Search

Vowel contrasts may be reduced or neutralized before coda laterals in English [Bernard (1985). The Cultivated Australian: Festschrift in Honour of Arthur Delbridge, pp. 319–332; Labov, Ash, and Boberg (2008). The Atlas of North American English, Phonetics and Sound Change (Gruyter Mouton, Berlin); Palethorpe and Cox (2003). International Seminar on Speech Production (Macquaire University, Sydney, Australia)], but the acoustic characteristics of vowel-lateral interaction in Australian English (AusE) rimes have not been systematically examined. Spectral and temporal properties of 16 pre-lateral and 16 pre-obstruent vowels produced by 29 speakers of AusE were compared. Acoustic vowel similarity in both environments was captured using random forest classification and hierarchical cluster analysis of the first three DCT coefficients of F1, F2, and F3, and duration values. Vowels preceding /l/ codas showed overall increased confusability compared to vowels preceding /d/ codas. In particular, reduced spectral contrast was found for the rime pairs /iːl-ɪl/ (feel-fill), /ʉːl-ʊl/ (fool-full), /əʉl-ɔl/ (dole-doll), and /aeɔl-ael/ (howl-Hal). Potential articulatory explanations and implications for sound change are discussed.

Glottalisation is an important cue to coda stop voicelessness, particularly for younger Australian English speakers who utilise glottalisation more than older speakers, suggesting a recent sound change. However, most previous studies of glottalisation in this variety of English have focussed on single word utterances, raising questions about whether glottalisation in those studies may have been prosodically conditioned rather than specific to the coda stop: Could the observed effect have been due to phrase-final creaky voice, which is acoustically similar to coda-related glottalisation? This study therefore explored the differential effects of phrase position on the production of glottalisation. Phrase-medially (where phrase-final creaky voice is not expected to occur), results confirmed previous findings that glottalisation cues coda stop voicelessness and that it does so more frequently for younger compared to older speakers. In phrase-final position, rates of glottalisation increased, but older speakers appeared more similar to younger speakers in use of glottalisation, suggesting that the change towards the increased use of glottalisation may be nearing completion in this prosodic position. Younger speakers appear to represent a more advanced stage of the change extending the use of glottalisation from phrase-final to phrase-medial position.

This study compares prosodic structural effects on nasal (N) duration and coarticulatory vowel (V) nasalization in NV (Nasal-Vowel) and CVN (Consonant-Vowel-Nasal) sequences in Mandarin Chinese with those found in English and Korean. Focus-induced prominence effects show cross-linguistically applicable coarticulatory resistance that enhances the vowel's phonological features. Boundary effects on the initial NV reduced N's nasality without having a robust effect on V-nasalization, whose direction is comparable to that in English and Korean. Boundary effects on the final CVN showed language specificity of V-nasalization, which could be partly attributable to the ongoing sound change of coda nasal lenition in Mandarin.

The present study addresses how listeners may be sensitive to temporal regularities associated with prosody in their perception of durational segmental cues. Specifically, this study tests whether expectations about phrase-final lengthening mediate listeners' categorization of a "coat"∼"code" vowel duration continuum, where vowel duration cues stop voicing. Prosodic position of the target was manipulated such that it was either medial, or final in a carrier phrase. Results indicate that this prosodic manipulation influences categorization such that listeners effectively require longer vowel duration for a "code" response when the target is phrase-final, suggesting that prosodic patterns can modulate listeners' processing of temporal cues.

A nonlinear ultrasonic test is proposed for material damage evaluation using thermal modulation. Temperature change excites and modulates nonlinear behavior of ultrasonic waves in concrete. Coda wave interferometry was used to analyze the relative velocity change of ultrasonic wave with temperature variations. On concrete samples with different levels of thermal damage, experimental results indicate that the samples with a higher damage level demonstrated higher sensitivity to temperature variations. In addition, a slow dynamics nonlinear behavior was observed. When the temperature changed and then remained constant, the wave velocity gradually approached to its equilibrium value.

The sound speed of a porous medium changes with fluid substitution when the fluids have different acoustic properties. The authors demonstrate that coda wave interferometry is capable of sensing subtle local sound speed changes associated with minute fluid displacements, Δh. In fact the resolution on fluid motion is given by a simple scaling relationship, Δ h min / λ ∼ t − γ e 2 α t, where t is the waveform time, λ is the wavelength, γ is a constant that varies based on the nature of the acoustic propagation, and α is a system specific acoustic attenuation coefficient. In contrast to the conventional notion that later arrivals (further into the coda) give greater sensitivity to fluid movement, this scaling relationship suggests that there is a temporal optimum in sensitivity to Δh. This is the case even though later arrivals exhibit signal intensities well above the noise floor. The authors elucidate the physical basis for determining the waveform time at which the sensitivity is optimal.

The nominally incoherent coda of a scattered wavefield has been shown to be a remarkably sensitive quantitive monitoring tool. Its success is, however, often limited to applications where only moderate or localised changes in the scattering properties of the medium can be assumed. However, the compressional deformation of a relatively homogeneous rock matrix towards failure represents for a monitoring wavefield pronounced changes in both velocity and scattering power often due to a distribution of inelastic changes. A rolling reference wavefield is implemented when applying coda-wave interferometry (CWI) and coda-wave decorrelation (CWD), allowing relative velocity and material scattering power monitoring for such applications. It is demonstrated how this modification enables the qualitative monitoring of stages in material deformation common to unconfined compressive strength tests. In addition, the precursory/subtle onset of material yielding is identifiable in both the CWI and CWD methods, which was not possible when comparing to a fixed reference wavefield. It is, therefore, expected that this approach will enable these coda-based methods to robustly monitor continuous, destructive processes at a variety of scales. Possible applications include critical infrastructure, landslide, and reservoir compaction monitoring where both the subtle continuous and sudden large changes in a material's scattering properties occur.

International audience; Mammals use binaural or monaural (spectral) cues to localize acoustic sources. While the sensitivity of terrestrial mammals to changes in source elevation is relatively poor, the accuracy achieved by the odontocete cetaceans' biosonar is high, independently of where the source is. Binaural/spectral cues are unlikely to account for this remarkable skill. In this paper, bone-conducted sound in a dolphin's mandible is studied, investigating its possible contribution to sound localization. Experiments are conducted in a water tank by deploying, on the horizontal and median planes of the skull, ultrasound sources that emit synthetic clicks between 45 and 55 kHz. Elastic waves propagating through the mandible are measured at the pan bones and used to localize source positions via either binaural cues or a correlation-based full-waveform algorithm. Exploiting the full waveforms and, most importantly, reverberated coda, it is possible to enhance the accuracy of source localization in the vertical plane and achieve similar resolution of horizontal- vs vertical-plane sources. The results noted in this paper need to be substantiated by further experimental work, accounting for soft tissues and making sure that the data are correctly mediated to the internal ear. If confirmed, the results would favor the idea that dolphin's echolocation skills rely on the capability to analyze the coda of biosonar echoes.

