Consonant Duration Research Articles

Durational cues for word segmentation Dutch

Word segmentation is a prerequisite for word recognition and subsequent comprehension of a connected speech utterance. Linguistic and contextual information provides insufficient means to this end. In this paper, the contribution of acoustic-phonetic word boundary markers to perceived word segmentation is investigated. In the first experiment, phonetically ambiguous word combinations were presented; linguistic and contextual information could not contribute to word segmentation of these stimuli. Yet, the observed 80% accuracy shows that phonetic boundary cues are used effectively. Post-hoc acoustic analysis suggests that three durational parameters may affect perceived segmentation, viz. pre-boundary vowel decay time, intervocalic consonant duration, and post-boundary vowel rise time. In the second experiment, the latter two of these parameters were manipulated. Results show a significant effect of consonant duration on perceived segmentation, but no effect of vowel rise time. Uncontrolled boundary markers (presumably, boundary segments and allophonic variants) also provided positive boundary cues for /C ≠ V/ context. In conclusion, some word boundary markers contribute to the perceived word segmentation. Nevertheless, results also suggest that non-sensory (linguistic and contextual) information is more important for retrieving the intended word segmentation.

Acoustic and perceptual characteristics of voicing in fricatives and fricative clusters.

Several types of measurements were made to determine the acoustic characteristics that distinguish between voiced and voiceless fricatives in various phonetic environments. The selection of measurements was based on a theoretical analysis that indicated the acoustic and aerodynamic attributes at the boundaries between fricatives and vowels. As expected, glottal vibration extended over a longer time in the obstruent interval for voiced fricatives than for voiceless fricatives, and there were more extensive transitions of the first formant adjacent to voiced fricatives than for the voiceless cognates. When two fricatives with different voicing were adjacent, there were substantial modifications of these acoustic attributes, particularly for the syllable-final fricative. In some cases, these modifications leads to complete assimilation of the voicing feature. Several perceptual studies with synthetic vowel-consonant-vowel stimuli and with edited natural stimuli examined the role of consonant duration, extent and location of glottal vibration, and extent of formant transitions on the identification of the voicing characteristics of fricatives. The perceptual results were in general consistent with the acoustic observations and with expectations based on the theoretical model. The results suggest that listeners base their voicing judgments of intervocalic fricatives on an assessment of the time interval in the fricative during which there is no glottal vibration. This time interval must exceed about 60 ms if the fricative is to be judged as voiceless, except that a small correction to this threshold is applied depending on the extent to which the first-formant transitions are truncated at the consonant boundaries.

Syllable internal timing: Effects of speaking rate, sentence position, and focal stress.

The duration of a vowel can be affected by speaking rate, sentence position, and focal stress. Having confirmed the independent effects of these factors on vowel duration, the goal of this study was twofold: (1) to describe their effects on the timing of consonants neighboring the affected vowel, and (2) to characterize the internal timing of the syllable when these factors are combined in the same sentence. Conversations were developed in which target CVCs occurred in initial or final sentence position and were either focused or nonfocused by the discourse. Twelve subjects produced each conversation at three speaking rates. The results indicate that (1) like vowels, consonant durations are affected by speaking rate, sentence position, and focal stress, and (2) when converging on a syllable, the investigated factors influence the syllable’s internal timing independently. These findings suggest that speaking rate, sentence position, and focal stress each globally affect the internal timing of a syllable, with the effects of each factor being superimposed on one another within flexible upper and lower limits.

The role of remote rate information and relative acoustic resemblance on phonetic representations.

