VOT

In 1964, Lisker and Abramson:

“the timing of glottal pulsing relative to supraglottal articulation would account for the great majority of homorganic consonantal distinctions traditionally said to depend on voicing, aspiration,”tensity“, and the like” (Abramson and Whalen 2017).

Abramson & Whalen (2017) argue that VOT is a cross-linguistically useful (and simple) measure, even if not a universal discriminator.

Considering a sample of languages, (Ladefoged & Cho 1999) show that VOT measurements do not directly depend on sound types and oppositions. Below is the sub-sample that has ejectives:

Ladefoged and Cho graph with data Abaza speakers

In our study, we want to consider how Abaza consonants perform in terms of temporal structuring (VOT etc.), comparing voiceless and ejective stops.

Data. Featuers of stops

• POA: labial, dental, velar, uvular

• Manner: stops

• Phonation type: voiceless, ejective

Data. Measurments

• Previous vowel duration (v1)

• Closure duration (cd)

• Voice Onset time (vot)

• Following vowel duration (v2)

Measurments

Examples of annotation in Praat

Data. List of words

• Words with two syllables (61 words)

• Phonectic context aCa

• Stressed position not controlled

• Eleciting words by a Russian translation

• Three repetition followed by a carrier phrase

• Four female speakers

List of words

• List of words was generated automalically from dictionary with R script

Analysis

• Data were annotated in Praat

• Extracted with Praat script

• Analysed and visualized with R by G. Moroz

Absolute values of CD

Absolute values of VOT

Absolute values of CD~VOT

Relative values: CD / CD + VOT

Relative values: VOT / CD + VOT

Generalized linear mixed model

Generalized linear mixed model

## Generalized linear mixed model fit by maximum likelihood (Laplace
##   Approximation) [glmerMod]
##  Family: binomial  ( logit )
## Formula: phonation ~ cd + vot + poa + (1 | speaker)
##    Data: df_for_glm
## Control: 
## glmerControl(optimizer = "bobyqa", optCtrl = list(maxfun = 1e+06))
## 
##      AIC      BIC   logLik deviance df.resid 
##    536.3    567.8   -261.2    522.3      655 
## 
## Scaled residuals: 
##      Min       1Q   Median       3Q      Max 
## -20.5505  -0.2826   0.2621   0.4298   4.3817 
## 
## Random effects:
##  Groups  Name        Variance Std.Dev.
##  speaker (Intercept) 0.7428   0.8619  
## Number of obs: 662, groups:  speaker, 4
## 
## Fixed effects:
##              Estimate Std. Error z value Pr(>|z|)    
## (Intercept)    0.9655     0.4765   2.026  0.04275 *  
## cd             0.7363     0.1607   4.580 4.64e-06 ***
## vot            1.5789     0.1906   8.283  < 2e-16 ***
## poadental      1.5357     0.2677   5.736 9.67e-09 ***
## poavelar_pal  -2.5637     0.4105  -6.245 4.24e-10 ***
## poauvular     -1.1142     0.3803  -2.930  0.00339 ** 
## ---
## Signif. codes:  0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
## 
## Correlation of Fixed Effects:
##             (Intr) cd     vot    podntl pvlr_p
## cd          -0.030                            
## vot          0.140  0.083                     
## poadental   -0.274  0.081  0.092              
## poavelar_pl -0.250  0.250 -0.365  0.316       
## poauvular   -0.265  0.133 -0.420  0.331  0.429

Thank you!

Tatiana Mamonova (t_mamonovaa@outlook.com)