Vowel dispersion and Kazakh labial harmony

Abstract

This paper uses novel data showing gradient labial harmony in Kazakh to compare Kaun's (1995) feature-based analysis with a dispersion-based analysis in a Maximum Entropy Harmonic Grammar. The paper demonstrates that the dispersion-based analysis better fits the Kazakh data than Kaun's analysis, and then extends it to account for four languages with harmony patterns different from that in Kazakh. The paper also argues that the dispersion-based account provides a better analysis of the typology of labial harmony than Kaun's feature-based analysis.

Mayak and the Typology of Labial Harmony

“…I claim herein that there is no one cross-linguistic threshold for similarity, but that languages with labial harmony implement different thresholds that all function in the same way-they promote harmony from perceptually similar contrasts, and penalize surface alternations that involve larger perceptual shifts. As noted above, the dispersionary analysis developed herein and in McCollum (2016) predicts that sequences of identical vowels should be strongly preferred cross-linguistically because the more similar a vowel is to its pair the more likely it will both initiate harmony and surface via harmony. This prediction is further explored in the following section.…”

“…I interpret this as strong confirmation that labial harmony preferentially, and perhaps fundamentally creates syntagmatic vowel identity, which is not explicitly predicted by Kaun's analysis. 6.2 High vowels One generalization that emerges in McCollum (2016), which is also evident in the Maa data used for the above analysis is that high vowels tend to be more dispersed from one another than non-high vowels. This relates to the claims made by Kaun (1995), namely that high round vowels are more perceptually salient than non-high round vowels.…”

“…Moreover, both progressive and regressive ATR harmony involve [+ATR] spreading, so the [-ATR] vowel, [ɔ], is not a possible trigger for ATR harmony in (6g,h). It is argued throughout the literature that labial harmony is parasitic on some other form of harmony or vowel agreement (Mester 1988;Cole & Trigo 1988;Jurgec 2011;van der Hulst & Moskal 2013;McCollum 2016). If this is the case, then labial harmony depends upon ATR harmony.…”

“…This warping reflects the more general perceptual salience of F1 over F2 contrasts (see Schwartz et al 1997;Padgett & Tabain 2005). As in McCollum (2016), I chose to compress the F2:F1 vowel space to 0.625:1, the middle value in the proposed range. To do this, the target ratio, .625:1, is divided by the range of attested F2:F1 contrasts to give the warping factor, λ.…”

“…Turkic languages exhibit a drastic durational difference between high and nonhigh vowels, where high vowels are often subject to complete elision while non-high vowels are more resistant to this reduction process (see Sjoberg 1963;Poppe 1964Poppe , 1968Comrie 1997;Levi 2005;McCollum 2015McCollum , 2016Washington 2016 for further evidence). McCollum (2016) demonstrates that in Kazakh non-high vowels are over twice as long as [+hi] vowels. Therefore, we need only expand the model to include duration as a cue to trigger harmony as well as a perceptual threshold by which to delimit undergoers, as well.…”