This research uses the acoustic coda phase delay method to estimate relative changes in air temperature between explosions with varying event masses and heights of burst. It also places a bound on source-receiver distance for the method. Previous studies used events with different shapes, height of bursts, and masses and recorded the acoustic codas at source-receiver distances less than 1 km. This research further explores the method using explosions that differ in mass (by up to an order of magnitude) and are placed at varying heights. Source-receiver distances also cover an area out to 7 km. Relative air temperature change estimates are compared to complementary meteorological observations. Results show that two explosions that differ by an order of magnitude cannot be used with this method because their propagation times in the near field and their fundamental frequencies are different. These differences are expressed as inaccuracies in the relative air temperature change estimates. An order of magnitude difference in mass is also shown to bias estimates higher. Small differences in height of burst do not affect the accuracy of the method. An upper bound of 1 km on source-receiver distance is provided based on the standard deviation characteristics of the estimates.

This study compared fundamental frequency (F0), H1*-H2*, H1*-A1*, and harmonics-to-noise ratio (HNR) measures during intervals of three types of segmental glottalization and intervals of prosodic, phrasal creak. Results show that phrasal creak differs from all segmental glottalization types with lower F0 and H1*-H2* and higher HNR. /t/ glottalization before syllabic nasals has lower H1*-A1* than all other creaky phonation types, indicating concurrent pre-nasalization of segments preceding nasals, and coda /t/ glottalization has lower HNR than vowel-initial glottalization. A positive relationship between rates of segmental glottalization and phrasal creak suggests that speakers do not avoid their co-occurrence despite potential perceptual confusability.

The fronting of the two high-back vowels /u:/ and /U/ in Southern British English is very well documented, but mainly in the acoustic domain. This paper presents articulatory (ultrasound) data, comparing the relative tongue position of these vowels in fronting and non-fronting consonantal contexts, i.e. preceding a coronal consonant (food, foot) and preceding a coda /l/ (fool, full ). Particular attention is paid to the comparison between articulatory results and corresponding acoustic measurements of F2 in both vowels. Results show thatthe average dierences between food and foot and their dynamic proles are similar in articulation and acoustics. In /u:l/ sequences (fool ), tongue position is more advanced than could be inferred from its low F2. In addition, even though the tongues position in fool and full is clearly distinct, there is no comparable corresponding dierence in F2. This suggests that the common articulatory metaphor that characterises F2 increase as fronting must be used cautiously. In the case of English high-back vowel fronting, special attention must bepaid to the anking consonants when estimating vowel distances. This paper also provides specic recommendations for recording and analysing ultrasound data in research on vowel variation and change.

This study investigates the capacity for targeted hyperarticulation of contextually-relevant contrasts. Participants communicated target words with final /s/ or /z/ when a voicing minimal-pair (e.g., target dose, minimal-pair doze) either was or was not available as an alternative in the context. The results indicate that talkers enhance the durational cues associated with the word-final voicing contrast based on whether the context requires it, and that this can involve both elongation as well as shortening, depending on what enhances the contextually-relevant contrast. This suggests that talkers are capable of targeted, context-sensitive temporal enhancements.

Mid-frequency (3.5 kHz) pulses (5 pings, 5 s-long, one per minute) were radiated by an omnidirectional source of opportunity in a deep (3 km) Arctic ocean at 30m depth under ice during the ICEX14 experiment, ALAS component [Williams et al., IEEE-JOE 43, 145–159 (2018)]. Data analysis and modeling results are presented for normalized (re source level) acoustic intensityreceived by an omnidirectional hydrophone located at a 719 m distance from and the same depth as the source. Received time series showed a clear direct blast signal followed by an about 30 s-long reverberation coda. Modeling of the under-ice propagation shows that magnitude of the direct signal is strongly affected by presence of a weak near-surface (within 50 m depths) acoustic channel and by reflectivity of the ice, controlled by its physical properties and acoustical parameters, particularly the ice layer thickness. Analysis of the reverberation coda shows effects of reflections and scattering from rough and/or heterogeneous bottom, as well as reflectivity of the ice layer. Some other effects that may be caused by variability of the sound speed profile and absorption in water are considered as well. Possibilities of inversions for water column, bottom and ice parameters based on this and similar experiments are discussed. [Work supported by ONR.]

Frequency-resolved experimental measurements of ultrasonic diffusivity in the MHz range are presented. The samples under study are two-dimensional random arrangements of parallel steel rods immersed in water and exhibit high-order multiple scattering. Their physical characteristics, particularly the density and pair-correlation functions of the scatterers, are well controlled. These synthetic samples are used as phantoms for actual inhomogeneous materials. The resonant nature of the scatterers has a strong effect on diffusivity, which is shown to vary significantly with frequency. This may affect the result of broadband measurements of apparent diffusivity, which can be expected to depend on time and sample thickness, whereas diffusivity is intrinsically an intensive parameter. Moreover, the transport speed is shown to vary drastically with frequency, sometimes by more than 50%, due to a very narrow resonance that slows down transport. Interestingly, this sharp resonance could only be revealed by experiments performed with coda waves, and not with ballistic or coherent waves whose frequency resolution is intrinsically limited from an experimental point of view.

Blast waves produced by 60 high-explosive detonations were recorded at short distances (few hundreds of meters); the corresponding waveforms show charge-configuration independent coda-like features (i.e., similar shapes, amplitudes, and phases) lasting several seconds. These features are modeled as reflected and/or scattered waves by acoustic reflectors/scatters surrounding the explosions. Using explosion pairs, relative coda phase delays are extracted and modeled as changes in sound speed due to changes in air temperature. Measurements from nearby weather towers are used for validation.

