OPTIMALNOSNA TEORIJA U FONOLOGIJI (2. dio)

This paper serves as an introduction to Optimality Theory and to the principles of its application to phonology. Optimality Theory (OT) is a formal theory of language, based on the interaction of universal, violable and ranked constraints. The core idea of Optimality Theory is that the surface language forms are the product of the resolution of the conflicting demands that are forced upon them by two types of constraints: faithfulness constraints that command the output form to be as similar to the input form as possible and markedness constraints which demand that the output forms be in agreement with the universal principles of simplicity and unmarkedness. The language form that best meets these conflicting requirements is the optimal candidate. The paper introduces four areas of significance. First, Optimality Theory is related to the earlier generative models of phonology, especially pointing out certain prominent phonological problems, such as the Duplication Problem and conspiracies, which have brought the traditional derivational honology into question and which consequently led to the founding of constraint-based phonological models such as OT. In the second part, the architecture of the OT grammars is explained and certain central ideas like optimality, constraints, evaluations and markedness are defined. Third, the main principles of Optimality Theory are applied on a wide variety of segmental processes such as assimilations, deletions, epenthesis etc. In the fourth part, suprasegmental processes like rhythm-induced vowel syncope and various tone alternations are described within the OT framework. Since Optimality Theory has been the dominant descriptive model in international phonology for the last two decades, the aim of this paper is to bring it one step closer to contemporary Croatian linguistics.