted and raised tongue dorsum, accompanied by a slightly retroflexed tongue tip. The retroflexed tongue tip is surprising for a uvular rhotic. It could be related to requirements for rhotic production or a remnant of the alveolar rhotic Upper Sorbian had historically. The environments with /i/ were especially striking: the uvular rhotic was palatalized intervocalically; the tongue body was raised near the palatal region, while the dorsum retracted strongly. In other word positions, the rhotic caused a lowering and retraction of /i/. The palatalized rhotic had a raised tongue body, ...

Troughs, which are a discontinuity in anticipatory coarticulation (Perkell 1968), have been shown to occur in the tongue body in /ipi/ and lips in /usu/. An electromagnetic articulometer (EMA) study of a single subject reported previously showed that intraoral pressure during the consonant could not be solely responsible for the downwards tongue body movement observed during the labial consonants (/p, b, m, f, v/); results supported the hypothesis of a secondary tongue gesture during the consonant. Lip troughs during /s/ in rounded contexts were more ambiguous, possibly because of inconsistencies in corpus design. Six new subjects have been recorded using a WAVE EMA system and a revised corpus. Tongue trough magnitudes decreased in the order /f, v/ > /p, b/ > /m/, serving to disprove the aerodynamic hypothesis; they were deeper for long than short consonants (e.g., [iffi] > [ifi]), and converged to the same point in asymmetric vowel contexts. The same type of convergence during the consonant was observed ...

The production of consonants of the same place but different manner of articulation can involve certain adjustments in the posture of the tongue shape. This can be due to requirements for specific gestures (e.g., lowering the tongue sides for laterals) or constraints on coordination of different gestures (e.g., the tongue-palate constriction and the velum lowering for nasals). This study used ultrasound imaging to examine sagittal tongue shape differences in the production of Kannada (Dravidian) laterals, nasals, and stops of two places of articulation—alveolar/dental and retroflex. Words with these consonants (as geminates) were produced multiple times by five female and five male native speakers of Kannada. The analysis of tongue shapes revealed a lower tongue body/blade for laterals than stops, but only in retroflexes. The opposite was observed for /l/ vs. /t/, likely reflecting the alveolar vs. dental constriction differences. The nasals were produced with a significantly more advanced tongue body tha...

This paper examined the articulatory properties of high vowels [i], [y], and [u] in Mandarin produced by four Taiwanese Mandarin native speakers and four English-speaking Chinese learners (L2 learners) with an Electromagnetic Articulagroph AG500. The articulatory positions of the tongue top (TT), the tongue body (TB), the tongue dorsal (TD), and the lips were investigated. The TT, TB, and TD of [y] produced by the L2 learner were further back than that by the native. In addition, the TD of [y] by the L2 learners was higher than the native. Further comparison found that the tongue positions of [y] was similar to [u] in L2 production. Regarding to the lip positions, the [y] and [u] were more protruded that [i] in the native production, while there is no difference among these three vowels in the L2 production. The findings suggested that most of the L2 learner were not aware that the lingual target [y] should be very similar to [i] but the lip articulators of [y] are more protruded than [i]. Some L2 learner...

A common simplifying technique in ultrasound studies of variation is to select a single representative frame for each token, sacrificing dynamic information that is often critical for understanding the phonetic motivations of phonological phenomena. We examine the phonetic motivations for tongue body raising in English /ae/ tensing (e.g., Labov et al. 2005) in 23 North American English speakers using phonetically meaningful time-varying articulatory signals extracted directly from ultrasound video. An articulatory measure of /ae/ tenseness is generated using regression to find the linear combination of articulatory principal components (found using EigenTongue Feature Extraction; Hueber et al. 2007) that best accounts for the F2-F1 difference in front vowels. We have previously shown (Carignan et al. 2015) that tensing before /m n/ involves tongue body raising that is timed to the vowel nucleus, whereas tensing before /ɡ ŋ/ involves anticipating the velar closure to different degrees in different dialects. ...

