Hubbry Logo
No audible releaseNo audible releaseMain
Open search
No audible release
Community hub
No audible release
logo
7 pages, 0 posts
0 subscribers
Be the first to start a discussion here.
Be the first to start a discussion here.
Contribute something
No audible release
No audible release
No audible release
View on Wikipedia
from Wikipedia
No audible release
◌̚
Encoding
Entity (decimal)̚
Unicode (hex)U+031A

A stop consonant with no audible release, also known as an unreleased stop, checked stop or an applosive, is a plosive with no release burst: no audible indication of the end of its occlusion (hold). In the International Phonetic Alphabet, lack of an audible release is denoted with an upper-right corner diacritic (U+031A ◌̚ COMBINING LEFT ANGLE ABOVE) after the consonant letter, as in: [p̚], [t̚], [k̚].[1]

Audibly released stops, on the other hand, are not normally indicated. If a final stop is aspirated, the aspiration diacritic ⟨◌ʰ⟩ is sufficient to indicate the release. Otherwise, the "unaspirated" diacritic of the Extended IPA may be employed for this: apt [ˈæp̚t˭].

English

[edit]
Word-final unreleased stops in American English
Problems playing this file? See media help.

In most dialects of English, the first stop of a cluster has no audible release, as in apt [ˈæp̚t], doctor [ˈdɒk̚tə], or logged on [ˌlɒɡ̚dˈɒn]. Although such sounds are frequently described as "unreleased", the reality is that since the two consonants overlap, the release of the former takes place during the hold of the latter, masking the former's release and making it inaudible.[2] That can lead to cross-articulations that seem very much like deletions or complete assimilation.

For example, hundred pounds may sound like [ˈhʌndɹɨb ˈpʰaundz] but X-ray[3] and electropalatographic[4] studies demonstrate that since inaudible and possibly-weakened contacts may still be made, the second /d/ in hundred pounds does not entirely assimilate a labial place of articulation but co-occurs with it.

In American English, a word-final stop is typically unreleased; that is especially the case for /t/,[5] but in that position, it is also analyzed as experiencing glottal reinforcement.

Such sounds may occur between vowels, as in some pronunciations of out a lot. The overlap there appears to be with a glottal stop, [t̚ʔ]: the /t/ is pronounced, and since it is between vowels, it must be released. However, its release is masked by the glottal stop.[6] The term for this is t-glottalization.

The term "unreleased" is also used for a stop before a homorganic nasal, as in catnip. In such cases, however, the stop is released as a nasal in a nasal release and so it would be more precisely transcribed [ˈkætⁿnɪp].

Other languages

[edit]

In most languages in East and Southeast Asia with final stops, such as Cantonese,[7] Hokkien,[8] Korean,[9] Malay,[10] Thai,[11] and West Coast Bajau,[12] the stops are not audibly released: mak [mak̚]. That is true even between vowels. That is thought to be caused by an overlapping glottal stop[6] and is more precisely transcribed [mak̚ʔ]. A consequence of an inaudible release is that any aspirated–unaspirated distinction is neutralized. Some languages, such as Vietnamese,[citation needed] which are reported to have unreleased final stops, turn out to have short voiceless nasal releases instead. The excess pressure is released (voicelessly) through the nose and so there is no audible release to the stop.

Formosan languages

[edit]

The Formosan languages of Taiwan, such as Tsou and Amis realize all obstruents as released but not aspirated, as in Tsou [ˈsip˭tɨ] "four" and [smuˈjuʔ˭tsu] "to pierce", or Amis [tsᵊtsaj] "one" and [sᵊpat˭] "four".[citation needed]

Gyalrong languages

[edit]

In Gyalrongic languages, plosives and nasal stops could be unreleased after a glottal stop,[13] for example:

  • /pʰaroʔk/ > [pʰaˈrɔʔk̚]
  • /təwaʔm/ > [t̪əˈwaʔm̚]

Pirahã

[edit]

In Pirahã, the only surviving dialect of the Mura language, there is a special register of speech using solely humming, which does not involve an audible release and may be transcribed as [m̚] of different length and tone.[14]

Munda

[edit]

Munda languages such as Santali are characterized by checked syllable-final plosives that are both unreleased and glottalized.

