Abstract
Vowel reduction occurs in a rapid speech in unstressed syllables or it happens when an affix is added to a stressed syllable and it attracts the stress. Vowel reduction was described through different approaches likeSonority-Based Reduction, Contrast Enhancement Reduction, Centripetal, Centrifugal and [AIU] model. In this article, we describe Persianâs vowel reduction by Andersonâs [AIU] model in which vowel phonetic information is encoded by phonemic categories and there is a direct relationship between vowels spectral profiles and phonemic categories [A,I,U]. Therefore, vowel reduction occurs when one of these categories delete (Deletion), the categories are reversed (Reversion) or even one category is replaced by another (Augmentation). All of these processes are applicable in Persian.
)