Whalen et al.’s (2010) ultrasound study has suggested that constriction degree “may be the best descriptor for height” in American English front vowels {i, ɪ, e, ɛ}. The present study took a further step, investigating the case of height and backness distinction in Taiwanese front and back vowels. Stimuli consisted in /CV/ sequences with V corresponding to one of the six vowels {i, e, a, ɔ, ɣ, u} (Southern variety). Five adult speakers were asked to repeat each sequence six times in random order. The articulatory data were collected using the Carstens AG500 EMA system. Our results show that vowel height and backness in Taiwanese vowels can be reliably differentiated as follows: (1) for front and central vowels, height is primarily distinguished by Tongue Body Constriction Degree (TBCD) (i>e>a) (less robustly by Tongue Tip Constriction Degree), whereas height distinction of back vowels is made in Tongue Dorsum Constriction Degree (TDCD) (u>ɣ>ɔ), (2) backness can be distinguished by Tongue Tip and Tongue Body Constriction Location. Therefore, Whalen et al.’s (2010) claim can be extended to back vowels as well (at least in Taiwanese). Furthermore, our results are compatible with Ladefoged’s (2001) “holistic” view of vowel height as long as his “highest point of the tongue” can be “locally” interpreted: TBCD for height of front and central vowels and TDCD for back vowels. In sum, the present study suggests that distinction in vowel height be relativized to different gestural constriction degree.

To help understanding tongue shape and motions, a physical figure model of tongue muscles using viscoelastic material of urethane rubber gel were made by improving previous models. Compare to previous shape tongue models that had been made and presented, the new model is constructed from tongue body (consisting of Transversus linguae, Verticalis linguae, Longitudinalis linguae superior, and Longitudinalis linguae inferior), and individual extrinsic tongue muscles (consisting of Genioglossus anterior, Genio glossus posterior, Hyoglossus, Styloglossus, and Palatoglossus) parts. Therefore, each muscle’s shape, starting and ending points, and relation to other muscles and organs inside mouth are more understandable than previous ones. As the model is made from viscoelastic material similar to human skin, reshaping and moving tongue are possible by pulling or pushing some parts of the tongue muscle by hand, that is, tongue shape and motion simulations by hand can be done. The proposed model is useful for speech science education or a future speaking robot using realistic speech mechanism.

This study examined tongue movements in consonant-vowel-consonant sequences drawn from real words in phrases as produced by 36 older children (three male and three female talkers at each age from 10 to 15 years) and 36 adults. Movements of four points on the tongue were tracked at 400 Hz using the Wave Electromagnetic Speech Research System (NDI, Waterloo, ON, CA). The four points were tongue tip (TT; 1 cm from tip on midline), tongue body (TB; 3 cm from tip on midline), tongue right (TR; 2 cm from tip on right lateral edge), and tongue left (TR; 2 cm from tip on left lateral edge). The phrases produced included the vowels /i/, /I/, /ae/, and /u/ in words (i.e., “see,” sit,” cat,” and “zoo”). Movement measures included 3D distance, peak and average speed, and duration of vowel opening and closing strokes. The horizontal curvature of the tongue was calculated at the trajectory speed minimum associated with the vowel production using a least-squares quadratic fit of the TR, TB, and TL positional coordinates...

It has been argued that Kazakh primarily distinguishes its anterior ("front") vowels from its posterior ("back") vowels through retraction of the tongue root. This analysis is at odds with the traditional assumption that the anteriority of Kazakh vowels is contrasted by tongue body position. The present study uses ultrasound imaging to investigate the extent to which the position of the tongue root and the tongue body are involved in the anteriority contrast in Kazakh. Native speakers of Kazakh were recorded reading words (in carrier sentences) containing target vowels, which were controlled for adjacent consonants and metrical position. An audio recording was also made of these sessions. Frames containing productions of the target vowels were extracted from the ultrasound video and the imaged surface of the tongue was manually traced. Analyses of tongue root and body position were analyzed for each vowel and will be presented together with formant measurements from the audio recordings.

The goals of this study were to characterize tongue surface displacement during production of bilabial stops and to refine current estimates of vocal-tract wall impedance using direct measurements of displacement in the vocal tract during closure. In addition, evidence was obtained to test the competing claims of passive and active enlargement of the vocal tract during voicing. Tongue displacement was measured and tongue compliance was estimated in four subjects during production of /aba/ and /apa/. All subjects showed more tongue displacement during /aba/ than during /apa/, even though peak intraoral pressure is lower for /aba/. In consequence, compliance estimates were much higher for /aba/, ranging from 5.1 to 8.5 x 10(-5) cm3/dyn. Compliance values for /apa/ ranged from 0.8 to 2.3 x 10(-5) cm3/dyn for the tongue body, and 0.52 x 10(-5) for the single tongue tip point that was measured. From combined analyses of tongue displacement and intraoral pressure waveforms for one subject, it was concluded that smaller tongue displacements for /p/ than for /b/ may be due to active stiffening of the tongue during /p/, or to intentional relaxation of tongue muscles during /b/ (in conjunction with active tongue displacement during /b/).

