A study of manner assimilation: Focusing on word-initial [l]-avoidance, lateralization, and nasalization in Korean

Abstract

The purpose of this paper is to provide an explanation of manner assimilation as a whole in terms of the interaction of well-defined OT constraints. It turns out that previous OT treatments of lateralization in Korean assumes the crucial role of a language-specific constraint to the effect that a sequence of ln or nl is prohibited in the output. I attempt to get rid of the language-specificity by offering to replace the constraint in question with a licensing one requiring a doubly-linked structure of [lateral], LICENSE([lateral]). With the introduction of LICENSE([lateral]) on the constraint ranking, the otherwise puzzling phenomenon of word-initial [l]-avoidance can be nicely explained. Also, by combining this strategy with syllable contact law and output-output correspondence, it will be shown that we can account for the whole range of relevant data. Clearly, the best aspect of this analysis is that it does not add any new stipulation to the theory but makes use of independently motivated constraints of universal nature, not to mention alleviating some burden on grammar thereby. © 2018 Institute for the Study of Language and Information.