Sank's Glossary of Linguistics 
 Vow-Vowel G

VOWEL CATEGORY COMPACTNESS

  1. (Phonetics) We obtained the compactness of each vowel category in the F1–F2 vowel space in two steps.

    1. The F1 and F2 (Hz) of all vowel tokens were fitted to a customised MatLab script (Kartushina and Frauenfelder 2014), which calculated the area of an ellipse (Hz2) for each vowel category, participant, and register (ellipse_area).
    2. For the sake of clarity and ease of interpretability, the largest ellipse in the sample was then used as a reference point (max_ellipse_area), and the vowel category compactness was computed as a ratio score against the reference point for each participant, category and register with the formula:
      vowel category compactness = max_ellipse_area ÷ ellipse_area
     Therefore, a high vowel category compactness score indicated more compact vowel categories with respect to the reference area, whereas a low vowel category compactness score indicated looser vowel categories. | Audun Rosslund, Julien Mayor, Gabriella Óturai, et al., 2021
  2. (Examples)
     ○ There is evidence that more proficient L2 speakers may have more compact L2 and L1 categories. Kartushina and Frauenfelder (2014) examine L1 and L2 vowel category compactness and L2 French front vowel accuracy in Spanish middle school students who had studied French for four years and were judged to be intermediate level speakers of French. They find that learners with more native-like L2 French front mid rounded vowels tended to have more compact L1 Spanish vowel categories, as measured with F1 and F2 formant data. Specifically, when L1 vowel space variability is assessed with a compound measure (a sum of individual compactness scores for all five Spanish vowels), they find that learners with more compact L1 vowel spaces had more accurate L2 French /œ/ and /ø/. In addition, speakers with a more compact L1 Spanish /e/ category showed more native-like production of L2 French /e/ and /ɛ/. | Marie K. Huffman and Katharina S. Schuhmann, 2020
     ○ We argue that not only are individuals with various L1s equipped differently for the task of non-native perception, but also individuals with the same L1 vary in how their native phonological categories are represented in the perceptual space. Such variability is observable in measures of compactness of L1 phonetic categories, and its effects on non-native perception can be assessed by relating the degree of compactness to the perceived dissimilarity between novel contrasting sounds. We hypothesized that compact L1 categories give an initial advantage in distinguishing non-native contrasts. | Vita Kogan, 2020

VOWEL CATEGORY DISTINCTIVENESS

  1. (Phonetics) We measured how distinct participants' vowel categories were from each other in the F1–F2 vowel space. Vowel category distinctiveness was computed as the between-vowel category Sum of Squares (the squared distances of category cluster centroids from the overall vowel space centroid) divided by the total Sum of Squares (squared distances of individual vowel tokens from the overall vowel space centroid), for each participant and register, for eight vowel categories (we omitted the category /y/, as it fully overlaps with the Norwegian /i/ in the F1–F2 space, as the distinguishing feature is F3). Thus, vowel category distinctiveness can be thought of as a clustering performance quotient, indexing the proportion of variance in F1 and F2 explained by the vowel category identity, ranging from 0 (cluster/category membership explains no variance) to 1 (cluster/category membership explains all variance). | Audun Rosslund, Julien Mayor, Gabriella Óturai, et al., 2021
  2. (Examples)
     ○ Native speakers when learning to read and write most often are confused as to the difference between some sounds currently written as ə and the [ɪ] sound, which may indicate that the phonological categories for these vowels are not clearly distinct. | Deborah Morton, 2010

VOWEL COPY EPENTHESIS

  1. (Phonology) it has been hypothesized that the phenomena of epenthetic vowel copy (i.e., when the epenthetic vowel shares quality with neighboring vowels) is due to a transfer of phonological features from neighboring vowels and/or consonants towards an undeterminate epenthetic vowel (Rose and Demuth 2006, Uffmann 2006). | Adriana Guevara-Rukoz, 2018
  2. (Examples)
     ○ The language isolate Huave as spoken in San Francisco del Mar, Oaxaca State, Mexico, has a process of vowel-copy epenthesis across an intervening consonant (V1CV2), where V2 is epenthetic. The examples in (1) show vowel epenthesis with the 3PL suffix (Kim 2008):

