Intervocalic Voicing Research Articles

Roman tablets as linguistic corpora: evidence for phonological variation in 2nd c. Latin

Abstract The Tabula Alimentaria Veleiana and the Tabula Ligures Baebiani are two 2nd century tablets which outline the regional implementation and execution of Trajan’s alimentary scheme. These tablets not only provide some of the longest Latin inscriptions from antiquity, but they also offer unique insight into regional linguistic patterns, as one tablet is from northern Italy and the other is from southern Italy. Given that each tablet demonstrates its own consistent patterns of orthographic peculiarities, it is possible that these communities had local epigraphic traditions. This, in turn, could suggest that consistent spelling anomalies reflect regional phonological patterns, such as intervocalic voicing and consonant degemination. This case study builds on recent literature studying sociolinguistic variation in Latin by using tablets from the same time period as regional corpora with which to study linguistic patterns.

Laryngeal realism and the voicing contrast in Khuzestani Arabic stops

In laryngeal realism (LR), laryngeal specification of stops is explained by direct maps of cues (e.g. VOT) onto privative phonological laryngeal features [voice] or [spread glottis]. Phonetic realization of the segments and speakers’ ‘control’ (e.g. the degree of intervocalic voicing and speech rate manipulation effects on VOT duration) are used as diagnostics of phonological specification. Similar to some Arabic vernacular dialects (e.g. Qatari Arabic), Khuzestani Arabic in Iran presents a case where three voiced stops /b d ɡ/ are in contrast with voiceless stops /p t k/, but two voiceless guttural plosives /tˁ q/ have no voiced homorganic counterparts. In this paper we examine the phonetic realization of voicing in these stops at word-initial and intervocalic position, as well as the effects of speech rate manipulation on VOT and closure voicing. The data came from 12 native speakers recorded in Khorramshahr, Iran. Our findings suggest an over-specified voicing system in this Arabic variety. We found that voiced /b d ɡ/ were produced with voicing lead in initial position and complete closure voicing word medially, voiceless /p t k/ had long lag VOT, while guttural /tˁ q/ had short lag VOT. Speech rate manipulation revealed that only duration of (pre)voicing and duration of aspiration increased in slower speech. Also, f0, F1, and F2 were measured at vowel onset to evaluate the glottal state in production of stops. The results support the predictions of LR that voiced stops are specified by [voice], voiceless stops are [sg], while gutturals lack underlying specification for voice.

The rarity of intervocalic voicing of stops in Danish spontaneous speech

Previous studies of the phonetics of Danish stops have neglected closure voicing. Danish is an aspiration language, but the aspirated stops /p t k/ are produced with shorter closure duration and less articulatory effort than the unaspirated stops /b d ɡ/. Furthermore, all Danish stops are characterized by some degree of glottal spreading during the closure. In this study, we use a corpus of Danish spontaneous speech (DanPASS) to investigate the intervocalic voicing—its distribution across the two laryngeal categories, whether it patterns as a lenition phenomenon, and whether the aerodynamic environment predicts its distribution. We find that intervocalic voicing is not the norm for either set of stops and is particularly rare in /p t k/. Voiced tokens are mostly found in environments associated with lenition. We suggest that the glottal spreading gesture found in all Danish stops is a phonological mechanism blocking voicing, which is probabilistically lost in spontaneous speech. This predicts our results better than relying on laryngeal features like [voice] or [spread glottis]. The study fills a gap in our knowledge of Danish phonetics and phonology, and is also one of the most extensive corpus studies of intervocalic stop voicing in an ‘aspiration language.’