Strident fricatives in Mandarin may be followed by fricative vowels, which are made with the tongue body in essentially the same position as in the corresponding fricative (Ladefoged and Maddieson, 1996). Specifically, the vowel [i] may occur only after the palatalized fricative, the apical vowel after the alveolar fricative, and the retroflex vowel after the palatoalveolar fricative. Given the complementary distribution, the three phonetically distinct vowels are generally considered allophones. Under the view that the implementation of discretely specified distinctive features may be enhanced by gradient, non‐contrastive articulatory gestures to maintain perceptual distinctiveness (Keyser and Stevens, 2003), it is proposed that the three vowels are derived from an underlying [i] and are modified to enhance the defining articulatory and acoustic attributes of the fricatives. Acoustic data from six speakers indicate that the defining acoustic attributes for the fricatives, i.e., spectral prominences in discrete frequency ranges, are associated with distinct tongue configurations. The articulatory and acoustic properties of the palatalized fricative is most compatible with those of [i], whereas the tongue body backing gesture in the alveolar and palatoalveolar fricatives are consistent with lowered F2 frequency in the apical and retroflex vowels. [Work supported by NIH.]

The present study investigates the articulatory and acoustic properties of apical vowels in Mandarin Chinese using ultrasound and acoustic measures. Instead of the high front vowel, dental and retroflex sibilants are known to be followed by apical vowels made with tongue tip/blade, rather than the tongue body. The ultrasound images of one Beijing speaker confirmed that the tongue tip/blade gesture indeed remains unreleased during the production of these vowels. However, it also revealed that the tongue body for apical vowels is significantly retracted with a relative order of the vowel following dentals being more retracted than the vowel following retroflexes. This gave rise to a consistent acoustic consequence; the second formant of the vowel after dentals of six Mandarin speakers was significantly lower than that of the vowel after retroflexes. In addition, the third formant of the vowel after a retroflex was significantly lowered and nearly merged with the second formant, indicating that the vowel is ...

This study examined the hypothesis of whether vowel‐to‐vowel coarticulation effects are stronger when a particular vowel segment is under focus in clarified speech than when it is not under a condition of focus. In one condition, native speakers of American English produced disyllabic word pairs that varied in the quality of their first vowel (stressed V1), while in the second condition, the same group of speakers produced the word pairs that varied in the quality of their second vowel (V2). The degree of coarticulation was determined by the amount of spectral and/or articulatory displacement that occurred during the production of V2‐position [schwa] due to the influence of V1. Due to the need to increase the likelihood of coarticulatory effects during speakers’ productions, recording trials were sped up to a conversational speaking rate. Spectral measurements (first and second formants) of ə in both conditions were taken, as well as tongue‐surface position measurements pertaining to the position of tongue body during the production of ə under the two conditions. Preliminary acoustic analysis indicates that speakers produced V2‐position ə with greater coarticulatory effects in the V1‐clarifying condition than in the V2‐clarifying condition.

This study examined lip and tongue coordination and tongue movement kinematics in Italian long and short consonants with native speakers of Italian as subjects. The linguistic material consisted of Italian word pairs that only differed in the duration of a labial or lingual consonant, which was either long or short. Recordings were made of lip and tongue movements using a magnetometer system. There was a very robust difference in the duration of the oral closure. The duration of the tongue body movement between the two vowels in a sequence of vowel‐bilabial nasal‐vowel was longer, and its movement path longer, for a long than for a short consonant; the average speed of the movement showed no reliable difference. The onset of the tongue movement relative to the oral closure started earlier for a short than for a long consonant. In lingual stop and fricative consonants, the tongue movement during the closure was longer, and its average speed slower, in the long than in the short consonants. Overall, these results for Italian are comparable to those found in long and short Japanese consonants. In particular, they provide additional evidence for task‐specific modulation of tongue movement speed. [Work supported by NIH.]