This paper examines the perceptual weight of cues to the coda glottal consonant contrast in Trique (Oto-Manguean) with native listeners. The language contrasts words with no coda (/Vː/) from words with a coda glottal stop (/VɁ/) or breathy coda (/Vɦ/). The results from a speeded AX (same-different) lexical discrimination task show high accuracy in lexical identification for the /Vː/-/Vɦ/ contrast, but lower accuracy for the other contrasts. The second experiment consists of a labeling task where the three acoustic dimensions that distinguished the glottal consonant codas in production [duration, the amplitude difference between the first two harmonics (H1-H2), and F0] were modified orthogonally using step-wise resynthesis. This task determines the relative weight of each dimension in phonological categorization. The results show that duration was the strongest cue. Listeners were only sensitive to changes in H1-H2 for the /Vː/-/Vɦ/ and /Vː/-/VɁ/ contrasts when duration was ambiguous. Listeners were only sensitive to changes in F0 for the /Vː/-/Vɦ/ contrast when both duration and H1-H2 were ambiguous. The perceptual cue weighting for each contrast closely matches existing production data [DiCanio (2012 a). J. Phon. 40, 162-176] Cue weight differences in speech perception are explained by differences in step-interval size and the notion of adaptive plasticity [Francis et al. (2008). J. Acoust. Soc. Am. 124, 1234-1251; Holt and Lotto (2006). J. Acoust. Soc. Am. 119, 3059-3071].

Nonnative accents are commonplace, but why? Ample research shows that perceptual representations of second-language speakers are shaped by their first language. But is production also affected? If perceptual representations perfectly control motor production, then second-language speakers should understand their own speech accurately. To test this, we recorded 48 native Mandarin speakers labeling pictures in English. We then played back their own recorded productions (e.g., “lock”) as they chose one of four pictures (lock, log, shape, and ship). They also heard a paired native English speaker. Words contained contrasts challenging for Mandarin speakers, principally coda voicing (lock, log) and similar-vowel (shape, ship) pairs. Listeners achieved 89% accuracy on both their own productions and native speakers,’ suggesting good matching between perception and production. However, errors were unevenly distributed: Mandarin speakers heard their own voiced codas (log) as voiceless (lock) more often than the reverse (10% vs. 5%, p = 0.002). This mirrors a similar but larger voiceless bias in native-English listeners hearing accented stimuli, suggesting that Mandarin speakers’ coda voicing perception is more nativelike than their production. Ongoing work attempts to differentiate interlanguage intelligibility effects from learning of idiosyncratic speech patterns, and we are exploring which acoustic features predict recognition.Nonnative accents are commonplace, but why? Ample research shows that perceptual representations of second-language speakers are shaped by their first language. But is production also affected? If perceptual representations perfectly control motor production, then second-language speakers should understand their own speech accurately. To test this, we recorded 48 native Mandarin speakers labeling pictures in English. We then played back their own recorded productions (e.g., “lock”) as they chose one of four pictures (lock, log, shape, and ship). They also heard a paired native English speaker. Words contained contrasts challenging for Mandarin speakers, principally coda voicing (lock, log) and similar-vowel (shape, ship) pairs. Listeners achieved 89% accuracy on both their own productions and native speakers,’ suggesting good matching between perception and production. However, errors were unevenly distributed: Mandarin speakers heard their own voiced codas (log) as voiceless (lock) more often than the re...

Speakers place focus on specific linguistic elements using systematic modulations to highlight new or important information. Focus effects have primarily been studied for domains larger than a segment, such as syllables or words. By eliciting focus on segment-sized units using a corrective focus task, this study examines at what granularity or scope speakers manifest focus and its acoustic consequences for consonants differing in gestural and tonal compositions. Data were obtained from 16 native Korean speakers producing corrective focus on segments having varying syllable-internal positions (i.e., onset, nucleus, versus coda focus) and varying segmental compositions (i.e., tense and lax consonants). The results indicate that the patterns of focus modulation do not differ depending the on anticipated sub-syllabic focus position but rather are exhibited through the syllable. This suggests that the domain of focus extends beyond segment-sized intervals. Moreover, focus is manifested with systematic variations as a function of the gestural compositions active during its domain. Finally, there is some support for an interaction among multiple prosodic gestures—focus gestures and accentual phrase tone gestures. Overall, evidence suggests that focus has a syllable-sized (or larger) scope of effect that is sensitive to co-active articulatory and prosodic (tonal) gestures. [Work supported by NIH.]

This study investigates differences in sentence and story production between native and non-native speakers of English for use with a system of Automatic Speech Recognition (ASR). Previous studies have shown that production errors by non-native speakers of English include misproduced segments (Flege, 1995), longer pause duration (Anderson-Hsieh and Venkatagiri, 1994), abnormal pause location within clauses (Kang, 2010), and non-reduction of function words (Jang, 2009). The present study uses phonemically balanced sentences from TIMIT (Garofolo et al., 1993) and a story to provide an additional comparison of the differences in production by native and non-native speakers of English. Consistent with previous research, preliminary results suggest that non-native speakers of English fail to produce flaps and reduced vowels, insert or delete segments, engage in more self-correction, and place pauses in different locations from native speakers. Non-native English speakers furthermore produce different patterns of intonation from native speakers and produce errors indicative of transfer from their L1 phonology, such as coda deletion and vowel epenthesis. Native speaker productions also contained errors, the majority of which were content-related. These results indicate that difficulties posed by English ASR systems in recognizing non-native speech are due largely to the heterogeneity of non-native production.

Low amplitude diffuse coda ultrasound probe signals received through unconsolidated glass bead packs under static vertical loads are used to measure the bead pack's slow dynamics, in which the speed of low amplitude probe waves varies slowly with time long after loading. Coda-wave interferometry reveals tiny stretches in the waveform, with precision of parts in 105. These stretches are ascribed to decreased contact stiffnesses or decreased sound speeds. The coda-wave stretch is observed to be sensitive to low frequency conditionings at strains of order 10-5. Such conditionings include taps on the sides, low frequency harmonic vibrations and impulses on top, and step changes in vertical load. Typical behavior after the conditioning is an initial loss of stiffness followed by a slow healing that proceeds with the logarithm of time. The initial loss of stiffness and its slow dynamic healing are investigated for sensitivity to sundry parameters. Low amplitude diffuse coda ultrasound probe signals received through unconsolidated glass bead packs under static vertical loads are used to measure the bead pack's slow dynamics, in which the speed of low amplitude probe waves varies slowly with time long after loading. Coda-wave interferometry reveals tiny stretches in the waveform, with precision of parts in 105. These stretches are ascribed to decreased contact stiffnesses or decreased sound speeds. The coda-wave stretch is observed to be sensitive to low frequency conditionings at strains of order 10-5. Such conditionings include taps on the sides, low frequency harmonic vibrations and impulses on top, and step changes in vertical load. Typical behavior after the conditioning is an initial loss of stiffness followed by a slow healing that proceeds with the logarithm of time. The initial loss of stiffness and its slow dynamic healing are investigated for sensitivity to sundry parameters.