Miller and Liberman [Percept. Psychophys. 25, 457–465 (1979)] demonstrated that listeners use vowel duration information when making the /ba/–/wa/ distinction. They also found that adding a final consonant to the syllables shifted the /ba/–/wa/ boundary. However, they did not examine the effects of changing the final consonant duration. Furthermore, by using stimuli that were voiced throughout, the potential effects of acoustic and auditory similarity between the initial consonant distinction and the succeeding phonemes could not be examined. In this experiment CVC syllables were used to examine the effect of this later-occurring speaking rate information on an initial /sh/–/ch/ distinction. The syllables ended with /s/, and both vowel and final fricative duration were varied independently. This allowed one to examine the effect of acoustic similarity on the coding of rate information. The relevance of these results to the auditory and/or phonetic coding of speaking rate information will be discussed. [Work supported by NIDCD Grant No. DC00219 to SUNY at Buffalo.]

Speech timing strategies in elderly adults

This study investigated two speech timing patterns in healthy elderly adults. Earlier research reported that young adults adjust consonant duration according to: (1) utterance length and (2) spatial distance between the consonant articulation and adjacent vowel target. Specifically, fricative durations decreased with increases in utterance length and fricative durations were found to be longer in final vowel /i/ environments than /α/ environments. DiSimoni (1974 a - c ) found that the first of these patterns emerged in children between three and six years of age but the second, " lineal compensation", was not yet statistically evident even in his nine-year-old subjects. The present investigation examined the speech of ten elderly subjects (67-81 years) for the presence of these temporal patterns; the results were compared to 21-28-year-old controls. Although overall speech rate of the older adults was significantly slower, older and younger subject groups both exhibited the two " compensatory" effects under investigation, supporting the view that speech timing control may be somewhat resistant to effects of aging. Maintenance of anticipatory processing by older subjects may depend on slowed rate as a compensatory mechanism for reduced structural and physiological capabilities. Alternatively, increasing tendencies toward caution and concern with precision may account for slowed rate as well as preserved temporal efficiency.

The influence of consonant duration on vowel‐to‐vowel coarticulation in Korean

This study examines the influence of durational differences associated with the three manner types (unaspirated, aspirated, fortis) on the coarticulatory effects of one vowel on another across an intervening consonant in Korean. F2 and consonant durations were measured in V1CV2 tokens read by five speakers, where Vn = {/i/,/a/} and C = [(bilabial, dentoalveolar, palatoalveolar, velar) × (aspirated, unaspirated, fortis)]. Repeated measures ANOVA tests show that Korean exhibits both anticipatory and carryover vowel‐to‐vowel coarticulation. There is also anticipatory coarticulation associated with the consonant's manner type, confined predominantly to /a/. The data show greater magnitude differences on V1 = /a/ when the consonant is unaspirated (224 Hz) than aspirated (143 Hz) or fortis (127 Hz). There is no statistically significant coarticulation between manner type and V2. Durationally, unaspirated C's are approximately half as long as the aspirated and fortis C's. The aspirated and fortis C's are durationally distinguished by differences in VOT. These observations will be discussed in the context of dual channel and single channel models of consonant and vowel production.

Timing of upper lip protrusion gestures for the vowel /u/

Timing of upper lip protrusion gestures and accompanying acoustic events was examined for multiple repetitions of word pairs such as “lee coot” and “leaked coot” for three speakers of American English. The duration of the intervocalic consonant string was manipulated by using various combinations of/s,t,k,h,♯/. Pairwise comparisons were made of times of acoustic /i/ offset to acoustic /u/ onset (consonant string duration), protrusion onset to acoustic /u/ onset (onset interval), maximum acceleration to acoustic /u/ onset (acceleration interval), and acoustic /u/ onset to protrusion offset (offset interval). In spite of considerable token-to-token and cross-speaker variation, several general observations were made. There were some consonant-specific effects, primarily for /s/. The non-s subset evidenced two modes of behavior. (1) The lip protrusion gesture for /u/ had a relatively invariant duration, but its timing varied with respect to the oral consonant gesture complex: The longer the consonant string, the earlier the lip protrusion gesture, or (2) the protrusion gesture duration correlated positively with consonant duration. The mode 1, which predominated, the slope of the timing relationship between oral and labial gestures differed across subjects. [Work supported by NIDCD.]