Change of Tōhoku dialect spoken in Hawaii

AbstractThis study investigates changes in Tōhoku dialect speakers’ phonology after their immigration to Hawaii, specifically concerning intervocalic voicing and alveolar/palatal mergers. Tōhoku dialect is known for its unique phonology compared to other Japanese dialects and, for this reason, it is often stigmatized. Previous studies of second dialect acquisition have suggested that older speakers tend to retain the phonological features of their original dialects during dialect contact situations. The results from adult Japanese plantation immigrants, as expected, suggested that adult Tōhoku dialect speakers demonstrated limitations in acquiring second dialect phonology in their contact with non-Tōhoku dialect speakers. However, there are different degrees of second dialect acquisition between the intervocalic voicing and alveolar/palatal mergers among the Tōhoku dialect immigrants who interacted with non-Tōhoku dialect speakers on a daily basis and those who did not; namely, the former eliminated the stereotypical stigmatized Tōhoku dialect feature – the mergers – more than those speakers who did not have such frequent interaction. This suggests that both sociolinguistic factors, e.g., dialect discrimination by peers, as well as relative phonological complexity, influence the rates of second dialect acquisition.

Voicing contrasts in the stops of Indian English produced by Assamese speakers

Speakers of most American and British varieties of English contrast word-initial stops using aspiration, with long-lag Voice Onset Time (VOT) for /ptk/ and short-lag or lead voicing for /bdg/ (Docherty 1992, Chodroff & Wilson 2017, etc.). However, Indian English (IndE) speakers of Hindi and Telugu backgrounds living in the US produce a short-lag versus lead voicing contrast instead (Davis & Beckman 1983, Sirsa & Redford 2013). We recorded 10 L1 Assamese speakers residing in India reading English wordlists, all bilinguals who attended English-medium schools starting at age 4.3 (±1.3), young enough to acquire the target English of their community (Flege 1991). We are measuring the VOT of ten tokens each of /ptkbdg/ in word-initial position. Results thus far support the interpretation of the contrast in IndE as voicing not aspiration, as unaspirated voiceless stops versus pre-voiced stops. We also measure two expected consequences of using [voice] for the contrast: full voicing of sonorant consonants after voiceless obstruents in onset clusters and full (>75% of closure), rather than passive, intervocalic voicing. Though neither occurs in aspirating languages (Beckman et al. 2013), we find both in the IndE of Assamese speakers, suggesting IndE has developed a distinct contrast from AmE/BrE.

Postaspiration in Sevillian Spanish: Beyond voiceless stops

Sevillian Spanish /s + ptk/ clusters are often produced with gemination of the consonant following /s/ and postaspiration (/pasta/:[pahta]-->[pat:ha]) (Ruch and Peters, 2016). Spanish voiceless stops are unaspirated, but some argue that postaspirated stops are phonologizing in Seville (O'Neill, 2010; Gylfadottir, 2015). Acoustically, postaspiration is long VOT. Articulatorily, it is described as a laryngeal gesture moving across the stop, continuing past the release (Torreira, 2007). Previous work focuses only on /s + ptk/. This acoustic study examines /sC/-clusters with voiceless stops, voiced stops, and sonorants, extending our understanding of the mechanisms behind postaspiration. Seven Sevillians read paragraphs containing /sC/ and /C/ words (/kaspa/-/kapa/). All clusters show gemination. Compared to /ptk/, /s + ptk/ clusters have long postaspiration, resist intervocalic voicing, and result in higher pitch on the following vowel (a cue to contrastive aspiration in many languages, Dmitrieva et al., 2015). /s + bdg/ clusters have stronger constriction, higher COG, and less voicing than intervocalic (spirantized) /bdg/. /s + mnl/ clusters differ from /mnl/ only in gemination. Gestural overlap thus occurs in voiced and voiceless clusters, and appears to result from higher-level realignment as opposed to unintentional mistiming (Parrell, 2012). Postaspirated voiceless stops are unlikely to phonologize because the process occurs with both /s + bdg/ and /s + ptk/, giving different results.

