Tongue Constriction Research Articles

Alveolar stops exhibit greater coarticulatory resistance than retroflexes and dentals in Malayalam

Tight coronal contrasts in Malayalam geminates (V1t̪:V2 vs. V1t:V2 vs. V1t:V2) are examined in an Ultrasound study of tongue contours to understand the nature of coarticulatory resistance and aggressiveness. Degree of Articulatory Constraint (DAC) predicts that articulatory complexity mitigates the nature of coarticulatory resistance (CR) cross-linguistically. Findings from our study of Ultrasound tongue contours are contrary to the predictions of the DAC, where the expectation is that the directionality of CR and aggressiveness will be Vt̪:V < Vt:V < Vt:V. We find that the order of CR is Vt:V < Vt̪:V < Vt:V. Greater variability in tongue constriction location in retroflex and dentals is found compared to alveolars. Low neighborhood density alveolars exhibit low contextual adjustments even at the constriction. This is a consequence of greater articulatory complexity. Malayalam alveolars are distributionally restrictive compared to dental and retroflex places of articulation. Alveolars in Malayalam have low neighborhood densities which may indeed govern coarticulatory resistance and aggressiveness of this place of articulation. These findings have been corroborated in an acoustic study of Malayalam stops (Dutta et al., 2019). We discuss the implications of our findings for DAC and propose that sparse lexical representation coerces coarticulatory resistance in tight coronal place contrasts.

Effects of vocal tract geometrical differences on flow and sound of sibilant fricatives

Sibilant fricatives are known to be produced by controlling a turbulent jet flow in a front part of the vocal tract. However, the manner of the control in each individual is still unclear. In this study, the vocal tract geometries of sibilant fricatives /s/ and /∫/ were extracted from magnetic resonance imaging (MRI) and large eddy simulation was conducted to clarify the relationship among the vocal tract geometries, flows, and acoustic characteristics. Five male Japanese subjects assumed a supine position in the MR machine, and images of the vocal tract were corrected while the subject was sustaining /s/ and /∫/. From the corrected images, computational grids were constructed, and large eddy simulation of compressible flow was applied to simulate both the turbulent flow and sound generation. Results of the simulations showed that the configuration of jet flow in each vocal tract was different, and the maximum amplitude of sound source distributed in different positions in the front part of the vocal tracts. By extracting the characteristic dimensions of each vocal tract, we found that a space between lower incisor and tongue constriction is important for the formation of individual acoustic characteristics of sibilant fricatives.

A study of high front vowels with articulatory data and acoustic simulations

The purpose of this study is to test a methodology for describing the articulation of vowels. High front vowels are a test case because some theories suggest that high front vowels have little cross-linguistic variation. Acoustic studies appear to show counterexamples to these predictions, but purely acoustic studies are difficult to interpret because of the many-to-one relation between articulation and acoustics. In this study, vocal tract dimensions, including constriction degree and position, are measured from cinéradiographic and x-ray data on high front vowels from three different languages (North American English, French, and Mandarin Chinese). Statistical comparisons find several significant articulatory differences between North American English /i/ and Mandarin Chinese and French /i/. In particular, differences in constriction degree were found, but not constriction position. Articulatory synthesis is used to model the acoustic consequences of some of the significant articulatory differences, finding that the articulatory differences may have the acoustic consequences of making the latter languages' /i/ perceptually sharper by shifting the frequencies of F(2) and F(3) upwards. In addition, the vowel /y/ has specific articulations that differ from those for /i/, including a wider tongue constriction, and substantially different acoustic sensitivity functions for F(2) and F(3).