This study investigated how listeners' native language [Saudi Arabic (SA)] could affect the perception of non-native (English) sound patterns. SA stopsare generally released, while English stops show free variation: released or unreleased. In a phoneme detection task, SA listeners were asked to detect English stops in coda position, and their accuracy and reaction time (RT) were measured. The influence of three factors of the final coda (i.e., release, voicing, and place of articulation) on the accuracy and RT was statistically analyzed using mixed effect models. The results indicated a significant interaction among the three factors on the listeners' accuracy but not on the RT. The results revealed that listeners relied mostly on release bursts as a phonetic cue to accurately detect the target phoneme. The significant interaction between release and voicing suggested that the unreleased voiced stops were easier to detect than the voiceless counterparts. Despite the difference in the accuracy, the listeners' RT was not significantly influenced by release, voicing, or place of articulation. The results suggested that L1 experience effects non-native perception if the target phoneme does not have enough acoustic information. Listeners can accurately detect unfamiliar sound patterns when the signal is acoustically rich. This study investigated how listeners' native language [Saudi Arabic (SA)] could affect the perception of non-native (English) sound patterns. SA stopsare generally released, while English stops show free variation: released or unreleased. In a phoneme detection task, SA listeners were asked to detect English stops in coda position, and their accuracy and reaction time (RT) were measured. The influence of three factors of the final coda (i.e., release, voicing, and place of articulation) on the accuracy and RT was statistically analyzed using mixed effect models. The results indicated a significant interaction among the three factors on the listeners' accuracy but not on the RT. The results revealed that listeners relied mostly on release bursts as a phonetic cue to accurately detect the target phoneme. The significant interaction between release and voicing suggested that the unreleased voiced stops were easier to detect than the voiceless counterparts. Despite the difference in the accuracy, the listen...

This paper reports on a three-dimensional/four-dimensional ultrasound investigation of American English laterals. Traditional descriptions of English distinguish between pre-vocalic (onset) laterals (‘light-l’), said to involve a coronal gesture creating an alveolar occlusion on the center line with lateral venting, and post-vocalic (coda) laterals (‘dark-l’), said to involve both the coronal gesture and an additional dorsal constriction gesture. Recent work suggests that this view of laterals may be overly simple: e.g., onset laterals in American English may also have a dorsal gesture [Rhodes et al., JASA 137(4), 2268–2269 (2015)]. Previous work also noted a persistent tongue configuration involving deep cupping of the midline in the palatal region yielding a raised tongue tip and dorsum around the cupped region. This configuration appears in both onsets and codas, even for a speaker who exhibited no coronal contact [Berkson et al., ICASSP (2017), pp. 5080–5084]. This study presents configurations for 20 college-aged native speakers of English (10 female) who show individual differences in (i) the presence of palatal grooving and (ii) consistency between onset and coda configurations. We contend that the onset versus coda lateral distinction cannot be reduced to a single description in terms of lingual configuration.

In the process of phonological development, fricatives are generally assumed to be later acquired than stops. However, most of the observational work on which this claim is based has concerned itself with word-initial onset consonants; little is known about how and when fricatives are mastered in word-final coda position (e.g., nose). This is all the more critical in a language like English, where word-final fricatives often carry important morphological information (e.g., toes, goes). This study examines the development of duration cues to the voicing feature contrast in coda fricatives, using longitudinal spontaneous speech data from CVC words (e.g., noise vs face) produced by three children (1;6–2;6 years) and six mothers. Results show that the children were remarkably adult-like in the use of duration cues to voicing contrasts in fricatives even in this early age range. Furthermore the children, like the mothers, had longer frication noise durations for morphemic compared to non-morphemic fricatives (e.g., toes vs nose) when these segments occurred in utterance-final position. These results suggest that although children's fricatives tend to be overall longer and more voiced compared to those of adults, the voicing and morphological contrasts for fricative codas are acquired early in production.

The shape of pitch contours has been shown to have an influence on the perceived vowel duration in that listeners tend to perceive vowels as shorter when they are produced as longer (compensatory listening strategy, Guessenhoven & Zhou, 2013). This study further investigates if the syllable structure allowed in one's native language and the frequency of tone types also affect perceived duration. An AX perceptual experiment was conducted using disyllabic structures (i.e., mepVC) in which the target vowel ([a] and [i]) in four duration steps (180, 210, 240, 270ms) and 5 pitch contours (LL, HH, HL, LH, HLH) and 7 codas ([p], [t], [k], [m], [n], [ŋ], [∅]) are embedded in the second syllable, to be compared with an anchor stimulus with a fixed vowel duration (225ms) and pitch (195Hz). Taiwanese Southern Min speakers, whose native language allows obstruent codas, and Mandarin speakers, whose native language only allows nasal codas, were recruited to participate in the experiment in which they were told to liste...

The sensitivity of odontocetes to changes in sound source azimuth and, in particular, elevation (minimum audible angles) has been reported to be superior to those of humans and bats. It has been suggested that binaural/spectral cues might be insufficient to account for this skill. We investigate bone-conducted sound in a short-beaked common dolphin’s mandible and attempt to determine whether and to what extent it could contribute to the task of localizing a sound source. Experiments are conducted in a water tank by deploying, on the horizontal and median planes of the skull, sound sources that emit synthetic clicks between 45 and 55 kHz. Elastic waves propagating through the mandible are measured at the pan bones and used to localize source positions via binaural cues, as well as a correlation-based full-waveform algorithm. We find that by making use of the full waveforms, and, most importantly, of their reverberated coda, the accuracy of source localization in the vertical plane can be enhanced. While further experimental work is needed to substantiate this speculation, our results suggest that the auditory system of dolphins might be able to localize sound sources by analyzing the coda of biosonar echoes.