The articulation of consonant-induced vowel duration changes in English

This reports on an articulatory investigation of acoustic findings that vowels tend to be longer before voiced than voiceless consonants [House and Fairbanks, J. Acoust. Soc. Am. (1953)], before fricatives than stops [Umeda, J. Acoust. Soc. Am. (1975)], and before single consonants than before multiple consonants [Fowler, J. Exp. Psychol. (1983)]. Measurements were taken from tongue movement traces obtained with the Wisconsin microbeam system during vowels in accented and unaccented conditions. Assuming a gestural theory of speech production, three mechanisms for producing these duration differences are considered: The duration of the closing gesture may be stretched, the consonantal gesture may begin later in the vowel, or the production of the entire vowel may be lengthened. One subject's vowel durations were affected by the voicing and manner of the following consonant. Another's were affected only by consonant voicing when the word is accented, and only by consonant duration when the word is unaccented. Moreover, the effect of duration and voicing on the vowel duration was due, in part, to differences in how early the closing gesture for the consonant begins, while the effect of manner of articulation was due entirely to differences in the duration of the closing gesture. [Work supported, in part, by NSF.]

Perception of consonant length: voiceless stops in Turkish and Bengali

The goal of the present research was to investigate the extent to which acoustic cues other than the silent gap corresponding to closure duration might figure in the discrimination of geminate and non-geminate voiceless stops. The method, adapted from earlier studies on duration cues, was to create two sets of stimuli with the length of the silent interval varying incrementally between that of a non-geminate and that of a geminate. One set was made by artificially lengthening the silent interval of an original non-geminate in 10 ms steps up to the length of a geminate, and the other set was made by shortening an original geminate in the same manner. Starting with recorded minimal word pairs with geminate and non-geminate stops in Turkish and Bengali, sets of stimuli were constructed as described above. These stimuli were presented to native speakers of the respective languages in a word identification task, and the results were charted to see whether the identification curves were the same for the original geminates as for the original non-geminates. The result was that the curves did differ, the original geminates being identified as geminates slightly more frequently than original non-geminates at closure durations between 120 and 160 ms. The difference was statistically significant for at least some points on the curve. However, in contrast to earlier studies on consonant duration contrasts, the displacements found were less than 10 ms on the time axis, and the region of significant difference between the identification curves confined entirely to a segment of the time continuum in which no naturally occurring stimuli are found. Our conclusion is that in actual speech recognition there is no evidence that cues other than closure duration play a role in the discrimination of geminate and non-geminate stops in these languages.

The relationship of stuttering severity and treatment length to temporal measures of stutterers' perceptually fluent speech.

The purpose of this study was to investigate the relationship between stuttering severity and treatment length to temporal measures of stutterers' perceptually fluent speech. Seventeen adult stutterers (4 mild, 8 moderate, and 5 severe) were divided into two treatment categories. One group with limited exposure to treatment (n = 7) consisted of stutterers who had completed 0-10 hr of treatment. Another group with an extensive period of treatment (n = 10) consisted of stutterers who had completed 47-212 hr of treatment. All subjects were asked to repeat a simple phrase and read a prose passage. Spectrographic measures of the stutterers' voice onset times, vowel, consonant, and phrase durations were obtained from perceptibly fluent productions of the isolated phrase and from a phrase extracted from the oral reading. Results showed that there were no significant differences among stuttering severity groups or the two groups with different lengths of treatment. These results suggest that changes in temporal measures of stutterers' fluency might occur through the process of treatment, but those changes are not related to stuttering severity or length of treatment.

