The present study analyzes various types of tensification in Korean within the framework of Optimality Theory with candidate chains (McCarthy 2006, 2007). Two important issues are addressed in this analysis: (1) Korean displays more than one type of t ...
The present study analyzes various types of tensification in Korean within the framework of Optimality Theory with candidate chains (McCarthy 2006, 2007). Two important issues are addressed in this analysis: (1) Korean displays more than one type of tensification: post-obstruent tensification, compound tensification, Sino-Korean tensification, modifier tensification, and predicate tensification (Kim 1967; Kim-Renaud 1974; Chung 1980; Ahn 1985, 1998, 2002; Oh 1988). (2) Compound tensification, Sino-Korean tensification, modifier tensification, and predicate tensification are characterized as having counterbleeding opacity in which the triggers are not visible in the surface structure.
In this paper, the historically motivated latent segment Fobstruent is proposed, triggering tensification in compound tensification, Sino-Korean tensification, modifier tensification, and predicate tensification. (1) As for compound tensification, the genetive, locative, temporal, and benefactive marker of the first word, 叱, developed into siasios (ㅅ), is proposed in the underlying representation (Nam 1971, Oh 1988). (2) In cases of modifier tensification, it is suggested that tensification is attributed to the predicate modifier in the Middle Korean -l (-ㄹᅙ ). (3) In turn, this paper posits that the coda /t/ of Ancient Chinese words was systematically realized as [l] in loanword adaptation in Ancient Korean, based on Choi (1999). (4) The speaker's attempt to differentiate the same phonetic forms as in /anta/ 'to embarrass' and /anta/ 'to know' (present) results in tensification.
The present study then proposes that (1) there is the latent (floating) feature F Obstruent in the underlying representation, based on historical data; it entails tensification. The paper further proposes that (2) the precedence constraint PREC (IDENT[tense], MAX) can account for the counterbleeding relation in tensification.