Effects of L1 on the voicing of intervocalic voiceless stops and its relation with L2 proficiency

This study investigates the correlation between the proficiency of L2 learners of English and the degree of L1 Korean intervocalic voicing rule transfer on the English intervocalic voiceless stop production by Korean English learners. Korean has a phonological rule that voiceless stops become voiced between vowels (Korean intervocalic voicing rule) but this rule does not apply in English. The aforementioned difference between the phonological rule of the two languages makes it possible to predict that Korean learners with lower L2 proficiency might produce English intervocalic voiceless stops with higher degree of voicing than the learners with higher L2 proficiency. The speech data of Korean L2 learners of English is extracted from GenieSpeeCor, rated and categorized English speech corpus of Korean L2 learners of English production, and the value of intensity drop and closure duration were used as acoustic cues for voicing difference. Alveolar stop is excluded from the data analysis because the flapped /t/ sound is difficult to distinguish from the voiced sound. As a result of the current study, when it comes to the closure duration as an acoustic cue, there was no meaningful statistical relation between learners' L2 proficiency and the degree of intervocalic stops voicing. However, in terms of the intensity drop learners with the highest L2 proficiency showed a lower degree of voicing than the learners with middle and lowest proficiency.This study investigates the correlation between the proficiency of L2 learners of English and the degree of L1 Korean intervocalic voicing rule transfer on the English intervocalic voiceless stop production by Korean English learners. Korean has a phonological rule that voiceless stops become voiced between vowels (Korean intervocalic voicing rule) but this rule does not apply in English. The aforementioned difference between the phonological rule of the two languages makes it possible to predict that Korean learners with lower L2 proficiency might produce English intervocalic voiceless stops with higher degree of voicing than the learners with higher L2 proficiency. The speech data of Korean L2 learners of English is extracted from GenieSpeeCor, rated and categorized English speech corpus of Korean L2 learners of English production, and the value of intensity drop and closure duration were used as acoustic cues for voicing difference. Alveolar stop is excluded from the data analysis because the flapped /...

Tongue root configuration during Seoul Korean stops: An ultrasound study

Tongue root advancement facilitates voicing during stop closure by enlarging the supralaryngeal cavity volume (Westbury, 1983). In a recent ultrasound study, Ahn (2018) reports that tongue is indeed more advanced during voiced than voiceless stops both in English and Brazilian Portuguese, suggesting that the articulatory adjustment aligns more with the abstract laryngeal distinctions than their acoustic implementation—the voiced” stops are typically not phonated in English but in Brazilian Portuguese. This study, using ultrasound, compares tongue positioning during Seoul Korean (SK) stops of three laryngeal categories: lenis, fortis, and aspirated. All three categories are voiceless phrase-initially, with the lenis being voiced intervocalically (Jun, 1993). This study asks whether (1) SK lenis, fortis, and aspirated stops have different tongue configurations when none are phonated and (2) the intervocalic voicing of lenis stops leads to tongue root advancement. Nine native SK speakers recorded phrase-initial and intervocalic stops. Results: lenis, fortis, and aspirated stops did not show different tongue position in phrase-initial positions. In intervocalic positions, lenis stops, acoustically voiced during closure, did not show more tongue root advancement than other types of stops. These results suggest that SK speakers use tongue positioning neither for laryngeal contrast nor as an adjustment for allophonic voicing of intervocalic lenis stops.Tongue root advancement facilitates voicing during stop closure by enlarging the supralaryngeal cavity volume (Westbury, 1983). In a recent ultrasound study, Ahn (2018) reports that tongue is indeed more advanced during voiced than voiceless stops both in English and Brazilian Portuguese, suggesting that the articulatory adjustment aligns more with the abstract laryngeal distinctions than their acoustic implementation—the voiced” stops are typically not phonated in English but in Brazilian Portuguese. This study, using ultrasound, compares tongue positioning during Seoul Korean (SK) stops of three laryngeal categories: lenis, fortis, and aspirated. All three categories are voiceless phrase-initially, with the lenis being voiced intervocalically (Jun, 1993). This study asks whether (1) SK lenis, fortis, and aspirated stops have different tongue configurations when none are phonated and (2) the intervocalic voicing of lenis stops leads to tongue root advancement. Nine native SK speakers recorded phrase-ini...

