Prosodic Position Research Articles

Prosodic Effects on Acoustic Cues for the Korean Stop Contrast: Evidence from Daejeon Korean

The paper examines acoustic realization of Daejeon Korean stops in three different prosodic positions, AP-initial, AP-medial, and word-medial. The results showed that Korean stops are differentiated by closure duration, VOT, and F0 and their realizations varied depending on the prosodic position. The distinction of closure duration among stops was prominent word- medially and the different realizations of F0s for stop contrast were prominent word-initially. The F0 contrast between lax and other stops in word- initial position was weakened word-medially and neutralized when they were preceded by word-initial aspirated or tense stops. Daejeon Korean stops showed a VOT merger of lax and aspirated stops in AP-initial and AP-medial position. In addition, VOTs for lax stops were even longer than those for aspirated stops in female AP-initial position.

Scandinavian umlaut and contrastive feature hierarchies

AbstractWith the application of the Contrastive Hierarchy Theory, the contrastive features of preliterary Scandinavian vowels are here inferred from the interaction between targets and triggers for metaphonic fronting, rounding and breaking. One Proto-Scandinavian feature hierarchy is reconstructed for prominent syllables and another for non-prominent ones. The former hierarchy sustained contrasts that differed from the latter, including contrast for rounding and a preserved distinction between Pre-Germanic */i/ and */e/. A prominence system is reconstructed that predicts both the outcome of syncope and the distribution of the two vowel systems between syllables. The analysis neatly accounts for many notorious cruxes of umlaut and breaking that correlate with the prosodic position of the trigger, including the frequent absence ofi-umlaut in light syllables.

Sibilant production in Hebrew-speaking adults: Apical versus laminal

ABSTRACTThe Hebrew IPA charts describe the sibilants /s, z/ as ‘alveolar fricatives’, where the place of articulation on the palate is the alveolar ridge. The point of constriction on the tongue is not defined – apical (tip) or laminal (blade). Usually, speech and language pathologists (SLPs) use the apical placement in Hebrew articulation therapy. Some researchers and SLPs suggested that acceptable /s, z/ could be also produced with the laminal placement (i.e. the tip of the tongue approximating the lower incisors). The present study focused at the clinical level, attempting to determine the prevalence of these alternative points of constriction on the tongue for /s/ and /z/ in three different samples of Hebrew-speaking young adults (total n = 242), with typical articulation. Around 60% of the participants reported using the laminal position, regardless of several speaker-related variables (e.g. tongue-thrust swallowing, gender). Laminal production was more common in /s/ (than /z/), coda (than onset) position of the sibilant, mono- (than di-) syllabic words, and with non-alveolar (than alveolar) adjacent consonants. Experiment 3 revealed no acoustical differences between apical and laminal productions of /s/ and of /z/. From a clinical perspective, we wish to raise the awareness of SLPs to the prevalence of the two placements when treating Hebrew speakers, noting that tongue placements were highly correlated across sibilants. Finally, we recommend adopting a client-centred practice, where tongue placement is matched to the client. We further recommend selecting targets for intervention based on our findings, and separating between different prosodic positions in treatment.

Vowels in infant-directed speech: More breathy and more variable, but not clearer

Infant-directed speech (IDS) is known to differ from adult-directed speech (ADS) in a number of ways, and it has often been argued that some of these IDS properties facilitate infants’ acquisition of language. An influential study in support of this view is Kuhl et al. (1997), which found that vowels in IDS are produced with expanded first and second formants (F1/F2) on average, indicating that the vowels are acoustically further apart in IDS than in ADS. These results have been interpreted to mean that the way vowels are produced in IDS makes infants’ task of learning vowel categories easier. The present paper revisits this interpretation by means of a thorough analysis of IDS vowels using a large-scale corpus of Japanese natural utterances. We will show that the expansion of F1/F2 values does occur in spontaneous IDS even when the vowels’ prosodic position, lexical pitch accent, and lexical bias are accounted for. When IDS vowels are compared to carefully read speech (CS) by the same mothers, however, larger variability among IDS vowel tokens means that the acoustic distances among vowels are farther apart only in CS, but not in IDS when compared to ADS. Finally, we will show that IDS vowels are significantly more breathy than ADS or CS vowels. Taken together, our results demonstrate that even though expansion of formant values occurs in spontaneous IDS, this expansion cannot be interpreted as an indication that the acoustic distances among vowels are farther apart, as is the case in CS. Instead, we found that IDS vowels are characterized by breathy voice, which has been associated with the communication of emotional affect.

