음성형태 병합분석과 다중문자화

PF Merger Analyses and Multiple Spell-Outs

The goal of this article is to investigate the notion of Multiple Spell-Out, most advocated in recent minimalist studies such as Epstein (1999), Uriagereka (1999), Chomsky (2000, 2001, 2005) and Bošković (2001), etc., and then to clarify how to apply this notion to some syntactic derivations, especially placing emphasis on the PF merger procedures. To begin with, we attested the adequacy of the Multiple Spell-Out model in syntactic derivations, with a comprehensive critique of the Single Spell-Out model included. Next, we also investigated how the input data for PF merger can be analyzed under the Multiple Spell-Out model in the aspects of both verbal morphology and syntactic derivations. Unlike Chomsky’s (1999, 2000) assumption, this study proposed that all the maximal nodes be candidates for Spell-Outs and that those Spell-Outs be activated only when each specifier position is filled with the moved syntactic object(s). This study proposed the Multiple Spell-Out Condition as an alternative approach to Ausin(1999)’s Matching Principle. In conclusion, this study showed us that this condition could be an effective and proper solution to the problems caused in the single Spell-Out model.