QUANTIFIED OPTIMALITY AND THE PHONOLOGICAL PARSING OF KOREAN SOV SENTENCES
Abstract. In an attempt to understand the variable nature of phonological phrasing in Korean, this study analyzes intuitional judgments of 53 native speakers of Korean who evaluated possible phonological phrasings of simple Subject-Object-Verb sentences: [S]-[OV], [SO]-[V], [SOV],
and [S]-[O]-[V]. Analysis of the quantified rating data reveals a strong preference for a subject-predicate phrasing ([S]-[OV]) and a distinct dispreference for the phrasing in which the subject and object were grouped into a single phonological constituent ([SO]-[V]). These preferences are
then analyzed in the context of a constraint-based theoretical framework; by extending the Optimality Theory (OT) notion of "ranking" to include not only constraints but also candidates, we corroborate the existence of preference patterns in native-speaker intuitions regarding the phrasing
of [SOV]. These patterns are explained by referencing three putatively universal constraints that govern the phonological phrase formation: one that aligns phonological phrases with syntactic phrases, a second that requires phonological phrases to be binary branching, and a third that limits
the weight of phonological phrases to five syllables. Although the quantitative data and the proposed OT account are not in complete agreement, the account put forward should encourage further research into a more comprehensive integration of variation studies and OT.