Intervocalic consonant voicing vs. lengthening in two Athabaskan languages

For over a century, various authors have described intervocalic consonants in Athabaskan languages as being long (e.g., Sapir 1914 on Chasta Costa, Cook 2004 on Dene Sųɬine (Chipewyan), McDonough and Ladefoged 1993 on Navajo). Such descriptions, even if instrumental, have generally not been very detailed, and have not adequately investigated the factors which may influence lengthening, such as stress and position in word. In this presentation, I first present results of an investigation of effects of position and stress on the duration of consonants with internal cues to duration (fricatives and consonantal sonorants) in two Athabaskan languages, Deg Xinag and Kwadacha Tsek'ene. Results show that intervocalic consonants are longer before stressed than unstressed syllables. Next, I present results of an instrumental study of [t] which shows that a competing and antagonistic process to intervocalic lengthening, intervocalic voicing, is also found in both languages. Intervocalic stops are voiced through approximately 80% of their closure duration before unstressed vowels in both languages. Intervocalic voicing of stops has been reported for Tahltan (Bob 1997) and Dene Sųɬine (McDonough and Wood 2008) but generally not for other Athabaskan languages except for those in which nasals have evolved into voiced obstruents (e.g., Jicarilla Apache, Tuttle 2000).For over a century, various authors have described intervocalic consonants in Athabaskan languages as being long (e.g., Sapir 1914 on Chasta Costa, Cook 2004 on Dene Sųɬine (Chipewyan), McDonough and Ladefoged 1993 on Navajo). Such descriptions, even if instrumental, have generally not been very detailed, and have not adequately investigated the factors which may influence lengthening, such as stress and position in word. In this presentation, I first present results of an investigation of effects of position and stress on the duration of consonants with internal cues to duration (fricatives and consonantal sonorants) in two Athabaskan languages, Deg Xinag and Kwadacha Tsek'ene. Results show that intervocalic consonants are longer before stressed than unstressed syllables. Next, I present results of an instrumental study of [t] which shows that a competing and antagonistic process to intervocalic lengthening, intervocalic voicing, is also found in both languages. Intervocalic stops are voiced through appr...

Variable production and indexical social meaning: On the potential physiological origin of intervocalic /s/ voicing in Costa Rican Spanish

Abstract In several dialects of Spanish, men tend to exhibit more intervocalic /s/ voicing than women, e. g., oso ‘bear’ as [ozo], and this difference may have a physiological basis. File-Muriel et al. (2015, Disentangling the physiological from the socially-learned in gradient, sociophonetic processes: Evidence from s-realization in Barranquilla, Colombia. Unpublished manuscript) found that vocal tract size conditions /s/ aspiration in Barranquilla, and Nadeu and Hualde (2013, Reinterpretation of biomechanics as gender-conditioned variation in the origin of diachronic intervocalic voicing. Available at http://washo.uchicago.edu/pub/workshop/nadeu.pdf) contend that speakers with larger vocal tracts may have greater difficulty controlling vocal fold cessation. The present work serves as a continuation of these studies, utilizing 18 sociolinguistic interviews to determine (i) what factors are most predictive of intervocalic [z] in Costa Rica and (ii) whether physiology can potentially explain its origin. The results of a statistical analysis using 1,647 tokens of /s/ show that both gender and physiological factors significantly condition voicing (p < 0.001), with more voicing in men’s speech, as F2 decreases, and as f0 decreases. However, one would expect more gradient voicing in men’s speech if physiological factors caused the gender-based voicing difference, but women voice more gradiently while men produce higher rates of 0 % and 100 % voicing. We conclude that while physiological factors may have been its original source, non-physiological factors currently condition /s/-voicing in Costa Rica, with male speakers aiming for categorical targets for social motivations.