On the psychoacoustic nature of the P-center phenomenon

It seems almost generally accepted that the P-center, defined as the psychological moment of occurrence of a syllable, is only determined by the duration of the syllable’s consonantal onset and of its rhyme, independently of one another and independent of their phonetic categorization. Summing up the results of a number of our own experiments—already described elsewhere—it will be shown that in contrast to this view (1) the effect of prevocalic consonant duration is not independent of vowel duration and vice versa and that both effects are non-linear, (2) the syllable’s rhyme does not affect its P-center as a unit, but that there are also different effects of vowel and final consonant duration, again interacting with one another, and (3) phonetically different syllables behave differently and that P-center shifts can also be induced by non-temporal variations. An elaboration of a purely psychoacoustic model—proposed elsewhere—based on the time course of the loudness within each critical band [sone/Bark] can account for all these effects fairly well. Furthermore, this model distinguishes between “points of onset” and “temporal centers of gravity”, which both seem to be psychologically relevant factors in experimental tasks involving subjectively regular rhythm. Finally the P-center phenomenon is discussed in connection with the predictions of this perceptual model and with current models of articulatory timing.

The effect of position of stress on the durations of stop consonants in American English

It has been documented for American English that /t/ and /d/ are realized as very short alveolars (flaps) following a stressed vowel nonfoot initially: [V_V(…)] or [VCV_V(…)] [Zue and Laferriere (1979)]. This study investigates the duration of other stop consonants in the flapping environment to see whether the shortening of /t/ and /d/ is a phonologization of a more general phonetic principle. Stathopoulos and Weismer (1983) report that in nonsense words, stop closure durations for all places of articulation are shorter in the post‐stress environment. In a study of two speakers using natural speech, one speaker's consonants shortened significantly (p<0.05) in the flapping environment. However, the fact that the second speaker did not exhibit the same shortening pattern suggests the need to expand the number of speakers; results from such a larger subject pool will be reported.

Effect of vowel duration on the perception of syllable‐initial /∫/ and /t∫/

The duration of an adjacent vowel has been demonstrated to affect the judgment of consonant duration and, hence, phonetic identity. For example, syllable‐initial stops can be distinguished from glides on the basis of transition duration and, when the following vowel is relatively long, longer transitions are required in order for a glide to be perceived. Apparently, the vowel provides a source of durational contrast, whereby a longer vowel makes the adjacent consonant seem shorter. This effect has been demonstrated for both speech and nonspeech signals [L. Diehl and M. A. Walsh, J. Acoust. Soc. Am. Suppl. 1 80, S125 (1986)], and is presumably grounded in a general auditory mechanism. In this study, the effect of vowel duration on perception of syllable‐initial consonants is also demonstrated for fricatives and affricates. Frication duration is an important cue to this distinction, with longer durations yielding more /∫/ responses. Two series of edited natural tokens of /∫a/ and /t∫a/ were created in which frication duration varied from 100 to 210 ms. The duration of the following vowel was 204 ms for one series, and 358 ms for the other. With the longer vowel context, significantly longer frication was required to yield a /∫/ percept. Consistent with earlier findings for other consonantal distinctions, durational contrast occurred between the vowel and fricative. [Work supported by NINCDS.]

Evaluation of two speech enhancement techniques to improve intelligibility for hearing-impaired adults.

This study reports the effects on speech intelligibility of two types of digital speech processing: amplitude enhancement of consonants to produce near-zero consonant/vowel intensity ratios and increased duration of consonants to provide an additional 30 ms of sound. Compensatory vowel shortening accompanied duration increases to maintain original overall duration. One hundred words from the California Consonant Test (CCT) list were recorded. The target consonant in each word was processed in four ways to produce tape recordings of (a) unenhanced speech, (b) speech with increased consonant amplitude, (c) speech with increased consonant duration, and (d) speech with increased consonant amplitude and duration. The tapes were played to a group of 20 listeners with moderate sensorineural hearing loss at 65 dB SPL and to another group of 10 such listeners at 95 dB SPL. Results indicated that amplitude processing was associated with 10% to 12% improvement in intelligibility at the lower level but failed to yield any significant effect at the high level of presentation. Increasing consonant duration, by contrast, provided no benefit at the low level but gave modest benefit (5%) at 95 dB. Further development of the speech-processing techniques is necessary before they can be incorporated into a useful hearing aid.