Boundary phenomena and variability in Japanese high vowel devoicing

Devoicing of high vowels (HVD) in Tokyo Japanese applies in two environments—between voiceless consonants, and between a voiceless consonant and a “pause”—and applies variably as a function of a number of factors. The role and definition of “pause” in this process, in terms of a physical pause or prosodic position (word or phrase boundary), remains unclear, as does what is expected when these environments overlap, and why HVD appears to be categorical in some environments and variable in others. This paper addresses three outstanding issues about HVD—the role of “boundary phenomena” (prosodic position and physical pauses), the relationship between the two environments, and the sources of variability in HVD—by examining vowel devoicing in a large corpus of spontaneous Japanese. We use mixed-effects logistic regression to model how boundary phenomena affect the likelihood of devoicing and modulate the effects of other variables, controlling for other major factors, including a measure of gestural overlap. The results suggest that all boundary phenomena jointly affect devoicing rate, and that prosodic phrase boundaries play a key role: variability in HVD looks qualitatively different for phrase-internal and phrase-final vowels, which are affected differently by word frequency, speech rate, and pause duration. We argue the results support an account of HVD as the result of two overlapping vowel devoicing processes, each widely-attested cross-linguistically: devoicing between voiceless consonants, and devoicing before prosodic phrase boundaries. Variability in the application of these two processes can then be partially explained in terms of aspects of phonetic implementation and processing: gestural overlap (Beckman 1996), which often plays a role in reduction processes, and the locality of production planning (Wagner 2012), a recent explanation for variability in the application of external sandhi processes.

Analysis of allophones based on audio signal recordings and parameterization

The aim of this study is to develop an allophonic description of English plosive consonants based on recordings of 600 specially selected words. Allophonic variations addressed in the study may have two sources: positional and contextual. The former one depends on the syllabic or prosodic position in which a particular phoneme occurs. Contextual allophony is conditioned by the local phonetic environment. Co-articulation overlapping in time demands a precise determination of allophonic pronunciation in the context of phonemic transcription. The presented study is focused on creation of speech recordings that may serve for the analysis of allophone variation. Two sets of recordings are prepared. The first one consists of words read by the non-native speakers. Tempo of reading is forced by a teleprompter. In the second case, every word is played back from the recordings of the phonology expert and then the speaker repeats a particular word. The last stage is the assessment of recordings by the same expert. Scores assigned by the expert are included as a reference for signal analysis and parametrization. [Research sponsored by the Polish National Science Centre, Dec. No.2015/17/B/ST6/01874.]

Prosodically-conditioned fine-tuning of coarticulatory vowel nasalization in English

This study explores the relationship between prosodic strengthening and linguistic contrasts in English by examining temporal realization of nasals (N-duration) in CVN# and #NVC, and their coarticulatory influence on vowels (V-nasalization). Results show that different sources of prosodic strengthening bring about different types of linguistic contrasts. Prominence enhances the consonant׳s [nasality] as reflected in an elongation of N-duration, but it enhances the vowel׳s [orality] (rather than [nasality]) showing coarticulatory resistance to the nasal influence even when the nasal is phonologically focused (e.g., mob-bob; bomb-bob). Boundary strength induces different types of enhancement patterns as a function of prosodic position (initial vs. final). In the domain-initial position, boundary strength reduces the consonant׳s [nasality] as evident in a shortening of N-duration and a reduction of V-nasalization, thus enhancing CV contrast. The opposite is true with the domain-final nasal in which N-duration is lengthened accompanied by greater V-nasalization, showing coarticulatory vulnerability. The systematic coarticulatory variation as a function of prosodic factors indicates that V-nasalization as a coarticulatory process is indeed under speaker control, fine-tuned in a linguistically significant way. In dynamical terms, these results may be seen as coming from differential intergestural coupling relationships that may underlie the difference in V-nasalization in CVN# vs. #NVC. It is proposed that the timing initially determined by such coupling relationships must be fine-tuned by prosodic strengthening in a way that reflects the relationship between dynamical underpinnings of speech timing and linguistic contrasts.