Fricative patterning in aspirating versus true voice languages

Building on the empirical insights of Beckman, Jessen & Ringen (2013), we compare the fricatives within the laryngeal systems of Russian and Turkish on the premise that the former is a final devoicing language, while the latter is not, but instead has alternations based on processes of intervocalic voicing and final fortition. This view has consequences for the analysis of fricatives in Russian vs. Turkish: Russian fricatives undergo final devoicing, while Turkish fricatives do not. By contrast, unlike Russian fricatives, Turkish fricatives induce [spread glottis] assimilation in following sonorants. We show that these differences are upheld in three phonetic studies, extending the relevance of the ‘laryngeal realism’ hypothesis to fricatives as well as stops.

Biomechanically Conditioned Variation at the Origin of Diachronic Intervocalic Voicing.

This paper addresses the question of how synchronic variation in intervocalic voicing of voiceless obstruents, as observed in several languages (e.g., Rome Italian /lato/ [lato] - [lado]), may initiate and give rise to a regular sound change (e.g., /t/ > /d/ between vowels). We hypothesize that a biomechanically motivated linkage between male gender, speech rate, and voicing may provide a way to accelerate the spread of the phenomenon and lead to an eventual generalized recategorization. In order to explore this hypothesis, first we reanalyze the results of a previous study on intervocalic voicing in Spanish, focusing on individual differences and, in particular, the possible role of gender. Then we report on a study of the same phenomenon in Basque, focusing also on interspeaker variation. Finally, we report on a controlled experiment where speech rate was manipulated.

Manner asymmetries in Central Catalan pre-vocalic voicing

In Central Catalan, word-final sibilants and stop + sibilant clusters undergo voicing when followed by a vowel in the next word. In word-final pre-vocalic singleton stops, however, no voicing is observed. This asymmetry is unexpected from the point of view of phonetic naturalness, and it invites a closer empirical investigation, especially in the light of recent phonetic findings which challenge earlier descriptions of a similar voicing process before sonorant consonants in Catalan. This study presents a systematic acoustic comparison of voicing in pre-vocalic stops /p, b/, sibilants /s, z/ and stop + sibilant clusters /bz/ with voicing before obstruents and before sonorant consonants. The results show that while pre-sonorant voicing is limited and highly variable, pre-vocalic sibilant and cluster voicing are both robust categorical processes for most speakers. Drawing on evidence from how voicing is realised in a subset of gradient cases, I propose that pre-vocalic cluster voicing developed from pre-vocalic sibilant voicing which in turn descended from an earlier intervocalic voicing process. I further question whether the outcome of the sound changes in Catalan can be synchronically analysed using grounded markedness constraints, and to what extent the pre-vocalic voicing may be regulated by universal vs. language-specific mechanisms of phonological abstraction.

The voicing of intervocalic stops in Old Tuscan and probabilistic sound change

AbstractThe article examines the dual outcome of Latin intervocalic /p t k/ in Old Tuscan. While these stops remain voiceless in most of the Tuscan lexicon, a significant number of words are voiced. The prevailing opinion is that preservation of voicelessness is the native outcome, and that the words displaying voicing were borrowed from Romance languages in which intervocalic voicing was systematic. However, a number of facts militate against this hypothesis, including the presence of voicing in Tuscan words that are unattested or maintain a voiceless stop in the supposed donor languages. The results of a corpus-based analysis of an Old Tuscan lexicon show that the distribution of the voiced outcomes is phonologically conditioned (contra the borrowing hypothesis), as they are more likely if the stop was velar, next to low(er) vowels, next to stress and if followed by a liquid consonant. These results are also an example of sound change with clear and fine-grained phonological conditioning but a non-systematic outcome; irregularity was probably caused by the allophonic nature of Old Tuscan voicing and by its interaction with another lenition process.

A fraction too much friction

Typological work shows that voiced fricatives like /β ð/ occur more often without their voiceless counterparts than with them, contrary to what would be expected on the basis of markedness relations between voicing and obstruents. This paper suggests that many of the offending fricatives are more appropriately viewed as sonorants, whose unmarked status is to be voiced. This view has an important consequence for the interpretation of intervocalic voicing (e.g. afa > ava), which we suspect is the diachronic origin of most of the fricatives in our corpus. We propose that intervocalic voicing is sonorization, formalized in terms of the suppression of melodic material.