Stress effects on stop consonant closure and release gestures

This study seeks to characterize the articulatory kinematics of stop consonant gestures in contrasting environments, with the aim of incorporating the results into the gesture‐based approach to synthesis under development at Haskins. The CVCVC tokens drawn from the Bell x‐ray micro‐beam corpus have been examined, representing combinations of stress, vowel, and consonant type spoken in isolation by a female speaker of English. Measurements were taken from peaks of the vertical component of pellet displacement and velocity corresponding to consonant closure and release gestures, using the lower lip pellet for labials and the tongue root pellet for velars. The ANOVA results show that stress affects consonant gesture duration independently of coproduced vowel type, but interacts significantly with consonant type, possibly because velars compete with coproduced vowels for control of the tongue. Results also showed different results for consonant duration depending on whether the gesture occurred in the first or final syllable, which is consistent with previous studies of prepausal lengthening [D. H. Klatt, J. Phonet. 3, 129–140 (1975)]. [Work supported by NSF Grant BNS‐8520709.]

Variability differences in the speech production in two age groups of adult females

The purpose of this study was to examine the variability between the speech produced by two age groups of adult females, a young adult group and a geriatric group. The 25 women in each group phonated three productions of the vowel /a/ for 5 s at three effort levels—maximum, comfortable, and minimum. Recordings of these productions were digitized and analyzed for mean intensity and intensity distribution. The women also repeated a series of CV, VCV, and VC syllables using the consonants /p/, /b/, /t/, /d/, /s/, and /z/ combined with the vowel /a/, in the carrier phrase “Speak_____again.” Intraoral air pressure and the voicing signal were recorded to provide vowel and consonant durations, voice onset times, and peak intraoral air pressure. The F‐max test for homogeniety of variance was used to compare the variability between the two age groups. The older speakers exhibited significantly greater variability than the younger speakers for the dependent variables, peak intraoral air pressure, consonant duration, vowel duration, voice onset time, and the minimum effort level of speech intensity. The greater variability of production exhibited by the older speakers could be a factor in the perception of older speakers.

Interacting effects of preceding stress and phonemic vowel quantity on consonant duration in Finnish

Finnish has a quantity opposition in both vowels and consonants, and demarcative stress on the initial syllable. Stressed vowels are longer than corresponding unstressed ones. A peculiar fact about short and long vowels is that they are significantly longer after a short initial syllable [CV.CV(V)] than after any kind of long initial syllable [CVV.CV(V), CVC.CV(V), CVVC.CV(V)]. A proposed explanation is that duration as a stress cue spills over to the second vowel in CV.CV(V) words in order to avoid neutralizing the phonemic vowel quantity on the first syllable. If this indeed is the source of the second‐vowel lengthening, then we would expect a similar effect in syllable final consonants: The consonant after a short stressed vowel (CVC.) should be longer than the consonant after a long stressed vowel (CVVC.). The duration of geminates was studied in four word structures, CVC.CV(V) and CVVC.CV(V), and they were found to be longer in the former group than in the latter.

Effects of acoustic modification on consonant recognition by elderly hearing-impaired subjects.

In a recent study [S. Gordon-Salant, J. Acoust. Soc. Am. 80, 1599-1607 (1986)], young and elderly normal-hearing listeners demonstrated significant improvements in consonant-vowel (CV) recognition with acoustic modification of the speech signal incorporating increments in the consonant-vowel ratio (CVR). Acoustic modification of consonant duration failed to enhance performance. The present study investigated whether consonant recognition deficits of elderly hearing-impaired listeners would be reduced by these acoustic modifications, as well as by increases in speech level. Performance of elderly hearing-impaired listeners with gradually sloping and sharply sloping sensorineural hearing losses was compared to performance of elderly normal-threshold listeners (reported previously) for recognition of a variety of nonsense syllable stimuli. These stimuli included unmodified CVs, CVs with increases in CVR, CVs with increases in consonant duration, and CVs with increases in both CVR and consonant duration. Stimuli were presented at each of two speech levels with a background of noise. Results obtained from the hearing-impaired listeners agreed with those observed previously from normal-hearing listeners. Differences in performance between the three subject groups as a function of level were observed also.