See also

[edit]

References

[edit]

Sources

[edit]
  • Browman, Catherine P.; Goldstein, Louis (1990), "Tiers in articulatory phonology, with some implications for casual speech", in Kingston, John C.; Beckman, Mary E. (eds.), Papers in laboratory phonology I: Between the grammar and physics of speech, New York: Cambridge University Press, pp. 341–376
  • Choo, Miho; O'Grady, William D. (2003), The Sounds of Korean: A Pronunciation Guide, Honolulu: University of Hawaii Press
  • Nolan, Francis (1992), "The descriptive role of segments: Evidence from assimilation.", in Docherty, Gerard J.; Ladd, D. Robert (eds.), Papers in laboratory phonology II: Gesture, segment, prosody, New York: Cambridge University Press, pp. 261–280
  • O'Neill, Gareth (2014). "Humming, whistling, singing, and yelling in Pirahã context and channels of communication in FDG1". Pragmatics. 24 (2): 349–375. doi:10.1075/prag.24.2.08nei.
  • Zsiga, Elizabeth (2003), "Articulatory Timing in a Second Language: Evidence from Russian and English", Studies in Second Language Acquisition, 25: 399–432, doi:10.1017/s0272263103000160, S2CID 5998807
[edit]
Revisions and contributorsEdit on WikipediaRead on Wikipedia

No audible release

View on Grokipedia
from Grokipedia
In , no audible release refers to the articulation of a —such as /p/, /t/, /k/, /b/, /d/, or /g/—in which the complete oral closure blocks airflow without a subsequent audible burst or noise upon separation of the articulators, often due to sustained closure or overlap with adjacent sounds. This variant, also termed an unreleased stop or applosive, is a common feature of natural speech and is denoted in the International Phonetic Alphabet (IPA) by the corner [◌̚] placed after the symbol, as in [t̚] for an unreleased alveolar stop. Unreleased stops frequently appear in , particularly at word or boundaries, in clusters, or before another stop, where the release is either suppressed or inaudible because it coincides with the onset of the next segment. In English, for example, word-final stops are routinely unreleased, as in "cat" transcribed as [kæt̚], "button" as [ˈbʌt̚n̩], or "widen" as [ˈwajd̚n̩], reflecting articulatory economy rather than a phonemic distinction. Studies of spontaneous speech show that stops are often unreleased due to overlap between closures, with releases absent in up to substantial portions of utterances, especially in non-homorganic sequences like stop-plus-stop. Although typically allophonic in like English—where released and unreleased forms contrast only phonetically—unreleased stops can be phonologically significant in other languages, such as Karitiana (a ), where voiceless stops are systematically unreleased word-finally, identified acoustically by transitions and prolonged closure durations rather than bursts. In such cases, production involves firmer articulatory contact and gradual pressure dissipation, sometimes with glottal reinforcement to "hold the breath" during closure. This feature influences perception, as listeners rely on cues like preceding transitions or following segment contexts to distinguish unreleased stops, highlighting their role in both phonetic detail and cross-linguistic variation.

Overview

Definition

A stop , also known as a , involves three main phases: a complete closure of the vocal tract that blocks , a hold phase during which builds behind the closure, and a release phase in which the articulators separate to allow the compressed air to escape, typically producing an audible burst of turbulent noise. No audible release occurs when the release phase lacks a perceptible burst, either because the articulators do not fully separate or the airflow is insufficient to generate turbulent sound, resulting in the stop ending abruptly without the characteristic . This distinguishes it from fully released stops, where the release burst provides key acoustic cues for identification. Such consonants are referred to as unreleased stops, and in certain linguistic traditions, as checked stops—particularly in contexts involving syllable-final position—or applosives, emphasizing the absence of the explosion. Common examples include the voiceless bilabial [p̚], alveolar [t̚], and velar [k̚], as well as their voiced counterparts [b̚], [d̚], and [g̚]; in the , the ̚ (a small left-angle mark above the symbol) denotes the lack of audible release.