Effect of Domain Initial Strengthening on Vowel Height and Backness Contrasts in French: Acoustic and Ultrasound Data.

Phonetic variation due to domain initial strengthening was investigated with respect to the acoustic and articulatory distinctiveness of vowels within a subset of the French oral vowel system /i, e, ɛ, a, o, u/, organized along 4 degrees of height for the front vowels and 2 degrees of backness at the close and midclose height levels. Acoustic and lingual ultrasound data were examined to characterize the production of vowels in intonational phrase initial position, compared with intonational phrase medial position, for 4 speakers. Formant values and estimates of lingual constriction location and degree differed according to the prosodic position independent of vowel duration, with a higher F1 for /a/, a higher F2 for /ɛ/, a backer constriction for /o/ and /a/ but a fronter constriction for /ɛ/, and a narrower constriction for /e, ɛ, u, o/ but a wider constriction for /a/. For most speakers, these variations enlarge the acoustic and/or articulatory distance between members of the pairs /e-ɛ/, /ɛ-a/, /u-o/, /i-u/, and /e-o/ but reduce the distinction within the pair /i-e/. These changes in intonational phrase initial position are vowel dependent and frequently contribute to augmenting the phonetic distinctiveness between vowels contrasting along the height and backness dimensions.

Bidirectional prosodic effects of code-switching in Spanish-Basque bilinguals

This study investigates the prosody of code-switching in Spanish-Basque bilinguals, which has not been previously documented. Peninsular Spanish and Basque have several prosodic differences, including peak alignment in pitch accents. Spanish has delayed peaks (Estebas Plana & Prieto (2008)), while Basque has early peaks (Elordieta & Hualde (2015)). The current study investigates whether code-switched words are conditioned by prosodic position and whether the peak alignment and f0 maximum of a pitch accent differ between monolingual context (Spanish only) and code-switched contexts. Bilinguals (aged 21-31) from a Basque-dominant region (Lekeitio, n = 3) and a Spanish-dominant region (Bilbao, n = 4) participated in the present study. The data was coded for the following predictor variables: prosodic position, peak alignment, and f0 maximum in pitch accents. Results reveal that code-switches occur almost equally in phrase-medial and final positions. For peak alignment, Lekeitio speakers produced early peaks in both code-switching contexts (Spanish-Basque, Basque-Spanish), which matches monolingual norms for Lekeitio varieties of Spanish and Basque. However, Spanish-dominant speakers also produced early peaks in both code-switching contexts, despite delayed peak norms for Peninsular Spanish. Finally, preliminary results reveal greater f0 during code-switching contexts for Lekeitio speakers only. Further results and implications will be discussed.

Production of neutral tones in three Mandarin dialects

Neutral tone (NT) in Mandarin refers to syllables with shorter duration and whose pitch realisations depend on preceding syllables (Chao 1968; Lee and Zee 2008). There are few acoustic studies on NT in Mandarin dialects outside the standard, and on how NT interacts with sentence modality and prosodic position. We report results from a reading task with 3 speakers each from Chongqing, Kunming, and Nanjing (N = 9). A set of 8 syllables with potential NT and control lexical tones are placed in phrase-medial and final positions and in statements and echo questions. Pitch contours of entire utterances are analysed to determine the variation in pitch height, slope and tonal coarticulatory influence on target syllables. We find NT-like behaviour in Chongqing and Kunming, despite the longstanding claim of their absence (Luo and Wang 1981; Zhai 1996): e.g., in Chongqing, NT syllables are high-falling medially, but undergo more pitch reduction than control and exhibit obligatory tonal spreading from preceding tone regardless of speech rate. Nanjing NT interacts strongly with modality: statement-final NT is almost invariably realised as strong creak if preceding tone is mid-falling. We also report hitherto unreported variation and hypothesise on the impact of standard Mandarin on NT in these dialects.