Listeners quickly adapt to foreign-accented speech in transcriptions and word/sentence judgments (Bradlow & Bent, 2008; Clarke & Garrett, 2004; Sidaras, Alexander, & Nygaard, 2009). However, continuous spoken-word recognition measures such as eye tracking have revealed limitations on foreign-accent adaptation. For example, Trude, Tremblay, and Brown-Schmidt (2013) found that English listeners’ adaptation to a second-order constraint in a Quebec French talker’s accent (/i/ = [ɪ] before coda consonants except voiced fricatives) was limited. This study investigates whether foreign-accent adaptation is less limited when the mapping between the accented and underlying words does not require higher-level inferencing. English listeners completed an eye-tracking experiment in which they heard a (female) French talker and a (male) English talker. Target and competitor words began with a voiced or voiceless stop and were otherwise temporarily ambiguous (e.g., timberand dimple). All stops were resynthesized: The French talker’s stops were prevoiced (voiced) and short-lag (voiceless) and the English talker’s stops were short-lag (voiced) and long-lag (voiceless). Preliminary results suggest an effect of voicing only for the French talker, with voiceless-stop targets generating more competition than voiced-stop targets. Importantly, this effect decreased only slightly from the first to the second half of the experiment, suggesting weak foreign-accent adaptation. Listeners quickly adapt to foreign-accented speech in transcriptions and word/sentence judgments (Bradlow & Bent, 2008; Clarke & Garrett, 2004; Sidaras, Alexander, & Nygaard, 2009). However, continuous spoken-word recognition measures such as eye tracking have revealed limitations on foreign-accent adaptation. For example, Trude, Tremblay, and Brown-Schmidt (2013) found that English listeners’ adaptation to a second-order constraint in a Quebec French talker’s accent (/i/ = [ɪ] before coda consonants except voiced fricatives) was limited. This study investigates whether foreign-accent adaptation is less limited when the mapping between the accented and underlying words does not require higher-level inferencing. English listeners completed an eye-tracking experiment in which they heard a (female) French talker and a (male) English talker. Target and competitor words began with a voiced or voiceless stop and were otherwise temporarily ambiguous (e.g., timberand dimple). All stops were resynthesized: The Fre...

This is a pilot acoustic study of the distinction between short and long vowels in Plains Cree, based on publically available recordings of two noted teachers of this Algonquian language. Both speakers strongly differentiate all pairs of short/long vowels in terms of duration: long vowels are 2 to 2.5 times longer than short ones. Moreover, both speakers differentiate pairs in terms of vowel quality, but one of the speakers shows considerable overlap in the vowel space. Finally, durational differences between short/long pairs are preserved when vowels occur with a coda consonant, but the overall rhyme shows considerable contrast reduction.

Vowel duration in English serves as a cue to the voicing distinction of the following consonant, in a way that vowels preceding voiced consonants are longer than those preceding voiceless consonants. This study will examine the differences in vowel duration in various phonetic contexts produced by 8 Korean-speaking L2 learners of English in comparison with the differences produced by 4 native English speakers from the K-SEC corpus (Korean-Spoken English corpus). Specifically, we will test for the durational evidence of the influence of different manners of articulation and syllable positions on preceding or following vowels and investigate how the influence is phonetically realized by two language groups. Our hypotheses and predictions are as follows: 1) vowels are longer when preceded by consonants in coda position than when followed by the same consonants in onset position, 2) vowels are longer when preceded by nasals than when preceded by non-nasal consonants, and 3) Korean speakers produce the differences in vowel length to a smaller extent than native English speakers.Vowel duration in English serves as a cue to the voicing distinction of the following consonant, in a way that vowels preceding voiced consonants are longer than those preceding voiceless consonants. This study will examine the differences in vowel duration in various phonetic contexts produced by 8 Korean-speaking L2 learners of English in comparison with the differences produced by 4 native English speakers from the K-SEC corpus (Korean-Spoken English corpus). Specifically, we will test for the durational evidence of the influence of different manners of articulation and syllable positions on preceding or following vowels and investigate how the influence is phonetically realized by two language groups. Our hypotheses and predictions are as follows: 1) vowels are longer when preceded by consonants in coda position than when followed by the same consonants in onset position, 2) vowels are longer when preceded by nasals than when preceded by non-nasal consonants, and 3) Korean speakers produce the differe...

Phonological studies on Sasak tend to agree that Sasak has a six-vowel system /i, u, e, o, ə, a/ (Archangeli et al., n.d.; Chahal, 1998; Jacq, 1998). However, in Ampenan Sasak, a dialect spoken in the provincial capital of Mataram, it is difficult to attribute vowel quality differences of tense mid-vowels [e, o] and lax mid-vowels [ɛ, ɔ] to allophonic variation of underlying phonemes /e, o/. While lax mid-vowels tend to appear in heavy syllables and tense mid-vowels appear in light syllables (see also Archangeli et al., n.d.), there are exceptions (e.g., /baɾeh/ "late," /əmbong/ "reservoir") and several minimal pairs (e.g., /bəɾəmbok/ "discuss," /bəɾəmbɔk/ "breathe;" /koboʔ/ "levening," /kɔbɔʔ/ "play with water"). Further, speakers’ intuitions about these distinctions are unreliable. In order to understand this relationship, we elicited wordlists across 13 female speakers of Ampenan Sasak aged 18 to 48. Wordlists were balanced for syllable weight, coda segment, and syllable position and subjected to acoustic analysis. Preliminary results show that the quality of the vowel is highly affected by the weight of the syllable and the coda segment. This suggests that mid-vowels are quasi-phonemic in Ampenan Sasak (Ladd 2013). While they generally behave as allophones, this paradigm is not strictly followed.Phonological studies on Sasak tend to agree that Sasak has a six-vowel system /i, u, e, o, ə, a/ (Archangeli et al., n.d.; Chahal, 1998; Jacq, 1998). However, in Ampenan Sasak, a dialect spoken in the provincial capital of Mataram, it is difficult to attribute vowel quality differences of tense mid-vowels [e, o] and lax mid-vowels [ɛ, ɔ] to allophonic variation of underlying phonemes /e, o/. While lax mid-vowels tend to appear in heavy syllables and tense mid-vowels appear in light syllables (see also Archangeli et al., n.d.), there are exceptions (e.g., /baɾeh/ "late," /əmbong/ "reservoir") and several minimal pairs (e.g., /bəɾəmbok/ "discuss," /bəɾəmbɔk/ "breathe;" /koboʔ/ "levening," /kɔbɔʔ/ "play with water"). Further, speakers’ intuitions about these distinctions are unreliable. In order to understand this relationship, we elicited wordlists across 13 female speakers of Ampenan Sasak aged 18 to 48. Wordlists were balanced for syllable weight, coda segment, and syllable position and subjected to aco...