Phonological Significance

In many languages, unreleased stops function as allophonic variants of their released counterparts, particularly in syllable-final positions where the lack of audible release is predictably conditioned by prosodic environment. This allophony often leads to neutralization of contrasts, such as voicing distinctions, restricting the inventory in codas to a single unreleased form without altering lexical meaning. For instance, in non-tonal languages like Korean, final stops across places of articulation converge phonetically into unreleased realizations due to aerodynamic factors like lowering, which prevents burst production while preserving the underlying phonemic categories. In contrast, certain tonal languages exhibit phonemic status for unreleased stops, where their presence in codas creates meaningful distinctions through association with checked tones. These checked tones, characterized by abrupt termination via an unreleased stop, contrast with unchecked tones in open or sonorant-closed syllables, impacting word meaning and contributing to the overall tonal inventory. Such systems, common in Sino-Tibetan and Austronesian languages, rely on the unreleased coda not only for perceptual cues like shortened duration but also for maintaining phonemic oppositions in dense tonal paradigms. The phonological implications of unreleased stops extend to syllable structure, where they frequently demarcate coda positions and enforce constraints on resyllabification or in consonant sequences. This role reinforces prosodic boundaries, as unreleased codas inhibit airflow continuation into adjacent onsets, thereby stabilizing syllable integrity without invoking . In languages with aspiration contrasts, unreleased stops typically align with unaspirated series, further integrating them into broader laryngeal feature systems while avoiding overlap with aspirated or glottalized variants.

Phonetic Characteristics

Articulatory Mechanics

The production of unreleased stops involves a complete oral closure where the active and passive articulators—such as the for bilabials, the tongue tip and alveolar ridge for alveolars, or the tongue body and velum for velars—come into firm contact, fully obstructing the through the vocal tract. This closure is sustained long enough for intraoral air pressure to build up behind it due to continued pulmonic from the lungs. Unlike released stops, however, the articulators do not separate or do so without producing an audible burst of frication or aspiration; instead, the closure is maintained until the articulators transition to the next segment or the utterance ends, resulting in no perceptible release. Unreleased stops commonly occur in articulatory contexts involving overlap or suppression of the release gesture. In consonant clusters, coarticulation allows the closure for the following consonant to begin before the release of the preceding one, effectively blocking any audible separation of the articulators for the first stop. For instance, in sequences like English [kt] as in "catnip," the velar closure for /k/ holds while the alveolar closure for /t/ forms, preventing the /k/ release. In word-final positions, the lack of a following segment eliminates the need for release, so speakers maintain the closure without separation, as the airflow simply ceases or transitions silently. In some languages and dialects, unreleased stops may involve glottal approximation or closure, where the vocal folds adduct to form a secondary , overlapping the oral closure and further inhibiting any oral release by equalizing pressures or ceasing airflow at the . This glottal involvement is particularly noted in coda positions, creating a glottalized unreleased stop without an audible oral burst, though the typically remains open in non-glottalized cases to allow buildup. The ease of producing unreleased stops varies by place of articulation due to differences in articulator mass, positioning, and control. Bilabial stops like [p̚] are more frequently unreleased than alveolar or velar ones in American English.