    1. a.
      uC-u
      a-ʃum-uhw
      TV-find-3PL
      'they find (it)'
      [aʃumuh]
      b.
      uCj-i
      a-mbuʎ-ihw
      TV-burn-3PL
      'they burn (it)'
      [ambuʎjʊɸ]

     | Yuni Kim, 2024
     ○ We focus on perceptual vowel epenthesis following /h/. This case is ideally suited for our objective as in Japanese loanwords these fricatives are typically adapted by adding a "copy" of the preceding vowel when they occur in a syllable coda. For instance, Bach, (van) Gogh, and Ich-Roman are adapted as /bahːa/, /gohːo/, and /ihːiroman/. In work on loanword adaptations, cases of vowel copy in epenthesis have been explained as a result of the spreading of phonological features from the preceding vowel onto the epenthetic vowel (i.e., vowel harmony), for instance in Shona (Niger-Congo; Zimbabwe, Zambia), Sranan (English-based creole; Suriname), and Samoan (Austronesian; Samoa) (Uffmann 2006), and Sesotho (Niger-Congo; Lesotho) (Rose and Demuth 2006). | Adriana Guevara-Rukoz, 2018
     ○ In some cases, epenthetic segments are copies of nearby elements:

      Selayarese (Malayo-Polynesian; South Sulawesi) (Mithun and Basri 1985)
      Copy Epenthesis
    1. a. /sahal/ → [sahala] 'profit'
      b. /potol/ → [potolo] 'pencil'
      c. /lamber/ → [lambere] 'long'

     | Catherine Kitto and Paul da Lacy, 1999

VOWEL DEVOICING
(Examples)
 ○ Vowel Devoicing in Tokyo Japanese (Fujimoto 2015)

ʃu̥taiseː
ʃi̥sen
ɸu̥soku
tʃi̥kai
katsu̥toki
aʃi̥ka
'individuality'
'eye gaze'
'shortage'
'pledge'
'win time'
'sea lion'
 
 
 
 
 
 
ʃudaika
ʃizen
ɸuzoku
tʃigai
katsudoː
saʃiga
'theme song'
'nature'
'affiliated'
'difference'
'life activities'
'inserted picture'
 Languages with more consistent devoicing have larger amplitude laryngeal gestures. | Jason A. Shaw, 2025
 ○ Devoicing of vowels is a striking feature of contemporary French, but it has received relatively little attention in the linguistic literature. Perhaps the most frequently-cited example of this phenomenon is the common pronunciation of oui with a voiced [w] and a voiceless [i̥] (Laver 1994). Fónagy (1989) describes the voiceless [i̥] at the end of merci as sounding like un sifflement aigu 'a sharp whistling'. He gives examples from television as well as from controlled reading of texts, and suggests that the devoicing serves as a way of marking the end of an utterance. | Caroline L. Smith, 2003
 ○ In Turkish a syllable containing any of the four high vowels [i y ı u] can be realized without any audible traces of voicing. The phenomenon is demonstrated in spectrograms and waveforms that show a contrast between two words produced by the same speaker, one containing a fully realized vowel, the other containing a fully devoiced vowel. As becomes clear from the spectrograms, the endpoint of a continuum of vowel devoicing can be interpreted as vowel deletion. On the left we see a spectrogram and waveform of the word tufek 'gun, rifle' spoken in a slow rate of speech. Vertical striations at the bottom of the spectrogram are the individual glottal pulses, showing that this first vowel is voiced. The presence of the vowel is also reflected in the waveform. The spectrogram on the right shows the same word produced in a normal rate of speech. Here, the vowel has completely disappeared; there are no voicing traces left so that this vowel is analyzed as completely devoiced. This phenomenon is previously undocumented for Turkish, but resembles a process noted for several other languages, including Svabian (Griffen 1983), Canadian French (Cedergren and Simeneau 1985, Cedergren 1986), Korean (Jun and Beckman 1993, 1994) and Japanese (Mccawley 1968, Jun and Beckman 1993, 1994). | Stefanie Jannedy, 1995
 ○ We offer this paper as a revision of earlier attempts to account for vowel devoicing based on new data which has come to light as a result of ongoing field work among the Northern Cheyenne (Algic; USA). In this paper we will start from the analysis of Frantz (1972).
 Frantz identified three basically different environments in which vowels devoice:

  1. Phrase-Final Devoicing
    a. mahpe̥ 'water'
    b. mahpe-ná-vóóhtḁ 'I see the water.'
    c. ná-vóóhta mahpe̥ 'I see the water.'
  2. Prepenultimate Devoicing
    a. mḁheóʔo̥ 'house'
    b. hóhke̥haʔe̥ 'hat'
    c. ko̥sáne̥ 'sheep' (PL)
    d. mḁxého̥ 'logs'
    e. mḁhpevḁ 'in the water'
    f. kaʔe̥škóne̥ 'child'
  3. Penultimate Devoicing
    a. hóhko̥xe̥ 'axe'
    b. hemáhkḁse̥ 'buffalo chip'
    c. have̥se̥ 'evil'
    d. móne̥še̥ 'Really!'
    e. véve̥tse̥ 'horn'
    f. voxo̥tse̥ 'meat'
    g. né-vóomḁtse̥ 'I see you.'
    h. ná-némenémḁse̥ 'I was said to be singing.'
 | Wayne Leman and Richard Rhodes, 1978