Education in acoustics and speech science using vocal-tract models

Several vocal-tract models were reviewed, with special focus given to the sliding vocal-tract model [T. Arai, Acoust. Sci. Technol. 27(6), 384-388 (2006)]. All of the models have been shown to be excellent tools for teaching acoustics and speech science to elementary through university level students. The sliding three-tube model is based on Fant's three-tube model [G. Fant, Acoustic Theory of Speech Production (Mouton, The Hague, The Netherlands, 2006)] and consists of a long tube with a slider simulating tongue constriction. In this article, the design of the sliding vocal-tract model was reviewed. Then a science workshop was discussed where children were asked to make their own sliding vocal-tract models using simple materials. It was also discussed how the sliding vocal-tract model compares to our other vocal-tract models, emphasizing how the model can be used to instruct students at higher levels, such as undergraduate and graduate education in acoustics and speech science. Through this discussion the vocal-tract models were shown to be a powerful tool for education in acoustics and speech science for all ages of students.

The discrimination, perception, and production of two German /r/ allophones by two groups of American English speakers.

The German /r/ sound is one of the most difficult sounds for American English (AE) speakers learning German as a foreign language. Part of this difficulty may be due to its rich phonetic variation. The standard German /r/ variant [R] and dialectal variant [R′] are achieved by varying the tongue constriction degree while keeping place of articulation constant [Schiller and Mooshammer (1995)]. The close articulatory proximity of these allophones provides an opportunity for testing the relationship between perception and production in L2 sound acquisition. The aim of this study is to investigate how well experienced AE speakers and naive AE speakers can discriminate and produce the difference between the uvular fricative [R] versus the uvular trill [R′]. Two groups of AE subjects who participated in an imitation study were prompted to produce single words beginning with either [R] or [R′]. Subjects also participated in a discrimination and categorization test. Preliminary results suggest that inexperienced AE can discriminate [R] versus [R′] well. They often perceive the sounds as /h/ and are more successful at producing [R′] than [R]. Experienced speakers also discriminate the two sounds well, perceive both sounds as the German /r/ and struggle more with producing [R′] than [R].

The acoustics of gestural overlap

Acoustic modeling of speech production focuses on the effects of a single tongue constriction and a simultaneous labial constriction on the formant frequencies of an acoustic tube. But research on dynamic speech production reveals that tongue gestures overlap in time, so two or more simultaneous tongue constrictions affect the acoustic output of the tube. Understanding the acoustics of natural speech therefore requires an investigation of the acoustic effects of multiple constrictions. The goal is to generate a generalized formant nomogram that would reveal how the acoustic effects of single constrictions combine and interact in determining the acoustic output of the whole tract. Results will be presented from a finite element simulation of the acoustic effects of two constrictions that vary independently in location and size. The location of each constriction varied in 34 steps and the size varied in 16 steps. Independent variation of the two constrictions yields 295<th>936 constriction combinations. The function relating the acoustic output of the multiply constricted tract to the acoustic effects of each of the constrictions when made alone is generally nonlinear. Several features of this function will be discussed, and conditions for its linearization will also be presented. [Work supported by NIH.]

Perception of speech gestures

Perception of a synthetic French /aya/ sequence, generated by coproduction of two simultaneous gestures (tongue constriction and lip rounding), was tested in a series of experiments. The tokens consisted of eight different degrees of gesture reduction in which the target /y/ was not reached. The /ay/ trajectory in the plane F1–F2 was not rectilinear (it pointed first to /i/ before reaching /y/), having to do with the fact that the acoustic effect caused by lip rounding (lowering F2) becomes manifest later than the effect due to constriction (lowering F1 and raising F2). The /aya/ transition is, therefore, an appropriate stimulus to study the role of the linearity of trajectories in speech perception. Results show that unless /y/ is present as a steady-state vowel for at least 30 ms, the lip rounding gesture is ignored and the percept is /aia/. Thus the results strongly suggest that formant trajectories are perceived as time averages through a mechanism that calculates the direction of transition toward an (unreached) tar-get from a running average of the direction of the formant trajectory. These findings point to the importance of the trajectory direction in vowel reduction and of the time dimension in phonetic distinctions. [Work supported by NATO.]

