The primary purpose of the paper is to show that such phenomena as the cliticization of weak pronouns, TH/EX in there constructions and the wh-movement from extracted nominals can fall under one generalization called "defocusing", a pragmatic operation that cannot be applied twice to the same syntactic object. We argue, in this paper, TH/EX is not a PF operation but a pragmatic operation before Spell-Out. In our model, syntactic objects should go through all the syntactic and/or pragmatic operations before they reach the interface and lexical insertion is a 24hr outlet rather than a one-fell-swoop operation.

Chegyong Im. 2003. Processing of Ambiguous Sentences: An Optimality Theoretic Approach. Studies in Modern Grammar 31, 71-88. Our approach claims that the same principles that govern syntactic structures also contribute to a theory of sentence processing. We also argue, following Gibson & Boihier (1998), that the ambiguities can be resolved by the strategy of Optimality Theory i.e., by picking out the most optimal candidate with the rank of the constraints converted from the principles of Minimalist Program. The constrains suggested here are Thematic Valency, Node Conservativity and Node Locality to explain the preference among the ambiguous interpretations as well as Deictic Condition to select the antecedent for the Pros and anaphors.

Chegyong Im. 2001. Person Constraint and Licensing of Expletive Constructions. Studies in Modern Grammar 25, 29-53. Some structures allow number agreement but not first/second person agreement. That would follow if the [person] feature of T reduces to [3 person] (the default choice). We will show, in this paper, the environment in which this phenomenon happens, suggesting a universal constraint for the locative inversion constructions and expletive constructions. The other major purpose of the paper is to illustrate the licensing of expletive constructions in the framework of Rizzi (1997) and Platzack (2000). We argue that the different syntactic behavior between there type expletives and it type expletives can be explained in terms of Multiple Interfaces. We will also try to show that the structure of TECs in Icelandic can be neatly identified in our proposal.

Che-Gyong Im. 2000. Licensing of "-ka/-i" at Morphophonemic Level. Studies in Modern Grammar 20, 47-69. There are a lot of controversies on how to explain the behaviors of "-ka/-i" in the framework of minimalist program suggested in Chomsky (I993, 1995a, 1995b). Recent studies on the Nominative Case in Korean, assuming Case feature checking, successfully explains the typical Korean sentences bearing structural Nominative Case as long as they classify Korean as a configurational language. These approaches, however, face some problems when they treat "-ka/-i" in Multiple Nominative Constructions(MNCs), Nominative Object Constructions(NOCs) only as Nominative marker. In this paper, we will suggest that the odd behaviors of "-ka/-i" can be explained if we assume two different levels at which this marker is licensed: syntactic level(before Spell-out) and morphophonemic(PF) level. Converting the hedges of Case theory and the functions of "-ka/-i" into the constraints of Optimality Theory, we attempt to solve the problems raised in the licensing of "-ka/-i".

Im Che-Gyong. 1999. Interpretation of "-self": An Optimality Theoretic Approach. Studies in Modern Grammar 16, 139-159. The distribution and the interpretation of the English reflexive "-self" have been controversial among scholars since it was first suggested in Chomsky (1980, 1982). In this paper, we try to show that the English reflexives can be divided into two groups: core vs. non-core reflexives, adopting Webelhuth`s (1995) suggestion. Core reflexives are the ones which obey the Binding Principles at morpho-syntactic level and non-core reflexives are the ones which are used logophorically in a discourse. We argue that as long as the interpretation of a reflexive is finding its antecedent, a non-core reflexive has its prominent antecedent in a dicourse. The hedges to maintain the Principles of Binding can be converted into the constraints of Optimality Theory, the hierarchy of which plays the decisive role to select the optimal candidate for the antecedent of a reflexive. To select the optimal antecedent of non-core reflexives, we suggest a pragmatic constraint which can save the logophoric use of the reflexives. We also suggest that if those constraints are ranked properly, not only the interpretation of core and non-core reflexives can be obtained in a systematic way but their complementary distribution can be clarified as well.