earticle

영어

While works on pronominal anaphora resolution are well established, and there have been considerable works on the definite noun phrase anaphora, Associative Anaphora which by an entity can be referred to by a definite referring expression without previously being mentioned in
the text. It is realized in text as an (definite)NP in Korean, and its referent is implicitly associated with previously introduced NP.
But this paper deals with the phenomenon known as Associative Anaphora in a critical perspective. Embedded into the variable categories developed in Kim Kwanghee(1997), the establishment of associative anaphora is unreasonable. Some reasons are as follows:
First, so called associative NP, anaphora, is a constant entity not variable one. Therefore, associative NP is introduce new informations into a textual situation. It is the point of difference between co-referential anaphora which leads old informations and associative anaphora. Second, a associative NP must be supported by an accessible precedence NP. Problems are arise here. The lexical lists derived by association are different depending on speakers and hearers. In this case the context may offer several pairs of ‘NP-NP’ candidates for the role
of antecedent and associative anaphora. So there may be no consistency.
We reach the conclusion from these several observations that associative anaphora must be treated as associative inference or