Vowel Harmony in Azeri Turkish of Moghan Region: An Auto-Segmental Analysis

The aim of this research is to study the vowel harmony in Azeri Turkish, spoken in Moghan region. It focuses on the role of opaque units in the process. Based on the generative phonology's linear approach, the vowel harmony is assumed to be the product of sequential assimilations of vowels in the string, whereas auto-segmental phonology considers harmonic features as being at a separate layer which is acting autonomously. For that, these features spread simultaneously on the whole vowels in the string, unless an opaque unit intervenes and blocks it. The analysis of the data collected from the region under investigation (namely from Germi city) showed that the vowel harmony can be described, according to the principles offered by Katamba (1989), as follows: a) vowel harmony could be observed in the features [±back] and [±round], b) [±back] harmony applies on the whole vowels in the system, while [±round] applies only on [+high] vowels, c) opaque units which block harmony feature's spreading are: 1) suffixes with a [+low] vowel in their underlying form and 2) the consonant /j/, occurring in the places other than root morpheme, which imposes [-back], [-round] and [+high] features on the preceding vowel, and this in turn influences the following vowels.