This article explores the nature of locality principle. In particular, we compare two recent conceptions of locality, which are Chomsky`s (2001a, 2001b) phase-based implementation of locality and Hiraiwa`s (2003) version of cyclic derivational locality. We present a new piece of empirical arguments against Hiraiwa (2003) from locality circumvention effects observed in passivization of applicative structures. We show that the locality circumvention phenomenon induced by wh-movement of a higher internal argument when a lower internal argument undergoes passivizing A-movement past it in low applicative structures cannot be explained under Hiraiwa`s cyclic locality computation, whereas it can be well accounted for under Chomsky`s phase-based evaluation view, which is a partly representational version of locality principle.

Lee Jae-Choel. 1999. The Locality Condition of Movement. Studies in Modern Grammar 16, 103-119. The main purpose of this paper is to point out the problems of the shortest move requirement and the Attract-F proposed in Chomsky (1993) and Chomsky (1995) respectively, and to propose an alternative. The existing locality conditions restrict the domain of either the moving elements or the feature checkers. But both approaches cannot explain Islands Violations, Superraing and Crossing Effects. To solve these problems, I argued that the existing Attract-F requirement be maintained as the locality condition of feature checkers, but the Minimal Link Condition be required as independent locality condition of moving elements. In contrast with Chomsky(1995), I assume that [+WH] feature of Wh-phrases is [-interpretable] in order to explain the asymmetries of Islands Violations and Crossing Effects.