Consonantal timing and release burst acoustics distinguish multiple coronal stop place distinctions in Wubuy (Australia).

Substantial research has established that place of articulation of stop consonants (labial, alveolar, velar) are reliably differentiated using a number of acoustic measures such as closure duration, voice onset time (VOT), and spectral measures such as centre of gravity and the relative energy distribution in the mid-to-high spectral range of the burst. It is unclear, however, whether such measurable acoustic differences are present in multiple place of articulation contrasts among coronal stops. This article presents evidence from the highly endangered indigenous Australian language Wubuy, which maintains a 4-way coronal stop place contrast series in all word positions. The authors examine the temporal and burst characteristics of / t̪ t ʈ/ in three prosodic positions (utterance-initial, word-initial but phrase medial, and word-medial). The results indicate that VOT, closure duration, and the spectral quality of the burst may indeed differentiate multiple coronal place contrasts, in most positions, although measures that distinguish the apical contrast in absolute initial position remain elusive. The authors also examine measures (spectrum kurtosis, spectral tilt) previously used in other studies of multiple coronals in Australian languages. These results suggest that the authors' measures perform at least as well as those previously applied to multiple coronals in other Australian languages.

A preliminary ultrasound study of apical consonants in stressed and unstressed position in Arrernte

In this study, we present ultrasound data from four female speakers of the Australian language Arrernte. We focus on the apical consonants—alveolar /t n l/ and retroflex /ʈ ɳ ɭ/—in stressed and unstressed position. Previous articulatory results for the Arrernte stops using electropalatography and electro-magnetic articulography (for two of the speakers in the current study) showed that the tongue is most retracted for the retroflex in unstressed prosodic context, but that both apicals have a more advanced tongue position in stressed context. In addition, jaw position was highest for the alveolar in stressed position, and lowest for the retroflex in unstressed position. Current ultrasound results show that the rear-most portion of the tongue is consistently further back for the retroflex in unstressed position. This result confirms our previous conclusion that the most prototypical retroflex is the one found in the weaker prosodic position. By contrast, the stressed retroflex has a more forward position of the rear-most portion of the tongue, while the stressed and unstressed alveolar consonants show varying patterns depending on speaker and on manner class. Results are discussed in light of the variability that is often found for apical articulations in Australian languages.

Vowel Reduction in Israeli Heritage Russian

This study examines vowel reduction patterns of Israeli Heritage Russian speakers (IHRs). Contemporary Standard Russian is well documented as having a complex system of vowel reduction (e.g., Barnes, 2002; Crosswhite, 1999; Jakobson, 1929; Padgett, 2004): specifically, underlying /o/ surfaces as [o] in stressed syllables, as [ɐ] in the first pretonic syllable, and as [ə] in other unstressed syllables. In Modern Hebrew, on the other hand, stressed and unstressed vowels differ in duration, but not in quality (Cohen, Silber-Varod, & Amir, in preparation; Maymon, 2001). We conducted a production experiment to determine the patterns of vowel reduction in the Russian of IHRs. Sixteen IHRs were exposed to audiobased forms of real and nonce words with stressed /o/ and were required to produce the forms with and without stressed suffixes. Thus, underlying /o/ was produced in three distinct prosodic positions: stressed (e.g., /nos/ ‘nose sg.’), pretonic (e.g., /nos- ̍ɨ/ ‘nose pl.’), and antepretonic (e.g., /nos-o ̍voj/ ‘nasal.’). The quantity (duration) and quality (F1, F2) of /o/ were acoustically analyzed and compared to a control group of five Russian-speaking adult immigrants to Israel. The results showed that IHRs reduced unstressed /o/ in both real and nonce words, but in producing nonce words they did not display the height contrast that is expected between pretonic and antepretonic vowels. We argue that IHRs’ productions of real words may be rote-learnt, whereas their treatment of nonce words better reflects their productive grammar. We propose that IHRs’ productive system of vowel reduction is a mixed system, combining aspects of their heritage language (i.e., Russian) and dominant language (i.e., Hebrew).

Deriving disyllabic word variants from a Chinese conversational speech corpus.