Effects of bite block speech on intrinsic segment duration.

The fact that low vowels tend to have longer intrinsic duration than high vowels is sometimes attributed to mechanical characteristics associated with the greater distance the jaw travels in producing low vowels. To investigate this tendency, 6 adult speakers of English each produced 72 [h]-vowel-consonant stimuli containing high and low vowels under 'normal' conditions and with their jaw fixed. If duration differences commonly observed between high and low vowels are a result of only mechanical effects of mandibular movement, it was hypothesized that stabilizing the jaw would eliminate these intrinsic vowel duration differences. Based on acoustic measurements that were made, it was found that the fixed-jaw conditions affected the intrinsic durational differences between high and low vowels to some extent, but the contrast was not entirely removed. In addition to the vowel duration changes, a number of systematic effects upon consonant closure duration were also observed as a result of the normal versus fixed-jaw conditions.

下顎前歯部排列位置の変化が発音に及ぼす影響

The oral cavity with such articulatory organs as the lips, tongue, uvula, and soft palate which are very versatile in mobility plays an important role of articulatory function in producing various sounds by changing their forms. When many teeth are lost by surgery or periodontosis, it is well known by experience that the articulatory function of the oral cavity is greatly impaired and speech impediment is induced. In practice in prosthodontics, therefore, it is very important to restore the articulatory function, as well as esthetic and occlusion, by carrying out a prosthetic treatment at the region with a tooth-loss. In view of the articulatory function, it is important to examine the relation between artificial tooth position and articulation. Four types of experimental dentures with over bite and over jet varied were therefore prepared for the cases of a tooth-loss at the lower anterior region. Five syllables of "S" followed by a vowel in Japanese (Sa, Si, Su, Se, So) were designated as the target syllables and analyses of "S" position on Mandibular Kinesiograph and of the voice on a sound spectrograph were made. Then, the effect of the changes in the artificial tooth position in the lower anterior tooth region on articulation was examined in detail. The results were as follows : 1. With the standard-type experimental denture with over bite at 2mm and over jet at 2mm, no positional changes in the anteroposterior "S" position were found in any of the "S" syllables. 2. The up-down "S" positions differed among the target syllables. The positions for "So" showed the greatest difference in comparison with other syllables. 3. The effect of increase in over jet on the anteroposterior "S" position was greater in comparison with the up-down "S" position, and the anteroposterior "S" position tended to be displaced forward. However, with the experimental denture with over bite at 4mm and over jet at 6mm, the tendency toward forward displacement of the anteroposterior "S" position decreased. 4. The effect of increase in over jet on the up-down "S" position was rather small, and no fixed patterns were seen in the displacement direction of the up-down "S" position. 5. The consonant duration differed among the target syllables. It was comparatively long for "Si" and "Su" and was the shortest for "Sa". 6. The effect of the artificial tooth position in the lower anterior region on the consonant duration was small. 7. The average amplitude differed among the target syllables. It was comparatively large for "Si" and "Su" and was the smallest for "Sa". 8. The effect of increase in over jet on the average amplitude was very marked. Although the average amplitude showed a decreasing tendency, the experimental denture with over bite at 4mm and over jet at 6mm, showed a decline in the rate of decrease. 9. Analyses of frequency showed that the energy level tended to decline in the hlgh frequency range form 4, 500&acd;5, 000Hz to 9, 500&acd;10, 000Hz for all target syllables of "S" with an increase in over jet. However with the experimental denture with over bite at 4mm and over jet at 6mm, the declining tendency of energy level decreased.