vowel harmony

Two of the models in the Serial Optimality Theory in which derivational paths compete and are evaluated by Prec(a, b) family of constraints are Optimality Theory- Candidate Chains (OT-CC) and Optimal Interleaving (OI). The former operates exclusively at the level of phonology, while the latter postulates the interplay between morphology and phonology. This article presents a case study of vowel harmony in Kalhori Kurdish that necessitates a level where phonological and morphological processes alternate to account for both the application and lack of application of a phonological process. In Kalhori Kurdish, the features of [o] are spread regressively all the way to the leftmost edge of the verb; however, this is not the case when the coordinating conjunction clitic [=o] is added to the verb. The analysis of the data collected from 15 Kalhori Kurdish speakers revealed that OT-CC cannot simultaneously explain the root-prefix vowel harmony and the absence of this process between the clitic and the verbal ending, unless the morphological information is duplicated at this level. However, benefiting from the possibility of morphological processes alternating with the phonological ones, the OI model successfully explains both the presence and the absence of the vowel harmony

Jam (2020) attempts to provide an Optimality-Theoretic analysis of Persian vowel harmony. As to cases where backness harmony does not appear to occur, his explanation appeals to Lexically Specific Constraint Theory, treating these instances as exceptional rather than systematic. In this reply, I argue that such an appeal to exceptionality is unnecessary, and I propose an alternative analysis in which regressive vowel harmony in Persian can be accounted for in a principled, unified manner without recourse to lexically indexed constraints. By introducing an appropriately formulated markedness constraint, it becomes possible to capture the full range of observed patterns while maintaining the parsimony and explanatory elegance expected of an Optimality-Theoretic framework. In addition to this central point, I identify several further weaknesses in Jam’s analysis. These include a conflation of rule ordering with constraint ranking, which obscures the theoretical distinction between derivational and constraint-based approaches, the problematic assumption that the phonological representation of loanwords should mirror their source-language forms, and the omission of critical information in both the representations and tableaux.