VOWEL DISHARMONY
(Examples)
 ○ In Finnish, front vowels (ä, ö, and y [/æ/, /ø/, /y/]) cannot appear in the same lexeme as back vowels (a, o, and u [/a/, /o/, /u/]). Finnish also has two "neutral" vowels (e, i [/e/, /i/]) that can accompany either back or front vowels. For instance, the word pouta /pouta/ 'dry weather' is composed of three back vowels and the word pöytä /pøytæ/ 'table' is composed of three front vowels—items like poyta or pöutä that combine front and back vowels (i.e., vowel disharmony) could not possibly be native words in Finnish. | Manuel Perea, Jukka Hyönä, and Ana Marcet, 2022
 ○ The diversity of disharmonic vowel sequences in Seto Estonian:

  1. Stem-Internal
     [ ä a ]
     [ e a ]
     [ i a ]
     
    impossible
    marked
    unmarked
     
     
     
     
     
    Stem + Ending
     [ ä ] a
     [ e ] a
     [ i ] a
     
    impossible
    impossible
    impossible
 | Paul Kiparsky and Karl Pajusalu, 2001
 ○ Apart from fossilized lexical items which may fail to conform to vowel harmony, disharmony may arise as a result of productive morphological processes that remain active in a given language. We present a preliminary typology here. Disharmony may arise under at least the following scenarios:

 | K. David Harrison, 1999
 ○ Suomi et al. (1997) found that listeners use vowel disharmony as a cue for speech segmentation. Thus, HYmy was easier to detect in PUhymy than in PYhymy. | Jean Vroomen, Jyrki Tuomainen, and Beatrice de Gelder, 1998

VOWEL FRICATIVIZATION

  1. (Phonetics) A brief discussion of the acoustics and articulation of fricativized vowels, which occur at two places of articulation, is provided here to familiarize the reader with this uncommon segment type.
     The first and most common subset of these sounds, the so-called apical vowels, are articulated coronally. They can broadly be characterized in articulation as having simultaneous tongue tip and tongue body constriction, have been a subject of concerted study for many years, and have been acknowledged as a distinct and unusual class of phonetic segment for some time (Gjerdman 1916, Karlgren 1926).
     The second broad type of fricativized vowel, somewhat less common than the coronal type, is produced with labiodental constriction. The labiodental fricativized vowels have not been the subject of concerted study in the same way as the coronal type but are commonly mentioned in discussions of syllabic consonants (Bell 1978).
     Fricativized vowels are in some ways quantifiably indeterminate between high vowels and voiced fricatives, having acoustic and articulatory characteristics of both. Acoustically, they generally resemble syllabic voiced fricatives, but with a more clearly audible formant structure resembling that of "high vowels rather than a schwa-like quality often associated with syllabic fricatives" (Connell 2007). This formant structure is not of a peripheral (very front or back) high vowel, however, and is more frequently reflective of a centralized quality approaching [ɨ] (Connell 2007, Cheung 2004, Engstrand et al. 1998/2000). For coronal fricative vowels, F1 is typically higher than might be expected and F2 lower than might be expected given the vowel's frontness and height (Engstrand et al. 1998/2000, Feng 2007). Despite the presence of fricative noise at or above F3 or F4, F3 is also salient enough to be useful in cuing distinctions between different types of coronal fricativized vowels, as in Standard Mandarin's contrast between apico-alveolar [z̩] and retroflex [z̢̩  ] (Cheung 2004).
     The frication associated with fricativized vowels is invariably strident (Connell 2007, Faytak 2013), with the directing of a jet of air at an obstacle. | Matthew Faytak, 2014
  2. (Examples)
     ○ Using frication as a yardstick, Faytak (2015) posits that there are two types of "fricativized" vowels:

    1. Those with a steady-state fricative noise of relatively low intensity.
    2. Those with high-intensity frication toward the beginning of the vowel and lower-intensity frication towards the end.

     | Sean Foley, 2022
     ○ Typologically, vowel fricativization, resulting in syllabic fricatives and sibilants, is found mainly in high vowels. | Alan C. L. Yu, 1999

 

Page Truncated April 11, 2026

 
 B a c k   T o   I n d e x