An articulatory-acoustic-aerodynamic analysisof [s] in VCV sequences

Previous studies of the effect of vowel context on fricatives showseeming contradictions in the case of /s/: acoustic analysis shows the greatest context effect, while aerodynamic analysis shows relatively little effect, for the same subject. In this study, aerodynamic, acoustic, and articulatory data for the same subject producing /s, z/ in a variety of contexts were compared systematically. The strong acoustic effect of the /u - u/ context exists with /z/ as well as /s/, and appears to arise from a whistle-like source mechanism caused by lip rounding; the main tongue constriction does appear to be immune to vowel context. Our interpretation of aerodynamic data as constrictions in series can be generalized to include the influence of lip rounding, thus: for this speaker and for these speech-like sequences, the area of the vocal tract constriction for /s/ is independent of the vowel context but the overall aerodynamic effect does vary with lip rounding. Our aerodynamic and acoustic data seem to be consistent; both support the view that some rounding extends into the /s/ fricative.

Limitation on labial anticipation in [i] in a language with an [i]–[y] contrast

In French, the acoustic effects of anticipation of a final high front rounded vowel, [y], can be observed in a preceding [i]. This labial anticipation induces a lowering of the formants being measured (F2,F3,F4) both in the middle and, to a greater extent, at the end of the [i] in [iCny] sequences. Although significant, this contextual influence does not jeopardize the acoustic contrast between [i] and [y]. To understand the articulatory basis of these results, acoustic consequences of the gestures involved in the so-called rounding contrast between [i] and [y] are investigated through articulatory modeling. An [i]-like area function is simulated with different degrees and combinations of anticipatory lipprotrusion, reductionoftheareaatthelips, and larynxlowering. [i] allows little variation in lip gestures; beyond a limited range of coarticulation [i] shifts acoustically to the region occupied by [y]. While [i] appears to be quantal when we consider variation in tongue constriction, it shows a sensitive response to lip perturbation. In the [iCy] sequences observed, it seems that constraints on distinctiveness prevent a greater labial anticipation.

Articulatory evidence for acoustic goals for consonants

Evidence has been found for trading relations between tongue-body raising and upper lip protrusion (measured with an EMMA system) for the vowel /u/, reflecting a ‘‘motor equivalence’’ strategy that should help to constrain acoustic variation [Perkell etal., J. Acoust. Soc. Am. 93, 2948–2961 (1993)]. Theoretically, analogous relations in the transformation between the area function and the acoustic transfer function are possible for the consonants /r/ and /∫/, which are also produced with two independently controllable constrictions, formed by the tongue and by the lips. Such relations might occur more among tokens that are least prototypical, i.e., closest to perceptual boundaries. In a preliminary test of these hypotheses, a single speaker has produced the sounds /r/, /∫/, and /u/, embedded in contexts that might induce differences in prototypicality. Motor equivalence was observed for the /u/ in /tut/ (least prototypical, with the highest F2 values) but not in /pub/ or /hud/. For /r/ and /∫/, anterior displacement of the tongue constriction was positively correlated with upper lip protrusion, providing initial support for the hypothesis that movement goals for the consonants also include acoustic components, which are manifested in a tendency to maintain sufficient front cavity length to help assure acoustic-phonetic distinctiveness. [Work supported by NIDCD.]