This paper reports speech articulator gestures in Swedish, Bulgarian, and Greenlandic to study universal and language‐specific components in articulation. This work is based on analysis of movements of individual articulators from x‐ray motion films of speech, and continues from previous reports from this study by Wood [J. Phon. 7, 25–43 (1979); 19, 281–292 (1991); Proc. 3rd Conf. I.C.P.L.A., 191–200, Helsinki (1994); Proc. 13th I.C.Ph.S. Vol. 1, 392–395, Stockholm (1995); J. Phon. 24, 139–164 (1996); Proc. 4th Speech Prod. Sem., 61–64, Grenoble (1996); Speech Commun., (in press)], and by Wood and Pettersson [Folia Ling. 22, 239–262 (1988)]. Common principles concerned utilization and integration of articulator gestures (articulator gestures executed in approach, hold, and withdrawal phases, four tongue body gestures used, all gestures available for vowels and consonants, gesture conflicts resolved by gesture queuing, coarticulation and assimilations implemented by coproduction). Language‐specific princip...

This paper presents electro‐magnetic articulography (EMA) data on the four coronal stops of Arrernte, an Australian language. The stops are: the lamino‐dental ‘‘th,’’ the apico‐alveolar ‘‘t,’’ the apico‐postalveolar (or ‘‘retroflex’’) ‘‘rt,’’ and the lamino‐palatal ‘‘ty.’’ Jaw, tongue tip (TT), and tongue body (TB) data were collected for two female speakers of the language. Results for the first speaker show a fronted tongue position for the laminal consonants, with the TT reflecting a similar location for both the dental and the palatal. However, for the palatal, the TB position is much higher, whereas for the dental, the TB is very low. For the apical consonants, the TT is not as far forward, and the TB is not quite as high as for the lamino‐palatal. For both TT and TB, apico‐postalveolar is further back than apico‐alveolar. For the second speaker, the TT sensor failed, but in line with the first speaker, the TB sensor showed a higher position for the palatal. The other stops were lower and more forward, with the postalveolar TB position higher than the laminal or alveolar stop position. For both speakers, the jaw position is lowest for the postalveolar. [Work supported by Australian Research Council and NIH: NIDCD.]

The idea of a language‐specific articulatory setting (AS), an underlying posture of the articulators during speech, has existed for centuries [Laver, Historiogr. Ling. 5 (1978)], but until recently it had eluded direct measurement. In an analysis of x‐ray movies of French and English monolingual speakers, Gick et al. [Phonetica (in press)] link AS to inter‐speech posture, allowing measurement of AS without interference from segmental targets during speech, and they give quantitative evidence showing AS to be language‐specific. In the present study, ultrasound and Optotrak are used to investigate whether bilingual English‐French speakers have two ASs, and whether this varies depending on the mode (monolingual or bilingual) these speakers are in. Specifically, for inter‐speech posture of the lips, lip aperture and protrusion are measured using Optotrak. For inter‐speech posture of the tongue, tongue root retraction, tongue body and tongue tip height are measured using optically‐corrected ultrasound. Segmental context is balanced across the two languages ensuring that the sets of sounds before and after an inter‐speech posture are consistent across languages. By testing bilingual speakers, vocal tract morphology across languages is controlled for. Results have implications for L2 acquisition, specifically the teaching and acquisition of pronunciation.

The relations between jaw and tongue movements were examined using words of the form sVd for eight different speakers from the X‐Ray Microbeam Speech Production Database. Measurements were examined at the maximum speeds during the release and during the closure of the tongue blade and tongue body. For nonhigh vowels the tongue blade traveled in the same direction as the jaw during release, and, to a lesser extent, the same was true during closure. Further, the magnitude of the projection of the tongue blade velocity onto the direction of the jaw movement was often large compared with the speed of the jaw. There was less consistency in the relation between tongue body and jaw movement. These results indicate that the jaw and tongue movements are not rigidly coupled. Rather the jaw, which can provide a hard boundary for the tongue, is getting out of the tongue’s way during release and following the tongue on closure for subsequent bracing. [Work supported by Grant NIDCD‐01247 to CReSS LLC.]

Speakers exhibit systematic errors when asked to imitate self‐produced vowels [B. Tuller and G. Vallabha, J. Acoust. Soc. Am. 107, 2855 (2000)]. These results were replicated (3 speakers, 45 systematically distributed self‐produced vowel targets, 10 imitations per target) and compared to a simulation of the effects of articulatory and perceptual noise. Articulatory noise was modeled using the Haskins Articulatory Synthesizer. The tongue body center (TBC) was interpolated between the configurations for /i/, /a/, and /u/ and 90 of the intermediate configurations were perturbed by adding independent Gaussian noise (sd=1 mm) to the x and y coordinates of the TBC. Perceptual noise was modeled by (1) transforming each of the 90 intermediate vowels to F1/F2 bark (‘‘perceptual’’) space and (2) adding independent Gaussian noise (sd=vowel discrimination limen=0.28 barks) to each bark coordinate. The models did not match the directions of the imitation errors made by the speakers. However, the distribution of the er...