Motivated by the quasi-categorical reduced forms of disyllabic words produced in Chinese conversational speech, a frequency-based selection procedure of typical pronunciation by disyllabic word type and reduction degree is proposed in this paper. This variant-selection algorithm utilizes techniques of free phone recognition and phonetic similarity score calculation to account for Chinese syllable structure. Four reduction types are suggested by considering the presence of a within-word syllable boundary: Citation form-like reduction, marginal segment deletion, nuclei merger, and syllable merger. The results show that the most frequent reduction types for disyllabic words in Chinese conversation are citation form-like reduction and syllable merger. In particular, high-frequency disyllabic words preferentially take the extreme syllable-merger form. As shown in the analysis, segmental reduction in Chinese disyllabic words is morphology-dependent. It is also related to the prosodic position at which a disyllabic word is produced as well as the temporal quality of the word. Finally, in the automatic speech recognition experiments, the performance was improved by adding a small number of variants selected by the algorithm to the pronunciation dictionary of the system.

Bono Bo and Fla Mingo: Reflections of Speech Prosody in German Second Graders' Writing to Dictation.

Written German is characterized by an underrepresentation of prosody. During writing acquisition, children have to tackle the question which prosodic features are realized by what means – if any. We examined traces of speech prosody in German children’s writing to dictation. A sample of 79 second graders were asked to write down eight sentences to dictation. We analyzed three potential reflections of speech prosody in children’s dictations: (a) Merging of preposition and definite article, potentially preferred after monosyllabic prepositions as in this case preposition and article may melt to the canonical trochaic foot in German. (b) The introduction of orthographically inadequate graphemic border markings within trisyllabic animal names, respecting borders of prosodic units like foot or syllable. (c) Omissions of the definite article in non-optimal prosodic positions, deviating from the preferred strong-weak rhythm. The occurrence of border markings was evaluated via graded perceptual judgments. We found no evidence for inter-word border markings being influenced by prosodic context, probably due to a ceiling effect. However, word-internal markings within animal names, although rarely occurring in general, were clearly influenced by prosodic structure: Most of them were produced at borders of feet or syllables, while significantly fewer markings were perceived at borders of syllable constituents or within consonant clusters. Moreover, we observed significantly more omissions of the definite article in non-optimal prosodic positions compared to potentially optimal positions. Thus, our results provide first evidence from writing acquisition for prosodic influences on writing in a language with scarce graphemic marking of prosody.

Lenition, perception and neutralisation

This paper argues that processes traditionally classified as lenition fall into at least two subsets, with distinct phonetic reflexes, formal properties and characteristic contexts. One type, referred to as loss lenition, frequently neutralises contrasts in positions where they are perceptually indistinct. The second type, referred to as continuity lenition, can target segments in perceptually robust positions, increases the intensity and/or decreases the duration of those segments, and very rarely results in positional neutralisation of contrasts. While loss lenition behaves much like other phonological processes, analysing continuity lenition is difficult or impossible in standard phonological approaches. The paper develops a phonetically based optimality-theoretic account that explains the typology of the two types of lenition. The crucial proposal is that, unlike loss lenition, continuity lenition is driven by constraints that reference multiple prosodic positions.

Articulatory reflexes of the three-way contrast in labial stops and kinematic evidence for domain-initial strengthening in Korean