Formant sensitivity in the vocal tract

Sensitivity of formants to perturbations of the location and degree of tongue constriction was studied. These two tongue parameters were extracted from vowel‐like midsagittal tongue shapes (N=321) generated using a tongue model [Harshman et al., J. Acoust. Soc. Am. 62, 693–707 (1977)]. The tongue shape data were arranged such that the tongue parameters vary montonically. Lip opening was varied from 0.3 to 6.0 cm2 in 0.3‐cm2 steps and the first three formants of each vocal tract were computed. Formant sensitivity was evaluated as dF/dA (Hz/mm), where dA is perturbation in a tongue parameter and dF is the corresponding formant change, and was plotted as a function of the location of tongue constriction. The results can be interpreted in terms of the formant variability, or stability, of a given vocal tract configuration. One finding is that at the tongue position associated with the vowel /u/ the formants are very stable under perturbation of the location of constriction. However, the same position is found to be the most unstable position for both F2 and F3 under perturbation of the degree of constriction. This may explain the largest F2 and F3 variations of /u/ in a set of normalized production data of vowels [Syrdal and Gopal, J. Acoust. Soc. Am. 79, 1086–1100 (1986)]. [Work supported by NIDCD, DC 00296.]

The geometric vocal tract variables controlled for vowel production: proposals for constraining acoustic-to-articulatory inversion

Studies on vocal tract shape estimation from the speech signal have thoroughly established that no unique inversion from the acoustic speech signal to the vocal tract area function is possible. However, by emphasizing the non-uniqueness of vocal tract shapes, these studies have exaggerated the possibilities of compensatory articulation, which in turn challenges the traditional classification in terms of place of articulation. We suggest, in the light of more recent studies, that this phonetic concept was too hastily discarded. These studies show that certain geometric variables are very important in the process of motor control and in the typology and description of vowel systems. The aim of this paper is to show the possibility and the value of deriving these controlled variables, rather than the entire vocal tract geometry, from acoustic parameters. For this purpose, using a statistical model of the vocal tract integrating articulatory constraints, we generated a 60 000 item look-up table of vocalic configurations. The analysis of 10 French vowels shows that, for the majority of them, it is possible to associate the formant patterns with the well known geometric parameters, namely the lip area and the location and dimension of the tongue constriction. This result seems to confirm that some geometric variables controlled during production can be inferred from the acoustic signal if articulatory constraints are taken into account in the inversion procedure.

Mechanisms for tongue-tip trills

Possible mechanisms for the production of tongue-tip trills are discussed. It is necessary that energy be transferred from the airstream to the tongue tip to enable it to vibrate. A two-mass model, similar to that of the vocal folds, is one possible model for tongue trills, but there are others that allow the tongue one-degree-of-freedom motion. It is possible that the boundary-layer separation point is affected by the acceleration of air, which is not accounted for in the quasistatic approximation. An argument will be given that the acceleration term produces a favorable energy transfer during trills. Also, the capacitive loading of the vocal tract behind the tongue constriction can provide necessary phase shifts between constriction volume velocity and intraoral pressure for favorable energy transfer. A numerical simulation of the latter model is presented with parameters derived from measurements of oral-nasal volume velocity and intraoral pressure during sustained trills. Some of the features of the trills are reproduced with this simulation.

Translating pellet positions into constriction features

Tongue constriction features can be estimated from sagittal x-ray pictures of the tongue surface and vocal tract wall. However, such records cannot be obtained in quantity, making them unsuitable for testing models such as Stevens's quantal theory. The x-ray microbeam allows larger data sets, but records flesh points rather than surfaces. This paper presents an algorithm for relating the two representations. The vocal tract wall is estimated from whole-head scans and a palate trace. Pellet positions are then “warped” into a Cartesian space where location along the tract and distance from it are the x and y values. The algorithm has been applied in a replication of Perkell and Nelson's test of quantal theory using principal component analysis. Quantal theory predicts that the pellet closest to the constriction site will show least variability, and that the most precision will be in the dimension perpendicular to the vocal tract wall for “quantal” vowels such as /i/. In the warped space, then, the principal component of variation for the relevant pellet for these vowels should be parallel to the x axis. This prediction is borne out. [Work supported by the NSF.]