Acoustic measurements have been made on stop consonants produced by several normal and dysarthric speakers. The acoustic data were previously recorded by H. P. Chang and H. Chen at MIT. In the present study, various aspects of production following release of the oral closure were quantified through the use of acoustic measures such as spectra and durations of noise bursts and aspiration noise, as well as shifts in frequencies of spectral prominences. Through comparison of these measurements from the normal and dysarthric speech, and based upon models of stop‐consonant production, inferences were drawn regarding articulator placement, rate of articulator release, tongue‐body movements, and vocal‐fold state. The dysarthric speakers deviated from normals particularly with respect to alveolar constriction location, rate of release, and tongue‐body movement into the following vowel. For example, the lowest front‐cavity resonance in the burst spectrum of an alveolar stop is normally in the range 3500–5500 Hz. For three of eight dysarthric speakers, the range was lowered to 1500–2800 Hz, indicating either placement of the tongue tip further back on the palate or formation of the constriction with the tongue body in a location similar to that of a velar stop. [Work supported in part by NIH Grant DC00075.]

The lateral consonant in English is generally produced with a backed tongue body, a midline closure of the tongue blade at the alveolar ridge, and a path around one or both of the lateral edges of the tongue blade. In prevocalic lateral consonants, the release of the closure causes a discontinuity in the spectral characteristics of the sound. Past attempts to synthesize syllable‐initial lateral consonants using formant changes alone to represent the discontinuity have not been entirely satisfactory. Data from prior research has shown rapid changes in not only the formant frequencies but also in the glottal source amplitude and spectrum and in the amplitudes of the formant peaks at the consonant release. Further measurements have been made on additional utterances, guided by models of lateral production. Based on these data, new synthesis attempts have incorporated changes in source amplitudes, formant bandwidths, and the location of a pole‐zero pair. Including these additional parameters improves the naturalness of the synthesized lateral‐vowel syllables in initial perception tests. Results of listening tests indicating the relative importance of the various factors to the naturalness of the synthesized syllables will be reported. [Work supported in part by a LeBel Fellowship and by NIH Grant No. DC00075.]

Earlier work has shown that the lips are moving at a high velocity at the instant of oral closure for bilabial stop consonants, resulting in tissue compression. The lips may thus have virtual targets that would require them to move beyond each other. The present study examines events at the oral closure for stops produced with the tongue. Tongue movements were recorded using a magnetometer system. Four subjects participated and produced ten tokens each of VCV sequences with all possible combinations of the three vowels /i, a, u/ and the consonants /t, d, k, g/. Consistent with the results for bilabial stops, the tongue was moving at a high velocity at the instant of oral closure. The tongue movement trajectories were more complex with a larger horizontal component than the lips. For velar stops, the tongue body was usually moving forward at stop closure, with the velocity influenced by vowel context. For alveolar stops, the tongue tip horizontal movement direction depended on the preceding vowel, but also differed across subjects. Between the onset and release of the oral closure, the tongue tip and tongue body moved through a trajectory of usually less than 1 cm. [Work supported by NIH.]

In the past, much articulatory movement data have been obtained using the x‐ray microbeam. Currently, however, magnetometer systems such as EMMA are becoming predominant. A unique data set, addressing concerns of consistency between these instruments is reported. A single speaker was recorded with both instruments reading the same utterances: ‘‘It’s a [pV’CVp] again.’’ Data from two x‐ray microbeam (XRMB) runs were collected on the same day (pellets re‐placed for the second run); 27 months later the parallel EMMA data were collected. Vertical movement of the lower lip, tongue tip, and tongue body during the VCV’s was analyzed. Extrema positions (amplitudes) and distance and time between these positions (gestural displacements and durations) were obtained. The results demonstrate that the correlation between EMMA and XRMB runs is very high, almost as high as that between the two XRMB runs. Using further analyses of variance it was concluded confidently that the results obtained using EMMA for articulatory ...