Acoustic and Perceptual Features

Unreleased stops exhibit a distinctive acoustic profile marked by the complete absence of a release burst, which in released stops consists of a transient noise segment typically lasting 10-50 ms and providing cues to place of articulation. This absence results in an abrupt cessation of the preceding vowel's formant transitions, without the characteristic transitional noise or frication that accompanies release, leading to a silent closure phase followed directly by silence or the next sound. In spectrographic analysis, this manifests as a sharp halt in formant trajectories—such as F2 lowering for velar stops or rising for alveolars—without the vertical striations or broadband energy spike indicative of a burst, allowing visualization of the unreleased closure's duration, often 15-20 ms longer than in released counterparts. Perceptually, listeners identify unreleased stops primarily through cues other than the burst, including the duration of the oral closure, which tends to be extended (averaging 60-110 ms depending on place and ), and shortened voice onset time approaching zero, as the lack of release eliminates the lag between consonant offset and any subsequent voicing. Contextual information from adjacent segments, such as transitions from the preceding or coarticulatory effects in following sounds, further aids differentiation of , with studies demonstrating high identification accuracy when these cues are intact. However, perception of unreleased stops presents challenges, particularly in noisy environments where subtle cues are masked, reducing intelligibility compared to released stops, especially for velar places. Non-native listeners often struggle more due to reliance on burst cues in their L1, leading to lower discrimination rates than natives, though boundaries remain robust, with listeners grouping stimuli into discrete categories based on closure and transition thresholds. The articulatory basis for this acoustic silence—sustained closure without release—underlies these perceptual dynamics but is secondary to the audible signals here.

Examples in English

Word-Final Positions

In , voiceless stops /p, t, k/ are typically produced without audible release in word-final positions, as in the realization [kʰæt̚] for "," where the articulators remain in place at the closure without a subsequent burst. For /t/ specifically, as [ʔ] is a common variant in this context, particularly in casual speech. Phonetic analyses of spontaneous speech reveal high rates of unrelease for word-final stops, with approximately 70-80% occurring without release, especially in phrase-medial environments where prosodic factors like juncture strength reduce the likelihood of a burst. For instance, Zsiga (2000) found release rates as low as 20% for word-final stops followed by a in , attributing this to articulatory overlap and prosodic timing constraints that prioritize efficiency over full realization. Dialectal differences influence release patterns; while unreleased stops predominate in casual , speakers tend to produce more releases in careful or formal styles, though unreleased variants remain common in informal contexts across both varieties. Unreleased word-final stops interact with intonation and prosody by enabling concise rhythmic structure, as the absence of a release burst avoids prolonging closure, thereby supporting English's stress-timed without disrupting overall timing or phrasing.

Consonant Clusters and Syllable Boundaries

In English consonant clusters, the release of a preceding stop consonant is often inaudible because it is temporally overlapped or masked by the articulatory closure of the following consonant, resulting in no audible burst. This phenomenon is particularly prevalent in stop-stop sequences within the same syllable or across syllable boundaries, where the gestures for the two consonants overlap substantially, preventing the transient release energy from being perceptible. For instance, the bilabial stop in "apt" is realized as [æp̚t], with the /p/ closure released silently into the /t/ closure, and the alveolar stop in "button" as [ˈbʌt̚n], where the /t/ release is obscured by the nasal /n/ onset. At syllable boundaries, unreleased stops are common in coda-onset junctions, where the coda stop of one transitions directly into the onset of the next without an intervening , facilitating resyllabification processes in . A representative example is "lucky," transcribed as [ˈlʌk̚.kɪ], where the /k/ in the coda position lacks an audible release before the following /k/ in the onset, contributing to smoother prosodic flow. Acoustic analyses confirm this pattern, showing substantial gestural overlap in such clusters, with the second often occupying over 50% of the cluster duration, which minimizes audible release transients. Empirical acoustic studies of spontaneous speech demonstrate that unreleased stops predominate in pre-consonantal contexts, with approximately 75% of stops followed by another consonant exhibiting no audible release, compared to isolated word-final stops that more frequently show positive voice onset time (VOT) values of 30–100 ms for voiceless stops. In clusters, the absence of a release burst corresponds to effectively zero VOT for the preceding stop, as the articulatory release coincides with or is hidden by the subsequent closure, distinguishing these from released variants with measurable positive VOT. and cluster directionality further modulate this, with fewer releases in back-to-front sequences like /kt/ due to increased overlap. Exceptions to no audible release in clusters occur rarely, primarily in hyperarticulated or emphatic speech, where speakers produce clearer releases to enhance intelligibility, increasing release rates to around 30–50% in careful elicitation tasks versus spontaneous production. These instances are context-dependent and do not alter the default unreleased pattern in typical connected speech.