Many previous studies have found that vowels are longer before voiced consonants than voiceless (Chen 1970). Fewer studies have examined vowel duration preceding aspirated vs. unaspirated stops, and results are inconsistent (Maddieson & Grandour 1976, Ohala & Ohala 1992). Durvasala and Luo (2012) find that both voicing and aspiration have a lengthening effect on the preceding vowel in Hindi, and that the voiced aspirated class has the longest preceding vowel duration. In the present study, I measure the duration of vowels preceding word final codas in Nepali, an Indo-Aryan language with the same four-way contrast as Hindi. A phonetic study of final stop realization has not yet been reported in the literature, though Nepali is described as variably neutralizing voiced aspirated to the plain voiced. I present an acoustic analysis of cues in word-final position including voicing, burst, and preceding vowel duration. I find evidence of neutralization of voiced and voiced aspirated stops in terms of aspiration duration, but also find results consistent with Durvasala and Luo (2012), that the voiced aspirated stops have longer preceding vowel than the voiced. The difference in vowel duration between the voiced and voiced aspirated classes suggests incomplete neutralization of voiced aspirated to voiced.Many previous studies have found that vowels are longer before voiced consonants than voiceless (Chen 1970). Fewer studies have examined vowel duration preceding aspirated vs. unaspirated stops, and results are inconsistent (Maddieson & Grandour 1976, Ohala & Ohala 1992). Durvasala and Luo (2012) find that both voicing and aspiration have a lengthening effect on the preceding vowel in Hindi, and that the voiced aspirated class has the longest preceding vowel duration. In the present study, I measure the duration of vowels preceding word final codas in Nepali, an Indo-Aryan language with the same four-way contrast as Hindi. A phonetic study of final stop realization has not yet been reported in the literature, though Nepali is described as variably neutralizing voiced aspirated to the plain voiced. I present an acoustic analysis of cues in word-final position including voicing, burst, and preceding vowel duration. I find evidence of neutralization of voiced and voiced aspirated stops in terms of aspiration...

In adult-directed speech (ADS), words are rarely produced canonically. Infant-directed speech (IDS) has been argued to contain more canonical productions. However, recent analyses show that IDS is as variable as ADS. Then, how could infants learn to privilege canonical forms, as has been shown for adult listeners? Previous research on variation in IDS has focused on word-final productions. In this study, we investigate whether the extent of variation in IDS differs by the position of a segment in a word. We sampled IDS to 6 infants between 16 and 24-mo-old from the Providence corpus. Utterances with /t/, /d/, /n/, /s/ and /z/ were identified orthographically, forced-aligned, corrected, and transcribed by 3 phonetically-trained native speakers of English. This yielded 28,775 segment tokens in word initial, medial and final position. Results confirmed that IDS is at least as variable as ADS (canonical pronunciations < 50%). However, variation was limited to coda positions; on average, over 90% of onsets wer...

Sperm whale (Physeter macrocephalus) clicks have a multi-pulse structure, a result of the reflection of sound energy between air sacs in the spermaceti organ. Although previous research revealed that usual clicks (used for echolocation) recorded away from a vocalizing whale's longitudinal axis have waveforms with poorly defined pulse structures, it has been unknown whether sperm whale coda clicks (used for communication) show similar off-axis effects. To address this knowledge gap, a hydrophone array was used to localize vocalizing sperm whales, and the waveforms of coda clicks recorded from different aspects were examined. Coda clicks recorded close to the whale's acoustic axis showed well-defined multi-pulsed waveforms, while those recorded off-axis did not. As for usual clicks, this suggests that sound energy radiates directly into the water upon reflection off the frontal sac.

In this paper, based on Korean vowel qualities measured in the production experiments, I point out that /o/-stems are also actively participating in the decay of Korean vowel harmony, and their speed of decay is much faster than /a/-stems' speed. How the late-starter /o/-stems could have the faster speed in the decay of Korean vowel harmony? The role of phonetic naturalness in language learning is the key. Measured vowel qualities show that /o/ is acoustically much closer to /ʌ/ as “disharmonic” pair than to /a/ as “harmonic” pair. In addition, in terms of articulatory, the raising of /o/ and the backing of /ʌ/ make them more similar. In the production experiments, 14 Seoul-Korean speakers participated (7 males, 7 females). Onset and coda of stems are controlled. For each condition, 1 existing word and 2 nonce words are used. Participants are asked to make a natural conjugation form of a given stem and to produce it three times. Unlike previous corpus and production studies reporting that /o/-stems show n...

The Seabed Characterization Experiment (SBC17) was conducted in February 2017 off the east coast of the United States in the New England Mudpatch. The overall mission of the experiment was to determine the effect of the seabed on acoustic propagation. However, it was determined from pre-surveys that the area also contained a significant concentration of benthic biology. In order to characterize the variability of the sediment and the influence of the biology, the Applied Research Laboratories at the University of Texas (ARL:UT), deployed an acoustic system to collect normal-incidence echo returns from a 6 to 20 kHz chirp. The system was deployed in various locations around the mud patch to obtain a diverse set of data. The data displayed a long coda, which has been previously linked to the presence of benthic biology. The ability of these data to predict and measure sediment type and benthic biology concentration will be assessed. [Work sponsored by ONR, Ocean Acoustics.]

In this talk, we report results on the nonlinear interactions of ultrasonic coda waves, in reverberating or multiple scattering mesoscopic solid media. Using the method of coda wave interferometry (CWI), we analyze the effect of mixing a coda wave with an additional lower frequency pump wave. The extracted CWI parameters, known to be highly sensitive to small geometric or elastic modifications of the tested medium, are shown to be pump-amplitude dependent and to capture finely the results of the nonlinear interactions. Although nonlinear self-action effects with coda waves have been reported in unconsolidated granular media, they are difficult to implement in cracked solids or concrete. Instead, the reported nonlinear CWI class of methods (NCWI) shows robustness, a high sensitivity, and has been applied successfully to various complex media and structures. We show through several examples on « model » media (cracked glass plates) and on concrete structures, that NCWI can be useful for the nondestructive e...

