Input 'Clusters' and Contrast Preservation in OT

The starting point of this paper is the Contrast Preservation Theory (PCT) of Lubowicz (2002), and its novel approach to phonological opacity in OT. A core claim of PCT is that various opaque mappings, which have been problematic in OT, can be readily explained as the effect of systemic contrast preservation. Thus, PCT is an OT model which evaluates not only the Markedness and Faitfhfulness of its candidates, but also their preservation of phonological contrasts. To do so, the EVAL component of PC theory assesses multiple input forms, in an input scenario.The goal of this paper is to provide an alternative to scenarios, which still captures the contrast-preserving patterns suggested by PCT. My alternative is a grammar-based algorithm that builds finite, language-specific sets of input forms called input clusters. In building clusters, the algorithm relies crucially on the existing core of OT: the language-specific ranking of Markedness and Faithfulness constraints, and the decision-making powers of EVAL. Working loosely within the framework of PCT, I use the algorithm and its resulting clusters to analyze a derived environment effect (one opaque pattern explained under PCT). The success of this analysis provides initial support for the algorithm, and for the broader claim that such an algorithm�s clusters will contain all the input forms necessary to capture contrast-preserving opacity.After the introduction, section 2 provides some minimal theoretical background on PCT and its notion of input scenarios, and in section 3 I propose my cluster-building algorithm. Section 4 introduces the derived environment effect (DEE), and uses data from a Campidanian Sardinian DEE to demonstrate how the algorithm builds clusters. Section 5 puts those clusters to work in the analysis of Campidanian Sardinian, using a contrast-preserving constraint based on those of Lubowicz (2002). The last section summarizes the results, and raises questions for future work