Examples in Other Languages

East and Southeast Asian Languages

In , unreleased stops (/p̚/, /t̚/, /k̚/) are phonemically significant in syllable codas, occurring exclusively in checked syllables that distinguish three short tones through pitch levels on the preceding vowel: high level (tone 1, [˥]), mid level (tone 3, [˧]), and low level (tone 6, [˨]). These unreleased codas create abrupt syllable endings, marking the checked tone category and preventing merger with the six contour tones of open syllables. The unreleased /t̚/ coda combines with tonal contrasts to signal distinct lexical items. In Korean, word-final stops are invariably unreleased ([t̚], [p̚], [k̚]) and undergo place neutralization, regardless of their underlying lenis, aspirated, or tense distinctions in onset position. This neutralization simplifies the coda inventory to three tense, unreleased variants, often accompanied by glottal overlap or pre-glottalization, especially before nasals or in careful speech. The unreleased finals contribute to prosodic structure by reinforcing syllable boundaries without audible bursts, aiding in the perception of rhythmic patterns in this . Thai exhibits unreleased final stops (/p̚/, /t̚/, /k̚/) in closed , which shorten the preceding and classify the as "dead" with short duration, influencing tone assignment within the five-tone system. Unlike initial stops, which contrast aspiration to condition high vs. low tone registers, final stops lack aspiration and serve primarily as prosodic markers, with their unreleased nature ensuring no burst disrupts tonal contours. This pattern underscores the role of coda unrelease in maintaining tonal clarity in short-voweled . Similarly, in Vietnamese, coda stops are unreleased and frequently glottalized, particularly in varieties, where the glottal closure precedes or overlaps the oral closure, enhancing perceptual cues for closure. These unreleased codas shorten vowels in closed s, affecting length perception and contributing to the six-tone register by marking "heavy" types that support falling or rising tones. Across these isolating languages of East and , unreleased stops function as a shared prosodic feature, delimiting syllables and integrating with tonal systems to convey lexical and intonational information without the need for bursts.

Formosan Languages

In , a branch of the Austronesian family spoken primarily in , voiceless stops are phonemically contrastive and typically realized as unaspirated, with unreleased variants appearing in coda positions to distinguish meanings in languages that permit codas. This pattern reflects the family's conservative retention of Proto-Austronesian distinctions, where is modulated by prosodic environment rather than serving as a primary voicing cue. In Amis, unreleased finals contribute to marking , creating heavy syllables that attract to the word's end. Releases may be audible in isolated words but are often suppressed in or potential cluster-like environments, enhancing rhythmic flow. The phonemic inventory of treats unreleased stops as allophones distinct from implosives or ejectives, which appear in specific dialects like Tsou's preglottalized voiced stops. This separation maintains clarity in minimal pairs without relying on glottal reinforcement. Within the Austronesian family, unrelease in correlates with vowel harmony systems, where backness or height features propagate across syllables, influencing stop articulation to align with harmonic constraints.

Gyalrong Languages

In of the Tibeto-Burman family, unreleased oral stops frequently follow glottal stops, forming preglottalized sequences such as [ʔk̚] that lack an audible oral burst due to the preceding glottal closure. In Japhug Gyalrong, these sequences appear in expressive forms and morphological contexts, as seen in reduplicated calls like [ʔwan ʔwan] 'calling pigs', where glottal stops interact with adjacent stops in clusters. Unrelease functions as a key phonological feature in the complex onset clusters typical of Gyalrongic languages, contributing to their rich consonant inventory of over 400 biconsonantal and triconsonantal onsets. Grammars document word-final unreleased stops like [p̚] following glottalization, contrasting with released variants in non-final positions; for instance, in the closely related Situ Gyalrong, /pʰaroʔk/ surfaces as [pʰɑˈrɔʔk̚] 'crow'. Similar patterns hold in Japhug, with coda stops such as -t in past tense markers (e.g., prɤt 'break') realized unreleased. Dialectal variations affect unrelease audibility, particularly tied to speech rate: in faster speech across Japhug varieties like Kamnyu and Tatshi, the oral release is more consistently inaudible, while slower articulation may permit partial release; Tshobdun dialects show heightened glottal involvement in such sequences.

