Retour vers liste

Détail de la contribution

Auteur: Ji-Hye KIM

Co-Auteur(s): James Yoon

TSC (Tensed S Condition) and Korean Anaphors: An Experimental Study of Binding Domain and VP ellipsis in Korean

Abstract/Résumé: Standard BT assumes that the binding domain (BD) is defined by the conjunction of SSC (Specified Subject Condition) and TSC (Tensed S Condition). However, the failure of local anaphors to violate TSC can also be explained by the fact that some languages (English) lack nominative anaphors (Pollard and Sag 1992). When this complication is controlled for, TSC-violating anaphors can become acceptable. (1) John said that [pictures of himself] are on display Chomsky (1981) assumed that TSC holds, but allowed these to be a principled exception to TSC, while others (Pollard and Sag 1992) have argued that TSC-violating anaphors are licensed as exempt anaphors. Evidence suggesting the latter view is correct comes from the fact that TSC-violating anaphors allow both strict and sloppy readings under VP-ellipsis while locally bound anaphors admit only sloppy readings. (2) a. John is proud of himself. So is Bill (=is proud of Bill/*?John) b. John said that pictures of himself are on display. So did B (=said that pictures of Bill/John are on display) In languages like Korean and Japanese, all anaphors can occur as nominative-marked embedded subjects. Therefore, it has been assumed that TSC plays no role in defining the BD in Korean: Long-distance anaphors (caki, casin) have the added ability to violate SSC, but TSC is ineffective even for local anaphors (caki-casin, pronoun-casin). However, given that TSC-violations are allowed as long as the anaphor can be construed as exempt, it might be that some or all of TSC-violating anaphors in Korean are exempt anaphors. If they are, we expect strict readings to become available, while if they are core anaphors, there should be no difference in the rate of strict readings between TSC-violating and local anaphors, and sloppy readings should be dominant for both. In this study, we asked 40 Korean speakers to judge the acceptability of sentences containing different types of binding (local, TSC-violating, SSC-violating) with 4 anaphors (caki, casin, caki-casin, pronoun-casin). We then asked them to rate the likelihood of strict/sloppy readings in a subsequent sentence containing VP-ellipsis. Overall results showed that the complex anaphors (caki-casin and pronoun-casin) and the simple anaphors (caki and casin) showed significantly increased strict readings in TSC and SSC conditions compared to the case of local binding, which implies that TSC-violating Korean anaphors are treated as exempt anaphors. This finding is contrary to previously held assumptions about the irrelevance of TSC for all anaphors in Korean and supports the Exempt Binding approach for the determination of BD, over that of standard BT (Chomsky 1981) or theories that countenance GC Parameterization.