Studies of vowel‐to‐vowel coarticulation in Russian have generally found vowels in VCV sequences to have little influence on each other’s articulation. This has been attributed to Russian’s contrastive palatalization, which is assumed to block coarticulation by preventing the pharynx and tongue body from moving smoothly from one vowel configuration to the next during stop closure. If this is correct, coarticulation should be minimal even for reduced vowels. This hypothesis was tested by examining coarticulation in the nonfront reduced vowels of Russian. Here, F1 and F2 were measured at the onset, center, and offset of five repetitions of prestress /o/, poststress /o/, and stressed /a/ produced by native speakers in /tVb——b(j)V‘‘tilded el’’/ and /pVd—— d(j)V‘‘tilted el’’/ nonsense words. Posttarget consonants were plain or palatalized; flanking vowels were /i,a,u/, symmetrical. All speakers showed substantial coarticulation throughout the poststress target, during the first half of prestress target, and in the onset of stressed target. These results suggest that contrastive palatalization does not preclude significant vowel‐to‐vowel coarticulation.

Previous work [Sands, J. Acoust. Soc. Am. Suppl. 1 86, S123 (1989)] reported that the acoustic properties of the primary release in clicks was not affected by the height or backness of the following vowel, due to the requirements on the tongue body during the production of the click. This paper will show that the velar or secondary closure also exhibits no such coarticulatory effects, unlike pulmonic velar consonants. Words containing dental, lateral, and alveolopalatal voiceless unaspirated clicks before the vowels /i, e, a, o, u/ from seven Xhosa speakers were digitized at 20 kHz. The noise of the primary click release was removed in order to isolate the release of the secondary closure. The edited tokens were presented in a fixed choice task to six trained phoneticians who evaluated the place of articulation of the prevocalic segment. The main vowel effect seen is that front vowels tend to favor a uvular interpretation. It will be argued that this is not a coarticulatory effect, but rather that listene...

By means of pellet‐tracking technique using an x‐ray microbeam system, observations of articulatory movements for a larget set of Japanese nonsense words containing various types of CV syllables, vowel sequences have been conducted. In particular, characteristics of mutual influences of the vowel and consonant articulations upon each other were investigated. It was observed that (1) in V1CV2 articulations, relative timing of consonant articulations and V1−V2 transitions varied depending on the type of consonant; and (2) for the vowel articulation in CVC, the degree of perturbation of tongue tip position for /t/, tongue body position for /k/, and jaw position for /p/ was approximately same, whereas the perturbation of lip position for /p/ was smaller than that of jaw position. Coordinations of the movements of the velum, tongue, and lip were also investigated with special reference to the difference in the syllable initial and syllable‐final nasal consonants in Japanese. Basic set of data on English is als...

The positioning of the tongue body is strongly influenced by coarticulation effects. The tongue body contact position for /k/ and /g/ varies especially widely, moving in the direction demanded by the surrounding phonemes. Coarticulation effects of post‐consonantal vowels on the place of tongue body contact for Japanese velar stop consonants were observed by dynamic velography. The dynamic velography system, developed by the authors is a 36‐electrode electropalatographic technique for observing lingual contact with the soft palate. The resulting velograms can be compared with spectrograms made from simultaneous recordings of speech. Three Japanese male subjects produced VCV utterances with stress on the second syllable. The velogram patterns for [aka], [ako], [aga], [ago] were similar, with contact located at the posterior part of the soft palate. Contact for [ake], [aki], [age], [agi] was located more anteriorly. Thus, as suggested by previous findings, the tongue contact position for Japanese /k/ and /g/ is strongly influenced by the vowels following it.

Examination of the vocal‐tract configurations that are available for the production of vowels, and of the properties of the sounds produced by these configurations, suggests that certain configurations are preferred in the sense that they give rise to sound outputs that are distinctive and whose properties are relatively insensitive to perturbations in the articulatory representation. It is shown that anterior‐posterior movement of the tongue body leads to two such preferred configurations independent of tongue height‐configurations corresponding to the traditional phonetic feature [± back]. Variation in tongue height is postulated to produce three distinctive positions achieved through movements of the tongue and the mandible: tongue maximally raised toward the hard or soft palate, tongue lowered but with some palate contact maintained with the edges of the tongue, and severed contact of the tongue with the palate. Shifts from one of these positions to another produces well‐defined contrasts in the acoustic output. Quantal adjustments leading to distinctive acoustic contrasts can also be made in the lips and tongue root, Implications of this point of view in the formulation of systems for the classification of vowels are discussed. [Supported in part by an NIH Grant.]