Pirahã

Pirahã is a spoken by the , a community of approximately 400–900 individuals (as of 2024) living along the Maici River in the Brazilian Amazon. The language features one of the world's smallest phonemic inventories, with three vowels (/i, a, o/) and seven or eight consonants depending on the speaker's gender (women: /p, t, k, b, g, ʔ, s/; men: /p, t, k, b, g, ʔ, h, s/), including voiceless and voiced stops. Its syllable structure is strictly (C)V, prohibiting codas and thus preventing word-final stops, which eliminates any phonemic contrast between released and unreleased plosives; unreleased stops occur allophonically in non-final positions as the default realization. This restricted sound system aligns with the language's overall simplicity, characterized by complex prosody including tones, stress, and length distinctions rather than consonantal complexity. A of Pirahã phonology is its use of multiple communicative channels, including a register employed for intimate conversations, lulling children to sleep, or maintaining during hunts. In this register, speech is produced solely through nasalized without oral articulation for vowels or full oral bursts for stops, resulting in no audible release for plosives; the stops are approximated with glottal or pharyngeal followed by hummed , substituting the typical transient burst. For instance, monosyllabic forms like those for basic terms (e.g., equivalents to "" or "thing") are rendered as hummed sequences where any initial stop lacks release, blending into the continuous hum. This mode preserves the language's prosodic structure while minimizing acoustic salience, tying into the culture's emphasis on immediate, sensory-based communication. Daniel Everett's research, particularly his 2005 study, connects these phonological traits to broader cultural constraints in Pirahã society, such as the "immediacy of experience" principle, which limits linguistic features to those directly observable or recent, contributing to the absence of embedded clauses, numbers, and a highly restricted that prioritizes efficiency over elaboration. This cultural-phonological interplay underscores how Pirahã's sound system, including its handling of stop unrelease in alternative registers, reflects adaptations to the Amazonian environment and social practices.

Munda Languages

In the , particularly Santali, word-final stops are characteristically glottalized and unreleased, often transcribed as [t̚ʔ], distinguishing them phonemically from plain codas in native vocabulary. This feature manifests as glottalized plosives like [p'], [t'], and [k'], where the closes simultaneously with the oral articulation, preventing audible release and marking syllable-final positions in Austroasiatic roots. Such unreleased stops contrast with released variants found in loanwords from , highlighting a phonemic opposition tied to lexical origins. Kharia exhibits a parallel pattern, with pre-glottalized stops and glottal stops appearing in syllable codas, including word-final positions, where unrelease reinforces the monosyllabic root structure typical of . These glottalized forms, such as pre-glottalized obstruents, contribute to the language's phonological inventory by enhancing prosodic boundaries without altering core segmental contrasts. Across , unreleased glottal stops serve phonemic functions, such as signaling boundaries or indicating tense distinctions, as seen in contrasts between plain and glottalized obstruents in Santali and Kharia. This plays a role in the articulatory mechanics of stop closure, aligning with broader family traits. Anderson's analysis in The Munda Verb (2007) documents the prevalence of unrelease in word-final stops.