The liquid phoneme in Korean is known to have an optional retroflex allophone in syllable-final positions (Lee 1999). However, there lacks literature evaluating when and how this retroflex allophone is realized. To define the distribution of the retroflex allophone of the liquid phoneme in Korean, this study uses a read speech corpus of Seoul Korean to track the rhotacization of the liquid phoneme in coda positions. Rhotacization of the liquid is detected and measured by the lowering of the third formant. Preliminary results show that the liquid phoneme in coda position is realized as retroflex when preceded by a back vowel, and when followed by another liquid phoneme (gemination). The results suggest that the liquid phoneme is not simply optionally rhotacized in coda positions; instead, the retroflex allophone is realized in contexts defined by the preceding vowel and the following consonant. Future research remains to further examine the articulatory and perceptual aspects of the rhotacization.

Changes in phonation patterns have long been studied as correlates of various linguistic elements, such as the occurrence of irregular pitch periods (IPPs) at significant locations in prosodic structure (in phrase-initial, phrase-final, and pitch accented contexts) and word-final voiceless stops, especially /t/. But less is known about the development of this phonation pattern in children [cf. Song et al., JASA, 131, 3036-50, 2012], particularly in toddlers between the ages of 2;6 and 3;6. The study of its course of acquisition may shed light on the mechanisms involved, since child vocal folds are very different physiologically from those of adults, and change strikingly during development. Monosyllabic target words from the Imbrie Corpus of speech from 10 toddlers 2 1/2 to 3 1/2 years old, ending in /t, d/ were examined for evidence of IPPs. Preliminary results based on three adult/child pairs suggest that both adults and children produce IPPs preceding coda /t/ about 50% of the time. But children produc...

This study contributes to evaluating the robustness and accuracy of Green's function (GF) reconstruction by cross-correlation of noise, disentangling the respective roles of ballistic and reverberated (“coda”) signals. We conduct a suite of experiments on a highly reverberating thin aluminum plate, where we generate an approximately diffuse flexural wavefield. We validate ambient-noise theory by comparing cross correlation to the directly measured Green’s function. We develop analytically a theoretical model, predicting the dependence of the symmetry of the cross correlations on the number of sources and signal-to-noise ratio. We validate this model against experimental results. We next study the effects of cross-correlating our data over time windows of variable length, possibly very short, and taken at different points in the coda of recordings. We find that, even so, a relatively dense/uniform source distribution could result in a good estimate of the GF; we demonstrate that this window does not have t...

This study presents a case from Korean which demonstrates that allophonic and phonotactic knowledge has an effect on listeners’ categorization choices. In Korean, voicing is generally treated as redundant on consonants since the only voiced consonants are nasals and liquids. However, voiced oral stops can occur intervocalically as allophones of the voiceless lenis counterpart. A further restriction comes from the velar nasal [ŋ], which is restricted to coda position. Two identification experiments, in which Korean listeners were presented with a 10-step continuum from voiced oral to nasal stop in three places of articulation and were asked to decide whether they heard a nasal or not, showed an earlier boundary shift in the nasal responses in the initial position than in the medial position. This finding suggests that Korean listeners do perceive the redundant/allophonic [voice] feature and are likely to interpret a stop containing this feature as nasal in initial position but as an allophonic oral variant...

A universal scaling law is known empirically to characterize the high-frequency decay of bandpassed seismic coda. An explanation advanced for this universality, based on multiple-Rayleigh scatter of nonuniform strength, is generic enough to suggest that similar scaling may apply to decay of the impulse response in other random media. A submerged sand layer is probed experimentally as a candidate for such a medium, and its incoherent acoustic impulse reflection is found to have an exponential decay regime similar to that in seismology.

A distinctive vocalization of the sperm whale, Physeter macrocephalus (=P. catodon), is the coda: a short click sequence with a distinctive stereotyped time pattern [Watkins and Schevill, J. Acoust. Soc. Am. 62, 1485-1490 (1977)]. Coda repertoires have been found to vary both geographically and with group affiliation [Weilgart and Whitehead, Behav. Ecol. Sociobiol. 40, 277-285 (1997)]. In this work, the click timings and repetition patterns of sperm whale codas recorded in the Mediterranean Sea are characterized statistically, and the context in which the codas occurred are also taken into consideration. A total of 138 codas were recorded in the central Mediterranean in the years 1985-1996 by several research groups using a number of different detection instruments, including stationary and towed hydrophones, sonobuoys and passive sonars. Nearly all (134) of the recorded codas share the same "3+1" (/// /) click pattern. Coda durations ranged from 456 to 1280 ms, with an average duration of 908 ms and a standard deviation of 176 ms. Most of the codas (a total of 117) belonged to 20 coda series. Each series was produced by an individual, in most cases by a mature male in a small group, and consisted of between 2 and 16 codas, emitted in one or more "bursts" of 1 to 13 codas spaced fairly regularly in time. The mean number of codas in a burst was 3.46, and the standard deviation was 2.65. The time interval ratios within a coda are parameterized by the coda duration and by the first two interclick intervals normalized by coda duration. These three parameters remained highly stable within each coda series, with coefficients of variation within the series averaging less than 5%. The interval ratios varied somewhat across the data sets, but were highly stable over 8 of the 11 data sets, which span 11 years and widely dispersed geographic locations. Somewhat different interval ratios were observed in the other three data sets; in one of these data sets, the variant codas were produced by a young whale. Two sets of presumed sperm whale codas recorded in 1996 had 5- and 6-click patterns; the observation of these new patterns suggests that sperm whale codas in the Mediterranean may have more variations than previously believed.

Sperm whale (Physeter macrocephalus) coda patterns from the northwestern Gulf of Mexico have been examined. Recordings were collected via a towed passive hydrophone array. Sperm whales are known to produce a characteristic ‘‘hammering’’ click. Codas are rhythmic patterns of clicks [W. A. Watkins and W. E. Schevill, J. Acoust. Soc. Am. 62, 1485–1490 (1977)]. Two types of codas have been proposed: identity codas and general use or shared codas [Watkins et al., Cetology 49, 1–15 (1985)]. Identity codas are patterns that are unique to an individual for at least a few hours, and general use codas are vocalizations that seem to be shared among groups of whales. In the southeast Caribbean, it was found that 50% of the codas analyzed were made up of two patterns; therefore, it was proposed that shared codas may have a function other than individual identity [Moore et al., Mar. Mam. Sci. 9, 1–9 (1993)]. There also appears to be geographical differences in codas’ spacing and composition between the Galapagos and th...