Vowel constriction effects on the perception of F 1 offset and vowel duration cues for word final stop consonant voicing

The perception of voicing in final velar stop consonants was investigated by systematically varying vowel duration and change in frequency of the final F 1 transition. Several CVC continua were synthesized for each of three vowels, [i, ɪ, æ] which represent a range of tongue constriction values. Subjects used a scaling procedure in responding to these stimuli under both open and closed response set formats. The findings of this study show that both vowel duration and F 1 offset cues influence the perception of final consonant voicing, with the latter more salient in open than constricted vowels and the former more important in constricted than open vowels. In addition, individual preferences for either vowel duration or F 1 offset cues were observed, and these predilections were related to the vowel constriction feature. In summary, the results show that F 1 offset cues were perceived more invariantly as + voice than vowel duration cues across vowel contexts.

Articulatory and acoustic correlates of the [s-š] distinction

Spectra for the consonants [s] and [š] show a distinct difference in the frequency of the lowest major spectral peak. Introspection suggests that the difference is due to the creation of a space between the tongue tip and the mandibular alveolar ridge for [š] but not for [s]. Acoustical calculations show that the abrupt decrease in the natural frequency of the cavity anterior to the tongue constriction as contact between tongue tip and lower teeth is broken has a magnitude consistent with acoustical observations. To test this hypothesis concerning the distinctive differences in tongue-tip contact associated with the [s-š] contrast, observations of contact of the tongue with the lower incisors and alveolar ridge were made by looking at the change in resistance caused by bridging across two electrodes fixed to the teeth. Although the transduction technique did not permit as detailed an examination as desired, data for three subjects indicate that there is a strong contact for /s/ and little or no contact for /š/. A side result is the finding of an apparent anatomically dependent contact for front vowels versus lack of contact for back vowels. [Supported by NINCDS.]

Articulatory and acoustic properties of /s, z/ consonants produced by esophageal speakers

The objective of the present study was to identify specific articulatory maneuvers and their acoustic consequences during production of /s, z/ consonants by esophageal speakers. The consonants were embedded in CVC syllables (V = i, ae, u/) in a sentence. Information on linguapalatal contact, lip and mandible position, and speech output were collected from three fluent male alaryngeal esophageal speakers who had received no radical neck dissection. Dimensions of the anterior tongue constriction and their variations were extrapolated from the linguapalatal contact patterns and subjects' dental models. Spectra of frication noises in the frequency range up to 8 Khz were characterized by a model with a set of resonances and antiresonances. The results will be discussed with respect to known functions of speech physiology and acoustics. [Supported by NIH grant NS‐11852‐02 and by NIH grant RR‐5349 awarded to the first author.]

Acoustic analysis of low vowels in the Boston dialect

A spectrographic and computer‐aided analysis was made of the low vowels [ae] (cat), [a] (cart), and [§] (caught, cot) of a speaker of Boston English. In this dialect, [ae] has higher F2 values than [a] and [§], [a] and [§] share the same F2 range but have different F1 values: [ae] [a] [§]F1 range 465–650 670–770 465–635F2 range 1440–1745 1250–1425 1135–1410[a] and [§] are thus equally backed (same F2) in relation to [ae]. [a] and [§] are distinguished from each other by tongue height (F1) and lip‐rounding rather than by distance of tongue constriction from the glottis (F2). The Boston data thus provide counterevidence to the traditional [Principles of the IPA, 5–6 (1949)] claim that [ae] and [a] are front vowels and [§] a back vowel. Further, the data support a three‐way distinction among low vowels: [ae]‐front, [a]‐back unrounded, and [§]‐back rounded.

Development of a Quantitative Description of Vowel Articulation

A set of parameters that yield a simple yet reasonably accurate description of the articulation of vowel sounds is developed. The articulatory description is potentially useful in a speech band-width compression system based on the coding of articulatory data. The parameters give information on the position of the tongue constriction, the size of the constriction formed by the tongue, and the dimensions in the vicinity of the mouth opening. An electrical analog of the vocal tract is utilized to obtain experimental relations between the articulatory parameters and the formant frequencies. Contours of vowel articulation are derived from these data. The relation of the contours to classical phonetics is discussed.