The purpose of this paper is to shed further light on the aetiology and speech specific nature of vowel intrinsic fundamental frequency (IF0), by investigating its presence in vedic chanting. This is one of the earliest forms of chanting (circa 5000 years old) which is still prevalent today, predating much of Indian classical music. It is interesting to study vedic chanting as it represents the borderline between speech and singing. There are three distinct tone levels in this chanting referred to as: (a) udatta high, (b) anudatta low, and (c) svarita, in‐between. Sanskrit grammarians classify vowels according to the position of the tongue body and use the same labels as above for marking the tones. The interaction of vowel height and tone level in this vocal mode is investigated. Several vedic passages from three chanters with different vowels occurring with the same tone level and different tone levels for the same vowel have been recorded. Initial measurements of F0 suggest that chanters can override the effect of IF0 to achieve the same relative pitch values. This implies that the effect of IF0 depends on vocal mode and an active control of IF0 is possible.

Kewley‐Port et al. [“Perception of static and dynamic acoustic cues to place of articulation in initial stop consonants,” J. Acoust. Soc. Am. 73, 1779–1783 (1983)] showed that listeners can reliably identify the place of articulation of initial bilabial and alveolar stops in stop‐vowel syllables with severely truncated vowels from just the first 20 ms of the syllable, which includes the burst and the beginning of the formant transitions. However, they showed that listeners must hear substantially more of the formant transitions, 40 ms or more, to identify the stop's place of articulation when the stop is velar. They speculate that their listeners' need for a longer transition may reflect a slower change in formant frequencies observed overall for velars compared to stops at other places of articulation. Slower formant frequency changes for velars probably arise from the greater inertia of the tongue body compared to the tongue tip or lip, as do the longer voice‐onset‐times (VOTs) observed for velars versus bilabials or alveolars. If a listener has other (spectral) cues that a stop is a velar, then he will accept stimuli with longer VOTs as voiced than if the other cues indicate another place of articulation. Though the longer VOTs are not themselves cues that the stop is a velar, they may enhance that percept [cf. K. Stevens et al. “Toward a phonetic and phonological theory of redundant features,” in Invariance and Variability in Speech Processes, edited by J. Perkell and D. Klatt (Erlbaum, Hillsdale, NJ, 1986), pp. 426–449]. Again assuming that other cues to place are present, slower transitions may enhance the identification of a stop as velar in a similar way, but could have either a negligible or deleterious effect on identifying a stop as bilabial or alveolar. In this study, ABX‐format discrimination experiments are being conducted in which listeners hear short (100 ms) CV syllables in which the second and third formant transitions match natural values for the syllables [ba, da, ga, bi, di, gi] in direction and magnitude, while their rate of change is varied in 15‐ms steps across a range of 15–90 ms. The stop bursts have been excised from these stimuli so the only cues to place are in the formant transitions. It is expected that listeners' ability to discriminate place of articulation to degrade in two directions, with artificially short transitions for velars and artificially long transitions for nonvelars.

A set of data of Japanese CV syllables has been obtained for two native speakers of the Tokyo dialect by use of the computer‐controlled x‐ray microbeam system [J. Acoust. Soc. Am. 57, 1516–1520 (1975)]. The material included /pV tV kV sV bV dV gV zV/, where V is /i, e, a, o or u/, repeated in immediate succession. For one subject, five pellets were placed on the tongue surface in the midsagittal plane, with another pellet on a lower tooth. The six pellets were tracked at a rate of 120 sample sets per second. The data are being analyzed, and tentative conclusions include: (1) the blade of the tongue (in reference to flesh points) is more advanced for [e] than for [i], (2) k‐g and t‐d oppositions reveal different tongue gestures, (3) for dental consonants, the tongue body makes a characteristic forward movement, and these movements are slower for fricatives than for stops. A slow motion picture of the movements of the pellets will be demonstrated.