This study examines articulatory characteristics of the three-way contrast in labial stops /p php*/ (lenis, aspirated, fortis, respectively) in Korean in phrase-initial and phrase-medial prosodic positions with a two-fold goal. First, it investigates supralaryngeal articulatory reflexes of the stops and explores articulatory invariance of these stops across prosodic positions. Second, it investigates Korean stops in kinematic terms from the perspective of domain-initial strengthening, and explores the nature of prosodically-conditioned speech production from a dynamical perspective. Results showed that the articulatory reflex of the three-way contrast was invariantly observed across prosodic positions with lip constriction degree (/p/</ph/</p*/), while lip constriction duration showed a binary distinction (/p/</php*/). Kinematically, there was only very weak articulatory evidence for the contrast across prosodic positions: The V-to-C lip closing movement tended to be faster for /ph/ than for /p/, and the C-to-V lip opening movement tended to be larger for /php*/ than for /p/. As for domain-initial strengthening, the consonantal lip closing gesture was characterized by a larger, longer and slower articulation, whereas the vocalic lip opening gesture (after the release) was larger and faster, but not longer. Kinematic relations indicated that the lip closing movement is most likely controlled by a rate of the clock (possibly modulated by a temporal modulation gesture, or π-gesture) comparable to boundary effects in English, but the boundary-induced lip opening movement was better accounted for by a change in target (possibly modulated by a spatial modulation gesture, or μ-gesture) which was comparable to prominence rather than boundary effects in English. The cross-linguistic difference was interpreted as coming from different prosodic systems between Korean and English, presumably instantiated in dynamical terms of how the temporal and the spatial modulation gestures are phased with constriction gestures in relation to boundary marking versus prominence marking.

The role of gesture duration in the initial unstressed vowel shift in Galician. A first approximation

This study examines a change in the Galician vowel system. This change has been described as the phonologization of the distinction between low‑mid-low and mid-high vowels of Galician in unstressed pretonic position. The principal aim of this study is to test whether this distinction can be triggered by the longer duration of those initial vowels, favoured by the structure of the syllable and by the initial prosodic position. To verify this hypothesis, an articulatory simulation study with TADA and an acoustic study with native speakers of Galician were carried out. Since there are no studies in which the unstressed (neither initial nor medial) vowel system of Galician is examined in terms of acoustical properties, the first step was to determine whether it differs from the stressed vowel system and whether there is an acoustical opposition between mid-low and mid-high vowels in initial position. The results showed that there are significant differences between vowels in unstressed syllable and vowels in stressed syllable and also between the unstressed initial vowels /e/ and /ɛ/, and /o/ and /ɔ/. A linear relation between duration of the gesture and degree of distance between /o/ and /ɔ/ and between /e/ and /ɛ/ was found in the simulation study with TADA, and a significant relation between duration and height was found for /ɔ/ in the acoustic study.

Production of English Voicing Contrast in Different Prosodic Conditions by Tonal and Non-tonal Korean Dialect Speakers

The use of phonetic correlates of Korean stops was found to be different among different dialect speakers, showing that Seoul speakers rely on both voice onset time (VOT) and fundamental frequency (f0), while Kyungsang speakers mainly use VOT in a three-way contrast among voiceless stops of Korean. Drawing on the relative use of phonetic cues by Korean tonal and non-tonal dialect speakers, this paper investigates how the two dialect speakers use phonetic cues for voicing contrast of English stops across different prosodic phrasings and accent conditions. Seoul and Kyungsang dialect speakers were asked to produce English voiceless and voiced stops in IP (intonational phrase) and Wd (word) levels under accented and unaccented conditions. The two dialect speakers showed different use of phonetic cues to signal the voicing contrast of English stops across different prosodic conditions. Compared to Seoul speakers, Kyungsang dialect speakers put more weight on the use of the VOT to enhance the voicing contrast of English stops in the higher prosodic position and accent condition. On the other hand, Seoul and Kyungsang dialect speakers used the onset of f0 in contrasting English voiced and voiceless stops across different prosodic conditions in a similar way. Both dialect speakers showed the significant difference in the onset f0 for English voiced and voiceless stops only in the accented condition regardless of prosodic phrasing. (Hanyang University)

Adult Second Language Learning of Spanish Vowels

The present study reports on the findings of a cross-sectional acoustic study of the production of Spanish vowels by three different groups of speakers: 1) native Spanish speakers; 2) native English intermediate learners of Spanish; and 3) native English advanced learners of Spanish. In particular, we examined the production of the five Spanish phonemic monophthongs (/i e a o u/) in two prosodic positions: stressed and unstressed. Since Spanish and English vary in the nature of the difference between stressed and unstressed vowels, a logical question is: How does the relationship between stressed and unstressed vowels differ across the three groups of participants? Do the learners transfer English patterns to their Spanish? The findings are discussed in terms of their significance for current models of second language speech learning and provide important information regarding the paths of pronunciation development in the second language acquisition of Spanish.