불투명성과 후보연쇄최적성이론

Abstract

ABSTRACT

Opacity and Optimality Theory with Candidate Chains

The purpose of this study is to analyze individual language's opacity from a different standpoint, which is based on Optimality Theory with Candidate Chains (OT-CC, McCarthy 2006a-d, 2007), a novel version of classic Optimality Theory (OT, Prince & Smolensky 1993, 2004). In terms of Chomsky & Halle (1968)'s rule-based theory (SPE), opacity can be found in two cases: overapplication and underapplication. The former refers to the case where a process applies even though its context is not present at the surface. Contrary to the former, the latter refers to the case where a process does not apply even though its context is present at the surface. Opacity has been a challenge to OT since it is an output-oriented theory. To handle opacity, many other theories such as Local Conjunction (Smolensky 1993, 1995; Kirchner 1996; Kager 1999; ？ubowicz 2005), OO-correspondence (Benua 1995; Kager 1995; McCarthy 1996), Lexical Constraint Domains (It？ & Mester 1995a-b), Two-level Well-formedness (Orgun 1995; Archangeli & Suzuki 1997; Cho 2001), Multi-stratal Evaluation (It？ & Mester 1999), Lexical Accent (Hammond 1999), and Sympathy Theory (McCarthy 1999, 2002), have been proposed. According to Kager (1999), McCarthy (2007), Seo & Jo (2006, 2008a-d), and Jo & Seo (2009a-b), however, there are also many problems in them. In this study to resolve the problems in SPE and constraint-based theories, I try to resolve opacity by employing the new account of OT-CC. Based on OT-CC, this study investigates and analyzes opacity, supporting the superiority of OT-CC. According to McCarthy (2007: 3-4), OT-CC is a synthesis of OT with derivations: OT-CC includes not only a surface form but also a series of intermediate forms, each of which is minimally different from the form that immediately precedes it. And Like OT, OT-CC is a constraint-based theory, in which ranked constraints evaluate a set of candidate outputs. The main difference from OT is in the nature of the candidates. Candidates in OT are individual forms. OT-CC, however, records the history of faithfulness violation in forming a valid candidate chain. The winning candidate chain is evaluated according to three well-formedness conditions: initial form, gradualness, and local optimality (harmonic improvement+best violation). Here candidate chain employs PREC(edence) constraints(A, B)(PREC(A, B)) in order to record the history of faithfulness violation. PREC(A, B) requires that every B-violating Localized Unfaithful Mapping (LUM) proceed and be not followed by A-violating LUM in the reduced LUM Sequence (rLUMSeq). If the order of faithfulness constraint violation is reversed or violation of constraint A is skipped, the candidate chain under question gets violation marks. This study is organized as follows: Chapter 1 provides an introductory remark about this study. In Chapter 2, I outline the concept of transparency and opacity while exemplifying individual language's data. In Chapter 3, I introduce OT and OT-CC. And after comparing OT-CC with SPE and OT, I consider the common properties and variances of the theories. In Chapter 4, I analyze individual language's opacity in the frameworks of SPE, OT, Local Conjunction, OO-correspondence, Lexical Constraint Domains, Two-level Well-formedness, Multi-stratal Evaluation, Lexical Accent, and Sympathy Theory. And then I point out the problems of the above theories. Chapter 5 is the applying part of OT-CC. In this chapter, I try to resolve individual language's opacity by utilizing OT-CC. By doing so, I prove that OT-CC can solve opacity more persuasively than any other theories. In Chapter 6, after examining English noun stress which shows opacity in the word-final tense vowels and analyzing the opacity in the frameworks of SPE and OT, I point out the problems of the theories. To resolve the problems, I apply the opacity to OT-CC. By doing so, I show OT-CC's firm explanation. Chapter 7 summarizes this study.