References

  1. https://home.cc.umanitoba.ca/~krussll/phonetics/narrower/unreleased-stops.html
  2. https://ecampusontario.pressbooks.pub/introducingtheipa/chapter/diacritics/
  3. https://phonetics.ucla.edu/course/chapter3/releasedstops.htm
  4. https://www.sciencedirect.com/science/article/abs/pii/S0167639311000756
  5. https://haskinslabs.org/sites/default/files/files/Reprints/HL0386.pdf
  6. https://journals.linguisticsociety.org/proceedings/index.php/BLS/article/download/3860/3559/5071
  7. https://open.maricopa.edu/essentialsoflinguistics/chapter/2-7-classifying-consonants/
  8. https://pressbooks.nvcc.edu/eng200h5p/chapter/3-4-describing-consonants-manner/
  9. https://www.mdpi.com/2226-471X/9/8/257
  10. https://escholarship.org/uc/item/2j23n6fg
  11. https://bpb-us-w2.wpmucdn.com/u.osu.edu/dist/2/123895/files/2024/04/23-hsieh-25955c1026ef86bd.pdf
  12. https://hal.science/hal-03481440v1/document
  13. https://www.internationalphoneticassociation.org/icphs-proceedings/ICPhS2023/full_papers/365.pdf
  14. https://journals.linguisticsociety.org/proceedings/index.php/BLS/article/view/3860
  15. https://journals.linguisticsociety.org/proceedings/index.php/amphonology/article/download/4904/4657/9208
  16. https://americanphonetics.ruhosting.nl/course/9-aspiration-glottalization-and-flapping/9-3-glottal-reinforcement-glottalization-and-glottal-stopping-glottaling/
  17. https://kuppl.ku.edu/sites/kuppl/files/documents/publications/Jongman_Blumstein_Lahiri_JPhon_1985.pdf
  18. https://www.internationalphoneticassociation.org/icphs-proceedings/ICPhS2023/full_papers/180.pdf
  19. https://haskinslabs.org/sites/default/files/files/Reprints/HL0487.pdf
  20. https://pubmed.ncbi.nlm.nih.gov/10450075/
  21. https://pubs.aip.org/asa/jasa/article/132/4/2700/831730/Evidence-for-language-transfer-leading-to-a
  22. https://doi.org/10.1016/j.specom.2011.05.010
  23. https://www.eleanorchodroff.com/articles/ChodroffWilson_CovariationOfVOT_JPhonPreprint2017.pdf
  24. https://www.journal-labphon.org/article/id/17746/
  25. https://scholarspace.manoa.hawaii.edu/bitstreams/b561dedd-5a33-4cd6-a425-f6b24cca6110/download
  26. https://hongkongvision.com/archives/14849
  27. https://zsigae.com/wp-content/uploads/2015/10/zsiga_externalsandhikorean_2011.pdf
  28. https://scholar.google.com/citations?user=f1NgR-YAAAAJ&hl=en
  29. https://www.isca-archive.org/interspeech_2004/tsukada04_interspeech.html
  30. https://www.researchgate.net/publication/8069656_Final_Consonants_and_Glottalization_New_Perspectives_from_Hanoi_Vietnamese
  31. http://www.lel.ed.ac.uk/~jkirby/docs/kirby2011vietnamese.pdf
  32. http://www.lel.ed.ac.uk/~jkirby/docs/brunelle2016tone-preprint.pdf
  33. https://referenceworks.brill.com/downloadpdf/display/entries/HFLO/COM-102020.pdf
  34. http://sealang.net/sala/archives/pdf8/edmondson2004phonological.pdf
  35. https://www.krisyu.org/papers/luyu24_phonetics_formosan_languages_proofs.pdf
  36. https://scholarlypublications.universiteitleiden.nl/access/item%3A2930665/view
  37. https://hal.science/hal-03190627/file/295-jacques-2021.pdf
  38. https://www.journals.uchicago.edu/doi/10.1086/431525
  39. https://roa.rutgers.edu/files/45-0195/45-0195-GREEN-0-0.PDF
  40. http://lisindia.ciil.org/santali/Santali_struct.html
  41. https://brill.com/display/book/9789004190092/Bej.9789004187207.i-474_001.pdf
  42. https://escholarship.org/uc/item/2xx3p7gk
Add your contribution
Related Hubs
Contribute something
User Avatar
No comments yet.