The Effect of Focus on the Realization of Intensity in Chinese

In this paper, the average intensity and the effect of focus on it for disyllabic words in Chinese is analyzed. It is found that, the average intensity of rhyme in the first syllable is comparatively great, while for the onset, due to intervocalic voicing, that in the second syllable is great. Focus has great effect on intensity. Analysis shows that, compared to the onset and the first syllable, the effect on the rhyme and on the second syllable is great.

An Acoustic Study on Focus in Chinese: The Variation of Intensity

In this paper, the average intensity and the effect of focus on it for disyllabic words in Chinese is analyzed. It is found that, under focused condition, the average intensity of rhyme in the first syllable is comparatively great, while for the onset, due to intervocalic voicing, that in the second syllable is great. Focus has great effect on intensity. Under focused condition, the overall intensity will increase. The results from this study will be applicable in speech engineering.

Empirical evidence for laryngeal features: Aspirating vs. true voice languages

It is well known that German utterance-initial lenis stops are voiceless but that German intervocalic (or intersonorant) lenis stops are sometimes produced with voicing. This variable voicing can be understood aspassive voicing, voicing that results because of the voiced context, rather than fromactive voicinggestures by speakers. Thus, speakers are not actively aiming to voice intervocalic stops, just as they are not actively aiming to voice utterance-initial stops (Jessen & Ringen 2002, Jessen 2004). If this is correct, the variable voicing that occurs in aspirating languages should be different from the voicing that occurs in true voice languages (such as Russian), in which speakers are actively aiming to voice both initial and intervocalic lenis stops. Since there is little data on the relative amount of intervocalic voicing in true voice languages, however, it has been difficult to evaluate this prediction. The purpose of this paper is to compare data on the voicing of intervocalic stops in German and English with data on the voicing of intervocalic stops in true voice languages. We find that the differences are substantial, supporting the claim that aspirating languages are not like true voice languages, in which the feature of contrast is [voice].

Cue Integration in the Perception of Korean Fricatives

Recent literature on Korean fricatives has shown that the perception of the plain /s/ over the fortis /s*/ correlates to low f0 onset of the following vowel, despite the fact that there is no such correlation in the production of /s/ over /s*/. The integration of f0 with other low frequency cues that covary more reliably with /s/ presents a possible explanation for this disparity between production and perception. In a fixed classification experiment, native Korean speakers distinguished between stimuli that differed in intervocalic voicing and f0 onset following /s/. Results showed the integration of these two cues into a single low-frequency intermediate perceptual property.

Voicing in Russian Stops: Cross-Linguistic Implications

This paper presents the results of an investigation of voicing in utterance-initial and intervocalic stops in monolingual Russian speakers. Prevoicing was found in over 97% of the lenis stops; over 97% of the intervocalic stops were fully voiced. Utterance-initial fortis stops were pronounced as voiceless unaspirated and had short positive VOT. Intervocalic fortis stops were completely voiceless except for a short voicing tail into closure. These results are relevant for typological studies of voicing. Some studies of languages with a two-way contrast between initial stops with prevoicing and short lag VOT have reported that prevoicing is less robust than what might be expected. These findings have been attributed to influence from another language without prevoicing. Our results with monolingual speakers of Russian support these claims. Our results are also relevant for the debate about the laryngeal feature in aspirating languages, which often have some voicing of intervocalic lenis stops. Such voicing has been attributed to passive voicing, in contrast with active voicing that occurs in true voice languages such as Russian. We found that the voicing in Russian is much more robust than the intervocalic voicing in aspirating languages. This difference is explained if the features of contrast are different in the two types of languages: [voice] in the case of Russian and [spread glottis] in the case of aspirating languages.