Several researchers have noted that a child's vocal tract is not simply a smaller version of an adult man's vocal tract, but that the ratio of pharynx‐cavity length to oral‐cavity length is substantially different. In order to study the acoustical implications of this difference, a static articulatory vocal‐tract model is being developed with dimensions that can be varied for male and female speakers over all age ranges. The model has parameters specifying the position of an elliptical tongue body relative to the angle of the jaw, similar to the model developed by Mermelstein for adult men [P. Mermelstein, J. Acoust. Soc. Am. 53, 1070–1082 (1973)]. Other parameters specify positions of tongue tip, lips, and hyoid bone. The sizes of the various parts of the vocal tract are specified by growth curves which express anatomical distances as a function of age. The growth curves were obtained by performing least‐squares fitting of two‐ through six‐parameter curves to measurements reported in the medical literature. This research has implications for a better understanding of infant phonetic capabilities, articulatory configurations used by women and children, and vowel normalization. [Work supported in part by the National Institutes of Health.]

The reason why, other things being equal, the pitch of vowels correlates with vowel “height,” has long been a subject of speculation. The effect is usually attributed either to acoustic coupling between vocal tract and vocal cords or to the pull of the tongue on the vocal cords. To determine which of these hypothesized factors is responsible we observed the effect on the pitch of vowels when one of these factors (impedance of the vocal tract was held constant, while the other (tongue stretch) was varied over an exaggerated range, viz., by having eight speakers utter special test utterances when their jaw was propped open by small wooden blocks ⩽ 10 mm thick. (Lindblom and Sundborg have shown that under such conditions apeakers can achieve acoustically normal vowels by increasing the distortion—and presumably the stretch—of the tongue body with respect to the jaw and larynx‐hyoid complex.) We found a small but consistent increase in the pitch interval between high and low vowels when subjects spoke with pr...

The articulatory behavior of the tongue body was observed and related to the phonemic representation of speech through the use of cineradiography at 100 frames/sec. Specific points on the tongue‐body surface were identified by the attachment of small radiopaque markers. Points on the jaw and each lip were also marked. The motion of these markers was observed during the production of nonsense trisyllables constructed from /i/, /ɑ/, /u/, /b/, /g/ in the form V1 CV2 CV1, with various conditions of stress as produced by a single speaker. Target positions were identified for each point marker corresponding to each phoneme. The transitions between target positions were observed. It was determined that the principal motion between vowel target positions could be characterized by a single characteristic transition function. Deviations from this principal intertarget motion that appear to be related to perturbations in oral pressure were observed.

The observations reported in this paper pertain to the articulatory movements—especially tongue movements—that are associated with the American English diphthongs /al/, /ɔI/, and /aU/. Flesh points on the lingual surface were simulated by attaching small radiopaque markers to the midline of the tongue body. Lateral‐view cinefluorography was used to record the displacement patterns of these markers during the production of diphthongs in isolated words and short sentences. Data were collected for three speech samples, each of which was read by a different set of two speakers. Consideration of these data emphasizes the general problem of identifying invariant phonemic correlates at the articulatory level, with illustrations of the variability in (1) the initial and terminal vocal‐tract configurations for a given diphthong, (2) the patterns of tongue‐point movement for a given diphthong, and (3) the tongue‐point velocities during a given diphthong. [Research supported in part by grants from NIH.]

The object of this study was to characterize the movements of the tongue body during consonant articulations by the displacement patterns of individual flesh points. Radiopaque markers were placed on the tongue body, and the displacement patterns of these markers during symmetric VCV utterances were recorded by means of cinefluorography. Two speakers recited utterances that were constructed with the vowels /i,u,a/ and the lingual consonants /g,j,d,z/. Additional data are presented for two other speakers to demonstrate the presence of tongue movement during symmetric VCV utterances in which the intervocalic consonant is the nonlingual fricative /v/. The discussion includes comments on consonant‐vowel interaction, remarks on certain peculiarities of the tongue‐point movements, and some speculations as to active and passive forces during lingual articulation. [Research supported in part by grants from NIH.]

EMG recordings from several muscles controlling tongue body movement for various VCV and VCCV utterances were examined to uncover the factors influencing the extent and direction of the coarticulatory reorganization of motor commands for speech gestures, and to reveal physiological correlates of syllable boundaries. The primary determiner of the direction of the reorganization appears to be the individual muscle's function in increasing or decreasing vocal tract constriction (closure and release). Muscles producing greater constrictions show anticipatory coarticulation while those producing greater openings show postcipatory coarticulation. Since consonants are produced with greater vocal tract constriction than are vowels, they limit the extent and direction of coarticulation. Thus, reorganization at the motor command level is shown for both anticipatory and postcipatory coarticulation. Syllabic units appear as VC sequences for muscles constricting the vocal tract and CV sequences for muscles expanding t...