An acoustic study of nasal consonants in three Central Australian languages

Abstract

This study presents nasal consonant data from 21 speakers of three Central Australian languages: Arrernte, Pitjantjatjara and Warlpiri. The six nasals considered are bilabial /m/, dental /n/, alveolar /n/, retroflex /ɳ/, alveo-palatal /ɲ/, and velar /ŋ/. Nasal formant and bandwidth values are examined, as are the locations of spectral minima. Several differences are found between the bilabial /m/ and the velar /ŋ/, and also the palatal /ɲ/. The remaining coronal nasals /n n ɳ/ are not well differentiated within the nasal murmur, but their average bandwidths are lower than for the other nasal consonants. Broader spectral shape measures (Centre of Gravity and Standard Deviation) are also considered, and comparisons are made with data for stops and laterals in these languages based on the same spectral measures. It is suggested that nasals are not as easily differentiated using the various measures examined here as are stops and laterals. It is also suggested that existing models of nasal consonants do not fully account for the observed differences between the various nasal places of articulation; and that oral formants, in addition to anti-formants, contribute substantially to the output spectrum of nasal consonants.

Young infants’ discrimination of subtle phonetic contrasts

“…This is problematic for several reasons. First, recent acoustic investigations of Australian languages 1 point to the possibility that the F2-F3 space might not be ideal to compare the distinctiveness of subtle place differences for either nasals (Tabain, Butcher, Breen & Beare, 2016a) or laterals (Tabain, Butcher, Breen & Beare, 2016b). Instead, Tabain and colleagues report greater success with measures that index the shape of the acoustic spectrum.…”

Acoustic characteristics of Punjabi retroflex and dental stops

“…The fourth spectral moment (spectral kurtosis) is a measure of "peakiness" in the spectral envelope of the release burst, and differentiates the tongue posture of apical and laminal articulations (Li et al, 2009;Tabain and Butcher, 2015). Within the class of coronals, laminals (dentals and palatals) show higher spectral CoG, spectral SD, and negative spectral skewness and spectral kurtosis, compared to apicals (retroflexes and alveolars) (Gordon et al, 2002;Tabain, 2012;Tabain et al, 2016). In retroflexes, raising the tongue tip towards a hard palate increases the length of the front cavity.…”

“…Unlike some other coronals, there is typically no constriction separating postlabial and sub-lingual cavities in retroflex production; rather, the whole space anterior to the coronal constriction forms a single large front cavity. Increasing the length of the front cavity lowers the spectral CoG, spectral SD, but increases the spectral skewness and spectral kurtosis (Laaksonen et al, 2011;Tabain et al, 2014). Spectral moments of alveolars more closely match those of retroflexes than dentals in some coronal-rich languages (Tabain, 2012;Tabain et al, 2016), but spectral CoG does not reliably differentiate retroflexes and dentals in word-medial (/aCa/) and word-initial (/a#Ca/) contexts in Wubuy, a language with a similar set of coronal contrasts (Bundgaard-Nielsen et al, 2016).…”

Acoustic Realization and Inventory Size: Kannada and Malayalam Alveolar/Retroflex Laterals and /ɻ/

“…As part of a larger cross-linguistic study of spectral cues to place of articulation across different consonant manner classes (for stop burst spectra, see Tabain et al, 2016a, and for nasal murmur, see Tabain et al, 2016b), Tabain et al (2016a) examined the lateral "ring" using spectral moment measures calculated in the 1-to 5-kHz range, and also using formant measures. Consistent with results for both stop bursts and nasals, the spectral centre of gravity (CoG -or first spectral moment) was found to be lower for retroflexes than for alveolars (although less relevant for the present study, the CoG was highest for the palatal consonant).…”