In General American English (GAE), tense/lax vowel contrasts occur before syllables ending in /l/ (e.g., peel, pill), but not those ending in /ɹ/ (e.g., peer; no [i]/[ɪ] contrast). No such restrictions occur after a syllable-initial liquid consonant (e.g., lip, leap, rip, reap). This study investigates the hypothesis that these phonotactic asymmetries arise from the interaction of differences in gestural coordination between nuclear vowels and liquid consonants in syllable onset versus coda position (Browman & Goldstein 1995) and differences in blending parameter strength for GAE laterals versus rhotics. A study of speech production by four GAE speakers was conducted using real-time Magnetic Resonance Imaging (Narayanan et al. 2004). Liquid consonants were produced in various vocalic environments and syllable positions (syllable-final, syllable-initial: word-initial and intervocalic). Analysis using a center of gravity metric defined by midsagittal lingual outlines points to greater articulatory stability...

Short-a, or /ae/, production varies substantially across American English dialects and is noted to be useful in describing regional pronunciation differences. The current study examines the production and perception of /ae/ in California English, which raises in words with a final nasal coda (both velar and non-velar nasals). We explore both the acoustic characteristics of this potential merger with /ɛ/, along with lexical confusability between words with /ae/ and /ɛ/ in prenasal position. Productions of 32 native California speakers reveal that the formant space of prenasal /ae/ indeed overlaps with /ɛ/. However, these vowels have reliably different patterns of two secondary acoustic features: duration and nasal coarticulation. /ae/ is produced with both a longer duration and a greater degree of nasal coarticulation than /ɛ/. Results from a lexical identification task presenting unaltered, as well as duration-neutralized, items to 50 listeners reveal that these vowels are not confusable despite considerable f...

In this talk, we report results on the nonlinear interactions of ultrasonic coda waves with lower pump waves, in reverberating or multiple scattering mesoscopic solid media. Using the method of coda wave interferometry (CWI), we analyze the effect of mixing a coda wave with an additional lower frequency pump wave. The extracted CWI parameters, known to be highly sensitive to small geometric or elastic modifications of the tested medium, are shown to be pump-amplitude dependent and to capture finely the results of the nonlinear interactions. Although nonlinear self-action effects with coda waves have been reported in unconsolidated granular media, they are difficult to implement in cracked solids or concrete. Instead, the reported nonlinear CWI class of methods (NCWI) shows robustness, a high sensitivity, and has been applied successfully to various complex media and structures. We show through several examples on « model » media (cracked glass plates) and on concrete structures, that NCWI can be useful fo...

Auditory Processing Disorder (APD), a neurological hearing impairment, often leads to delays in speech development when diagnosed in childhood. It is unclear whether children with APD make articulation errors due to auditory processing deficits. The purpose of this study was to analyze the articulation error patterns of children diagnosed with APD. We examined speech samples from 26 children diagnosed with APD who participated in a previous research study involving speech perception testing. During this task, each child repeated each of 50 monosyllabic words presented to them in quiet. For each incorrect response, we examined the errors by type of articulation error (substitution, deletion, insertion, or distortion) and the syllable position (onset or coda). Speech perception scores were normal in quiet. However, 22 subjects responded incorrectly for one or more words, with substitution errors (40-50%) being the most prevalent type of errors. Articulation errors in the coda position (81%) were more common...

Cantonese inaudibly released oral stop codas, whose place cues occur only before constriction and not during or after constriction, appear to be merging among young-adult speakers (Law, Fung, & Bauer, 2001). Prior work (Yip, 2015) revealed substantial articulatory variation among five talkers, ranging from gestural preservation, to coproduction of lingual-lingual sequences, to gestural loss. Coda contrasts were cued by first- and third-formant loci (both lower into /k/ than into /t/) and preceding vowel duration (longer before /t/ than before /k/). In the present study, 18 Cantonese-speaking listeners’ perception of codas produced by talkers in the prior study was tested in an AXB task in which ambiguous coda productions were judged as more similar to each talker’s best coda /t/ or /k/ production, as assessed with ultrasound tongue imaging. Comparisons of listener sensitivity to cues—based on LMER estimates for change in accuracy with respect to relevant acoustic response variables—with their accuracy rat...

Child speakers of American English in the age range 2;6-3;6 often produce vowel final noise (VFN), or preaspiration, for [-voice] labial and velar stops in coda position of monosyllabic words [e.g., Shattuck-Hufnagel, Hanson, and Zhao, “Feature-cue-based processing of speech: A developmental perspective,” in Proc. ICPhS 2015]. In contrast, the female caregivers of these children produce VFN less often than the children, with significantly shorter durations. The current work extends the earlier work by studying a fricative, /s/, in coda position. Data from the same nine child/adult pairs are used. While the coda stops were all preceded by the vowel /ʌ/, the corpus includes two words with coda /s/, Gus and goose, allowing the observation of possible vowel effects. Preliminary analysis of a subset of the data shows that, as for coda stops, children are more likely than adults to produce VFN, with longer durations. However, adults show a vowel effect, being much more likely to produce VFN when /s/ follows /ʌ/...

Mandarin and Cantonese, though are both Chinese, exhibit differences in L2 English pronunciation, especially in obstruent codas. For the reason that only nasals -n, -m and -ng can be found at coda in Mandarin whereas apart from nasals, unreleased consonants -p -t and -k exits as well in Cantonese, it is reasonable to believe that L1 transfer makes influences on modifying codas for L2 learners. This present study aims at comparing the strategies used by Mandarin and Cantonese speakers in coda modification and resyllabification. By embedding the stimuli in a natural conversation, room-cleaning task, data are elicited and analyzed by Praat. Results indicates that up to 93% of codas are released by Mandarin speakers and 41% of those are unreleased by Cantonese speakers. Further, instead of epenthesizing vocalic elements, Mandarin speakers are found to strongly aspirate the codas with 58%, which contradicts most of the previous studies claiming that epenthesis is the main modification used among Mandarin L2 le...

This study explores the phoneme distribution and syllable structure of entry words in the CMU Pronouncing Dictionary to provide linguists with fundamental phonetic data on English word components. Entry words in the dictionary file were syllabified using an R script and examined to obtain the following results: First, English words preferred consonants to vowels in their word components. In addition, monophthongs occurred much more frequently than diphthongs. When all consonants were categorized by manner and place, the distribution indicated the frequency order of stops, fricatives, and nasals according to manner and that of alveolars, bilabials and velars according to place. Second, from the analysis of syllable structure, two-syllable words were most favored, followed by three- and one-syllable words. Of the words in the dictionary, 92.7% consisted of one, two, or three syllables. Third, the English words tended to exhibit discord between onset and coda consonants and between adjacent vowels. Dissimila...